All Downloads are FREE. Search and download functionalities are using the official Maven repository.

io.reactivex.rxjava3.core.Flowable Maven / Gradle / Ivy

There is a newer version: 3.1.9
Show newest version
/*
 * Copyright (c) 2016-present, RxJava Contributors.
 *
 * Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in
 * compliance with the License. You may obtain a copy of the License at
 *
 * http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software distributed under the License is
 * distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See
 * the License for the specific language governing permissions and limitations under the License.
 */

package io.reactivex.rxjava3.core;

import java.util.*;
import java.util.concurrent.*;
import java.util.stream.*;

import org.reactivestreams.*;

import io.reactivex.rxjava3.annotations.*;
import io.reactivex.rxjava3.disposables.*;
import io.reactivex.rxjava3.exceptions.*;
import io.reactivex.rxjava3.flowables.*;
import io.reactivex.rxjava3.functions.*;
import io.reactivex.rxjava3.internal.functions.*;
import io.reactivex.rxjava3.internal.jdk8.*;
import io.reactivex.rxjava3.internal.operators.flowable.*;
import io.reactivex.rxjava3.internal.operators.maybe.MaybeToFlowable;
import io.reactivex.rxjava3.internal.operators.mixed.*;
import io.reactivex.rxjava3.internal.operators.observable.ObservableFromPublisher;
import io.reactivex.rxjava3.internal.operators.single.SingleToFlowable;
import io.reactivex.rxjava3.internal.schedulers.ImmediateThinScheduler;
import io.reactivex.rxjava3.internal.subscribers.*;
import io.reactivex.rxjava3.internal.util.*;
import io.reactivex.rxjava3.operators.ScalarSupplier;
import io.reactivex.rxjava3.parallel.ParallelFlowable;
import io.reactivex.rxjava3.plugins.RxJavaPlugins;
import io.reactivex.rxjava3.schedulers.*;
import io.reactivex.rxjava3.subscribers.*;

/**
 * The {@code Flowable} class that implements the Reactive Streams {@link Publisher}
 * Pattern and offers factory methods, intermediate operators and the ability to consume reactive dataflows.
 * 

* Reactive Streams operates with {@code Publisher}s which {@code Flowable} extends. Many operators * therefore accept general {@code Publisher}s directly and allow direct interoperation with other * Reactive Streams implementations. *

* The {@code Flowable} hosts the default buffer size of 128 elements for operators, accessible via {@link #bufferSize()}, * that can be overridden globally via the system parameter {@code rx3.buffer-size}. Most operators, however, have * overloads that allow setting their internal buffer size explicitly. *

* The documentation for this class makes use of marble diagrams. The following legend explains these diagrams: *

* *

* The {@code Flowable} follows the protocol *


 *      onSubscribe onNext* (onError | onComplete)?
 * 
* where the stream can be disposed through the {@link Subscription} instance provided to consumers through * {@link Subscriber#onSubscribe(Subscription)}. * Unlike the {@code Observable.subscribe()} of version 1.x, {@link #subscribe(Subscriber)} does not allow external cancellation * of a subscription and the {@link Subscriber} instance is expected to expose such capability if needed. *

* {@code Flowable}s support backpressure and require {@code Subscriber}s to signal demand via {@link Subscription#request(long)}. *

* Example: *


 * Disposable d = Flowable.just("Hello world!")
 *     .delay(1, TimeUnit.SECONDS)
 *     .subscribeWith(new DisposableSubscriber<String>() {
 *         @Override public void onStart() {
 *             System.out.println("Start!");
 *             request(1);
 *         }
 *         @Override public void onNext(String t) {
 *             System.out.println(t);
 *             request(1);
 *         }
 *         @Override public void onError(Throwable t) {
 *             t.printStackTrace();
 *         }
 *         @Override public void onComplete() {
 *             System.out.println("Done!");
 *         }
 *     });
 *
 * Thread.sleep(500);
 * // the sequence can now be cancelled via dispose()
 * d.dispose();
 * 
*

* The Reactive Streams specification is relatively strict when defining interactions between {@code Publisher}s and {@code Subscriber}s, so much so * that there is a significant performance penalty due certain timing requirements and the need to prepare for invalid * request amounts via {@link Subscription#request(long)}. * Therefore, RxJava has introduced the {@link FlowableSubscriber} interface that indicates the consumer can be driven with relaxed rules. * All RxJava operators are implemented with these relaxed rules in mind. * If the subscribing {@code Subscriber} does not implement this interface, for example, due to it being from another Reactive Streams compliant * library, the {@code Flowable} will automatically apply a compliance wrapper around it. *

* {@code Flowable} is an abstract class, but it is not advised to implement sources and custom operators by extending the class directly due * to the large amounts of Reactive Streams * rules to be followed to the letter. See the wiki for * some guidance if such custom implementations are necessary. *

* The recommended way of creating custom {@code Flowable}s is by using the {@link #create(FlowableOnSubscribe, BackpressureStrategy)} factory method: *


 * Flowable<String> source = Flowable.create(new FlowableOnSubscribe<String>() {
 *     @Override
 *     public void subscribe(FlowableEmitter<String> emitter) throws Exception {
 *
 *         // signal an item
 *         emitter.onNext("Hello");
 *
 *         // could be some blocking operation
 *         Thread.sleep(1000);
 *
 *         // the consumer might have cancelled the flow
 *         if (emitter.isCancelled()) {
 *             return;
 *         }
 *
 *         emitter.onNext("World");
 *
 *         Thread.sleep(1000);
 *
 *         // the end-of-sequence has to be signaled, otherwise the
 *         // consumers may never finish
 *         emitter.onComplete();
 *     }
 * }, BackpressureStrategy.BUFFER);
 *
 * System.out.println("Subscribe!");
 *
 * source.subscribe(System.out::println);
 *
 * System.out.println("Done!");
 * 
*

* RxJava reactive sources, such as {@code Flowable}, are generally synchronous and sequential in nature. In the ReactiveX design, the location (thread) * where operators run is orthogonal to when the operators can work with data. This means that asynchrony and parallelism * has to be explicitly expressed via operators such as {@link #subscribeOn(Scheduler)}, {@link #observeOn(Scheduler)} and {@link #parallel()}. In general, * operators featuring a {@link Scheduler} parameter are introducing this type of asynchrony into the flow. *

* For more information see the ReactiveX documentation. * * @param * the type of the items emitted by the {@code Flowable} * @see Observable * @see ParallelFlowable * @see io.reactivex.rxjava3.subscribers.DisposableSubscriber */ public abstract class Flowable<@NonNull T> implements Publisher { /** The default buffer size. */ static final int BUFFER_SIZE; static { BUFFER_SIZE = Math.max(1, Integer.getInteger("rx3.buffer-size", 128)); } /** * Mirrors the one {@link Publisher} in an {@link Iterable} of several {@code Publisher}s that first either emits an item or sends * a termination notification. *

* *

* When one of the {@code Publisher}s signal an item or terminates first, all subscriptions to the other * {@code Publisher}s are canceled. *

*
Backpressure:
*
The operator itself doesn't interfere with backpressure which is determined by the winning * {@code Publisher}'s backpressure behavior.
*
Scheduler:
*
{@code amb} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
* If any of the losing {@code Publisher}s signals an error, the error is routed to the global * error handler via {@link RxJavaPlugins#onError(Throwable)}. *
*
* * @param the common element type * @param sources * an {@code Iterable} of {@code Publisher}s sources competing to react first. A subscription to each {@code Publisher} will * occur in the same order as in this {@code Iterable}. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Amb */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable amb(@NonNull Iterable<@NonNull ? extends Publisher> sources) { Objects.requireNonNull(sources, "sources is null"); return RxJavaPlugins.onAssembly(new FlowableAmb<>(null, sources)); } /** * Mirrors the one {@link Publisher} in an array of several {@code Publisher}s that first either emits an item or sends * a termination notification. *

* *

* When one of the {@code Publisher}s signal an item or terminates first, all subscriptions to the other * {@code Publisher}s are canceled. *

*
Backpressure:
*
The operator itself doesn't interfere with backpressure which is determined by the winning * {@code Publisher}'s backpressure behavior.
*
Scheduler:
*
{@code ambArray} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
* If any of the losing {@code Publisher}s signals an error, the error is routed to the global * error handler via {@link RxJavaPlugins#onError(Throwable)}. *
*
* * @param the common element type * @param sources * an array of {@code Publisher} sources competing to react first. A subscription to each {@code Publisher} will * occur in the same order as in this array. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Amb */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs public static <@NonNull T> Flowable ambArray(@NonNull Publisher... sources) { Objects.requireNonNull(sources, "sources is null"); int len = sources.length; if (len == 0) { return empty(); } else if (len == 1) { return fromPublisher(sources[0]); } return RxJavaPlugins.onAssembly(new FlowableAmb<>(sources, null)); } /** * Returns the default internal buffer size used by most async operators. *

The value can be overridden via system parameter {@code rx3.buffer-size} * before the {@code Flowable} class is loaded. * @return the default internal buffer size. */ @CheckReturnValue public static int bufferSize() { return BUFFER_SIZE; } /** * Combines a collection of source {@link Publisher}s by emitting an item that aggregates the latest values of each of * the source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* Note on method signature: since Java doesn't allow creating a generic array with {@code new T[]}, the * implementation of this operator has to create an {@code Object[]} instead. Unfortunately, a * {@code Function} passed to the method would trigger a {@link ClassCastException}. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* If the provided array of source {@code Publisher}s is empty, the resulting sequence completes immediately without emitting * any items and without any calls to the combiner function. * *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatestArray} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the common base type of source values * @param * the result type * @param sources * the collection of source {@code Publisher}s * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @NonNull public static <@NonNull T, @NonNull R> Flowable combineLatestArray(@NonNull Publisher[] sources, @NonNull Function combiner) { return combineLatestArray(sources, combiner, bufferSize()); } /** * Combines a collection of source {@link Publisher}s by emitting an item that aggregates the latest values of each of * the source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* Note on method signature: since Java doesn't allow creating a generic array with {@code new T[]}, the * implementation of this operator has to create an {@code Object[]} instead. Unfortunately, a * {@code Function} passed to the method would trigger a {@link ClassCastException}. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* If the provided array of source {@code Publisher}s is empty, the resulting sequence completes immediately without emitting * any items and without any calls to the combiner function. * *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatestArray} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the common base type of source values * @param * the result type * @param sources * the collection of source {@code Publisher}s * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @param bufferSize * the internal buffer size and prefetch amount applied to every source {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code combiner} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: CombineLatest */ @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) public static <@NonNull T, @NonNull R> Flowable combineLatestArray(@NonNull Publisher[] sources, @NonNull Function combiner, int bufferSize) { Objects.requireNonNull(sources, "sources is null"); if (sources.length == 0) { return empty(); } Objects.requireNonNull(combiner, "combiner is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableCombineLatest<>(sources, combiner, bufferSize, false)); } /** * Combines a collection of source {@link Publisher}s by emitting an item that aggregates the latest values of each of * the source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* Note on method signature: since Java doesn't allow creating a generic array with {@code new T[]}, the * implementation of this operator has to create an {@code Object[]} instead. Unfortunately, a * {@code Function} passed to the method would trigger a {@link ClassCastException}. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* If the provided iterable of source {@code Publisher}s is empty, the resulting sequence completes immediately without emitting * any items and without any calls to the combiner function. * *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the common base type of source values * @param * the result type * @param sources * the collection of source {@code Publisher}s * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @NonNull public static <@NonNull T, @NonNull R> Flowable combineLatest(@NonNull Iterable<@NonNull ? extends Publisher> sources, @NonNull Function combiner) { return combineLatest(sources, combiner, bufferSize()); } /** * Combines a collection of source {@link Publisher}s by emitting an item that aggregates the latest values of each of * the source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* Note on method signature: since Java doesn't allow creating a generic array with {@code new T[]}, the * implementation of this operator has to create an {@code Object[]} instead. Unfortunately, a * {@code Function} passed to the method would trigger a {@link ClassCastException}. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* If the provided iterable of source {@code Publisher}s is empty, the resulting sequence completes immediately without emitting any items and * without any calls to the combiner function. * *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the common base type of source values * @param * the result type * @param sources * the collection of source {@code Publisher}s * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @param bufferSize * the internal buffer size and prefetch amount applied to every source {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code combiner} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: CombineLatest */ @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) public static <@NonNull T, @NonNull R> Flowable combineLatest(@NonNull Iterable<@NonNull ? extends Publisher> sources, @NonNull Function combiner, int bufferSize) { Objects.requireNonNull(sources, "sources is null"); Objects.requireNonNull(combiner, "combiner is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableCombineLatest<>(sources, combiner, bufferSize, false)); } /** * Combines a collection of source {@link Publisher}s by emitting an item that aggregates the latest values of each of * the source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* Note on method signature: since Java doesn't allow creating a generic array with {@code new T[]}, the * implementation of this operator has to create an {@code Object[]} instead. Unfortunately, a * {@code Function} passed to the method would trigger a {@link ClassCastException}. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* If the provided array of source {@code Publisher}s is empty, the resulting sequence completes immediately without emitting * any items and without any calls to the combiner function. * *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatestArrayDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the common base type of source values * @param * the result type * @param sources * the collection of source {@code Publisher}s * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @NonNull public static <@NonNull T, @NonNull R> Flowable combineLatestArrayDelayError(@NonNull Publisher[] sources, @NonNull Function combiner) { return combineLatestArrayDelayError(sources, combiner, bufferSize()); } /** * Combines a collection of source {@link Publisher}s by emitting an item that aggregates the latest values of each of * the source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function and delays any error from the sources until * all source {@code Publisher}s terminate. *

* Note on method signature: since Java doesn't allow creating a generic array with {@code new T[]}, the * implementation of this operator has to create an {@code Object[]} instead. Unfortunately, a * {@code Function} passed to the method would trigger a {@link ClassCastException}. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* If the provided array of source {@code Publisher}s is empty, the resulting sequence completes immediately without emitting * any items and without any calls to the combiner function. * *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatestArrayDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the common base type of source values * @param * the result type * @param sources * the collection of source {@code Publisher}s * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @param bufferSize * the internal buffer size and prefetch amount applied to every source {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code combiner} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: CombineLatest */ @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) public static <@NonNull T, @NonNull R> Flowable combineLatestArrayDelayError(@NonNull Publisher[] sources, @NonNull Function combiner, int bufferSize) { Objects.requireNonNull(sources, "sources is null"); Objects.requireNonNull(combiner, "combiner is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); if (sources.length == 0) { return empty(); } return RxJavaPlugins.onAssembly(new FlowableCombineLatest<>(sources, combiner, bufferSize, true)); } /** * Combines a collection of source {@link Publisher}s by emitting an item that aggregates the latest values of each of * the source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function and delays any error from the sources until * all source {@code Publisher}s terminate. *

* Note on method signature: since Java doesn't allow creating a generic array with {@code new T[]}, the * implementation of this operator has to create an {@code Object[]} instead. Unfortunately, a * {@code Function} passed to the method would trigger a {@link ClassCastException}. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* If the provided iterable of source {@code Publisher}s is empty, the resulting sequence completes immediately without emitting * any items and without any calls to the combiner function. * *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatestDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the common base type of source values * @param * the result type * @param sources * the collection of source {@code Publisher}s * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @NonNull public static <@NonNull T, @NonNull R> Flowable combineLatestDelayError(@NonNull Iterable<@NonNull ? extends Publisher> sources, @NonNull Function combiner) { return combineLatestDelayError(sources, combiner, bufferSize()); } /** * Combines a collection of source {@link Publisher}s by emitting an item that aggregates the latest values of each of * the source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function and delays any error from the sources until * all source {@code Publisher}s terminate. *

* Note on method signature: since Java doesn't allow creating a generic array with {@code new T[]}, the * implementation of this operator has to create an {@code Object[]} instead. Unfortunately, a * {@code Function} passed to the method would trigger a {@link ClassCastException}. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* If the provided iterable of source {@code Publisher}s is empty, the resulting sequence completes immediately without emitting * any items and without any calls to the combiner function. * *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatestDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the common base type of source values * @param * the result type * @param sources * the collection of source {@code Publisher}s * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @param bufferSize * the internal buffer size and prefetch amount applied to every source {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code combiner} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: CombineLatest */ @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @NonNull public static <@NonNull T, @NonNull R> Flowable combineLatestDelayError(@NonNull Iterable<@NonNull ? extends Publisher> sources, @NonNull Function combiner, int bufferSize) { Objects.requireNonNull(sources, "sources is null"); Objects.requireNonNull(combiner, "combiner is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableCombineLatest<>(sources, combiner, bufferSize, true)); } /** * Combines two source {@link Publisher}s by emitting an item that aggregates the latest values of each of the * source {@code Publisher}s each time an item is received from either of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the first source * @param the element type of the second source * @param the combined output type * @param source1 * the first source {@code Publisher} * @param source2 * the second source {@code Publisher} * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SuppressWarnings("unchecked") @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T1, @NonNull T2, @NonNull R> Flowable combineLatest( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull BiFunction combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(combiner, "combiner is null"); return combineLatestArray(new Publisher[] { source1, source2 }, Functions.toFunction(combiner), bufferSize()); } /** * Combines three source {@link Publisher}s by emitting an item that aggregates the latest values of each of the * source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the first source * @param the element type of the second source * @param the element type of the third source * @param the combined output type * @param source1 * the first source {@code Publisher} * @param source2 * the second source {@code Publisher} * @param source3 * the third source {@code Publisher} * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SuppressWarnings("unchecked") @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull R> Flowable combineLatest( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Function3 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(combiner, "combiner is null"); return combineLatestArray(new Publisher[] { source1, source2, source3 }, Functions.toFunction(combiner), bufferSize()); } /** * Combines four source {@link Publisher}s by emitting an item that aggregates the latest values of each of the * source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the first source * @param the element type of the second source * @param the element type of the third source * @param the element type of the fourth source * @param the combined output type * @param source1 * the first source {@code Publisher} * @param source2 * the second source {@code Publisher} * @param source3 * the third source {@code Publisher} * @param source4 * the fourth source {@code Publisher} * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SuppressWarnings("unchecked") @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull R> Flowable combineLatest( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Function4 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(combiner, "combiner is null"); return combineLatestArray(new Publisher[] { source1, source2, source3, source4 }, Functions.toFunction(combiner), bufferSize()); } /** * Combines five source {@link Publisher}s by emitting an item that aggregates the latest values of each of the * source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the first source * @param the element type of the second source * @param the element type of the third source * @param the element type of the fourth source * @param the element type of the fifth source * @param the combined output type * @param source1 * the first source {@code Publisher} * @param source2 * the second source {@code Publisher} * @param source3 * the third source {@code Publisher} * @param source4 * the fourth source {@code Publisher} * @param source5 * the fifth source {@code Publisher} * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SuppressWarnings("unchecked") @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull R> Flowable combineLatest( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Function5 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(combiner, "combiner is null"); return combineLatestArray(new Publisher[] { source1, source2, source3, source4, source5 }, Functions.toFunction(combiner), bufferSize()); } /** * Combines six source {@link Publisher}s by emitting an item that aggregates the latest values of each of the * source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the first source * @param the element type of the second source * @param the element type of the third source * @param the element type of the fourth source * @param the element type of the fifth source * @param the element type of the sixth source * @param the combined output type * @param source1 * the first source {@code Publisher} * @param source2 * the second source {@code Publisher} * @param source3 * the third source {@code Publisher} * @param source4 * the fourth source {@code Publisher} * @param source5 * the fifth source {@code Publisher} * @param source6 * the sixth source {@code Publisher} * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5}, {@code source6} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SuppressWarnings("unchecked") @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull T6, @NonNull R> Flowable combineLatest( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Publisher source6, @NonNull Function6 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(source6, "source6 is null"); Objects.requireNonNull(combiner, "combiner is null"); return combineLatestArray(new Publisher[] { source1, source2, source3, source4, source5, source6 }, Functions.toFunction(combiner), bufferSize()); } /** * Combines seven source {@link Publisher}s by emitting an item that aggregates the latest values of each of the * source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the first source * @param the element type of the second source * @param the element type of the third source * @param the element type of the fourth source * @param the element type of the fifth source * @param the element type of the sixth source * @param the element type of the seventh source * @param the combined output type * @param source1 * the first source {@code Publisher} * @param source2 * the second source {@code Publisher} * @param source3 * the third source {@code Publisher} * @param source4 * the fourth source {@code Publisher} * @param source5 * the fifth source {@code Publisher} * @param source6 * the sixth source {@code Publisher} * @param source7 * the seventh source {@code Publisher} * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5}, {@code source6}, * {@code source7} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SuppressWarnings("unchecked") @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull T6, @NonNull T7, @NonNull R> Flowable combineLatest( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Publisher source6, @NonNull Publisher source7, @NonNull Function7 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(source6, "source6 is null"); Objects.requireNonNull(source7, "source7 is null"); Objects.requireNonNull(combiner, "combiner is null"); return combineLatestArray(new Publisher[] { source1, source2, source3, source4, source5, source6, source7 }, Functions.toFunction(combiner), bufferSize()); } /** * Combines eight source {@link Publisher}s by emitting an item that aggregates the latest values of each of the * source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the first source * @param the element type of the second source * @param the element type of the third source * @param the element type of the fourth source * @param the element type of the fifth source * @param the element type of the sixth source * @param the element type of the seventh source * @param the element type of the eighth source * @param the combined output type * @param source1 * the first source {@code Publisher} * @param source2 * the second source {@code Publisher} * @param source3 * the third source {@code Publisher} * @param source4 * the fourth source {@code Publisher} * @param source5 * the fifth source {@code Publisher} * @param source6 * the sixth source {@code Publisher} * @param source7 * the seventh source {@code Publisher} * @param source8 * the eighth source {@code Publisher} * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5}, {@code source6}, * {@code source7}, {@code source8} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SuppressWarnings("unchecked") @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull T6, @NonNull T7, @NonNull T8, @NonNull R> Flowable combineLatest( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Publisher source6, @NonNull Publisher source7, @NonNull Publisher source8, @NonNull Function8 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(source6, "source6 is null"); Objects.requireNonNull(source7, "source7 is null"); Objects.requireNonNull(source8, "source8 is null"); Objects.requireNonNull(combiner, "combiner is null"); return combineLatestArray(new Publisher[] { source1, source2, source3, source4, source5, source6, source7, source8 }, Functions.toFunction(combiner), bufferSize()); } /** * Combines nine source {@link Publisher}s by emitting an item that aggregates the latest values of each of the * source {@code Publisher}s each time an item is received from any of the source {@code Publisher}s, where this * aggregation is defined by a specified function. *

* If any of the sources never produces an item but only terminates (normally or with an error), the * resulting sequence terminates immediately (normally or with all the errors accumulated until that point). * If that input source is also synchronous, other sources after it will not be subscribed to. *

* *

*
Backpressure:
*
The returned {@code Publisher} honors backpressure from downstream. The source {@code Publisher}s * are requested in a bounded manner, however, their backpressure is not enforced (the operator won't signal * {@link MissingBackpressureException}) and may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code combineLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the first source * @param the element type of the second source * @param the element type of the third source * @param the element type of the fourth source * @param the element type of the fifth source * @param the element type of the sixth source * @param the element type of the seventh source * @param the element type of the eighth source * @param the element type of the ninth source * @param the combined output type * @param source1 * the first source {@code Publisher} * @param source2 * the second source {@code Publisher} * @param source3 * the third source {@code Publisher} * @param source4 * the fourth source {@code Publisher} * @param source5 * the fifth source {@code Publisher} * @param source6 * the sixth source {@code Publisher} * @param source7 * the seventh source {@code Publisher} * @param source8 * the eighth source {@code Publisher} * @param source9 * the ninth source {@code Publisher} * @param combiner * the aggregation function used to combine the items emitted by the source {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5}, {@code source6}, * {@code source7}, {@code source8}, {@code source9} * or {@code combiner} is {@code null} * @see ReactiveX operators documentation: CombineLatest */ @SuppressWarnings("unchecked") @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull T6, @NonNull T7, @NonNull T8, @NonNull T9, @NonNull R> Flowable combineLatest( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Publisher source6, @NonNull Publisher source7, @NonNull Publisher source8, @NonNull Publisher source9, @NonNull Function9 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(source6, "source6 is null"); Objects.requireNonNull(source7, "source7 is null"); Objects.requireNonNull(source8, "source8 is null"); Objects.requireNonNull(source9, "source9 is null"); Objects.requireNonNull(combiner, "combiner is null"); return combineLatestArray(new Publisher[] { source1, source2, source3, source4, source5, source6, source7, source8, source9 }, Functions.toFunction(combiner), bufferSize()); } /** * Concatenates elements of each {@link Publisher} provided via an {@link Iterable} sequence into a single sequence * of elements without interleaving them. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, it may throw an * {@link IllegalStateException} when that {@code Publisher} completes.
*
Scheduler:
*
{@code concat} does not operate by default on a particular {@link Scheduler}.
*
* @param the common value type of the sources * @param sources the {@code Iterable} sequence of {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable concat(@NonNull Iterable<@NonNull ? extends Publisher> sources) { Objects.requireNonNull(sources, "sources is null"); // unlike general sources, fromIterable can only throw on a boundary because it is consumed only there return fromIterable(sources).concatMapDelayError((Function)Functions.identity(), false, 2); } /** * Returns a {@code Flowable} that emits the items emitted by each of the {@link Publisher}s emitted by the source * {@code Publisher}, one after the other, without interleaving them. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. Both the outer and inner {@code Publisher} * sources are expected to honor backpressure as well. If the outer violates this, a * {@link MissingBackpressureException} is signaled. If any of the inner {@code Publisher}s violates * this, it may throw an {@link IllegalStateException} when an inner {@code Publisher} completes.
*
Scheduler:
*
{@code concat} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * a {@code Publisher} that emits {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Concat */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable concat(@NonNull Publisher<@NonNull ? extends Publisher> sources) { return concat(sources, bufferSize()); } /** * Returns a {@code Flowable} that emits the items emitted by each of the {@link Publisher}s emitted by the source * {@code Publisher}, one after the other, without interleaving them. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. Both the outer and inner {@code Publisher} * sources are expected to honor backpressure as well. If the outer violates this, a * {@link MissingBackpressureException} is signaled. If any of the inner {@code Publisher}s violates * this, it may throw an {@link IllegalStateException} when an inner {@code Publisher} completes.
*
Scheduler:
*
{@code concat} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * a {@code Publisher} that emits {@code Publisher}s * @param prefetch * the number of {@code Publisher}s to prefetch from the sources sequence. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see ReactiveX operators documentation: Concat */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable concat(@NonNull Publisher<@NonNull ? extends Publisher> sources, int prefetch) { return fromPublisher(sources).concatMap((Function)Functions.identity(), prefetch); } /** * Returns a {@code Flowable} that emits the items emitted by two {@link Publisher}s, one after the other, without * interleaving them. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, it may throw an * {@link IllegalStateException} when that source {@code Publisher} completes.
*
Scheduler:
*
{@code concat} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be concatenated * @param source2 * a {@code Publisher} to be concatenated * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1} or {@code source2} is {@code null} * @see ReactiveX operators documentation: Concat */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable concat(@NonNull Publisher source1, @NonNull Publisher source2) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); return concatArray(source1, source2); } /** * Returns a {@code Flowable} that emits the items emitted by three {@link Publisher}s, one after the other, without * interleaving them. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, it may throw an * {@link IllegalStateException} when that source {@code Publisher} completes.
*
Scheduler:
*
{@code concat} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be concatenated * @param source2 * a {@code Publisher} to be concatenated * @param source3 * a {@code Publisher} to be concatenated * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code source3} is {@code null} * @see ReactiveX operators documentation: Concat */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable concat( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); return concatArray(source1, source2, source3); } /** * Returns a {@code Flowable} that emits the items emitted by four {@link Publisher}s, one after the other, without * interleaving them. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, it may throw an * {@link IllegalStateException} when that source {@code Publisher} completes.
*
Scheduler:
*
{@code concat} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be concatenated * @param source2 * a {@code Publisher} to be concatenated * @param source3 * a {@code Publisher} to be concatenated * @param source4 * a {@code Publisher} to be concatenated * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3} or {@code source4} is {@code null} * @see ReactiveX operators documentation: Concat */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable concat( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); return concatArray(source1, source2, source3, source4); } /** * Concatenates a variable number of {@link Publisher} sources. *

* Note: named this way because of overload conflict with {@code concat(Publisher>}). *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, it may throw an * {@link IllegalStateException} when that source {@code Publisher} completes.
*
Scheduler:
*
{@code concatArray} does not operate by default on a particular {@link Scheduler}.
*
* @param sources the array of source {@code Publisher}s * @param the common base value type * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs @NonNull public static <@NonNull T> Flowable concatArray(@NonNull Publisher... sources) { Objects.requireNonNull(sources, "sources is null"); if (sources.length == 0) { return empty(); } else if (sources.length == 1) { return fromPublisher(sources[0]); } return RxJavaPlugins.onAssembly(new FlowableConcatArray<>(sources, false)); } /** * Concatenates a variable number of {@link Publisher} sources and delays errors from any of them * till all terminate. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, it may throw an * {@link IllegalStateException} when that source {@code Publisher} completes.
*
Scheduler:
*
{@code concatArrayDelayError} does not operate by default on a particular {@link Scheduler}.
*
* @param sources the array of source {@code Publisher}s * @param the common base value type * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs @NonNull public static <@NonNull T> Flowable concatArrayDelayError(@NonNull Publisher... sources) { Objects.requireNonNull(sources, "sources is null"); if (sources.length == 0) { return empty(); } else if (sources.length == 1) { return fromPublisher(sources[0]); } return RxJavaPlugins.onAssembly(new FlowableConcatArray<>(sources, true)); } /** * Concatenates an array of {@link Publisher}s eagerly into a single stream of values. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * source {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them * in order, each one after the previous one completes. *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, the operator will signal a * {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources an array of {@code Publisher}s that need to be eagerly concatenated * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs @NonNull public static <@NonNull T> Flowable concatArrayEager(@NonNull Publisher... sources) { return concatArrayEager(bufferSize(), bufferSize(), sources); } /** * Concatenates an array of {@link Publisher}s eagerly into a single stream of values. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * source {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them * in order, each one after the previous one completes. *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, the operator will signal a * {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources an array of {@code Publisher}s that need to be eagerly concatenated * @param maxConcurrency the maximum number of concurrent subscriptions at a time, {@link Integer#MAX_VALUE} * is interpreted as an indication to subscribe to all sources at once * @param prefetch the number of elements to prefetch from each {@code Publisher} source * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code prefetch} is non-positive * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SuppressWarnings({ "rawtypes", "unchecked" }) @SafeVarargs public static <@NonNull T> Flowable concatArrayEager(int maxConcurrency, int prefetch, @NonNull Publisher... sources) { Objects.requireNonNull(sources, "sources is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapEager(new FlowableFromArray(sources), Functions.identity(), maxConcurrency, prefetch, ErrorMode.IMMEDIATE)); } /** * Concatenates an array of {@link Publisher}s eagerly into a single stream of values * and delaying any errors until all sources terminate. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * source {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s * and then drains them in order, each one after the previous one completes. *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, the operator will signal a * {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources an array of {@code Publisher}s that need to be eagerly concatenated * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @since 2.2.1 - experimental */ @CheckReturnValue @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) @SafeVarargs @NonNull public static <@NonNull T> Flowable concatArrayEagerDelayError(@NonNull Publisher... sources) { return concatArrayEagerDelayError(bufferSize(), bufferSize(), sources); } /** * Concatenates an array of {@link Publisher}s eagerly into a single stream of values * and delaying any errors until all sources terminate. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * source {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s * and then drains them in order, each one after the previous one completes. *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the source {@code Publisher}s violate this, the operator will signal a * {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources an array of {@code Publisher}s that need to be eagerly concatenated * @param maxConcurrency the maximum number of concurrent subscriptions at a time, {@link Integer#MAX_VALUE} * is interpreted as indication to subscribe to all sources at once * @param prefetch the number of elements to prefetch from each {@code Publisher} source * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code prefetch} is non-positive * @since 2.2.1 - experimental */ @SuppressWarnings({ "rawtypes", "unchecked" }) @CheckReturnValue @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) @SafeVarargs @NonNull public static <@NonNull T> Flowable concatArrayEagerDelayError(int maxConcurrency, int prefetch, @NonNull Publisher... sources) { return fromArray(sources).concatMapEagerDelayError((Function)Functions.identity(), true, maxConcurrency, prefetch); } /** * Concatenates the {@link Iterable} sequence of {@link Publisher}s into a single sequence by subscribing to each {@code Publisher}, * one after the other, one at a time and delays any errors till the all inner {@code Publisher}s terminate. * *
*
Backpressure:
*
The operator honors backpressure from downstream. Both the outer and inner {@code Publisher} * sources are expected to honor backpressure as well. If the outer violates this, a * {@link MissingBackpressureException} is signaled. If any of the inner {@code Publisher}s violates * this, it may throw an {@link IllegalStateException} when an inner {@code Publisher} completes.
*
Scheduler:
*
{@code concatDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources the {@code Iterable} sequence of {@code Publisher}s * @return the new {@code Flowable} with the concatenating behavior * @throws NullPointerException if {@code sources} is {@code null} */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable concatDelayError(@NonNull Iterable<@NonNull ? extends Publisher> sources) { Objects.requireNonNull(sources, "sources is null"); return fromIterable(sources).concatMapDelayError((Function)Functions.identity()); } /** * Concatenates the {@link Publisher} sequence of {@code Publisher}s into a single sequence by subscribing to each inner {@code Publisher}, * one after the other, one at a time and delays any errors till the all inner and the outer {@code Publisher}s terminate. * *
*
Backpressure:
*
{@code concatDelayError} fully supports backpressure.
*
Scheduler:
*
{@code concatDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources the {@code Publisher} sequence of {@code Publisher}s * @return the new {@code Flowable} with the concatenating behavior * @throws NullPointerException if {@code sources} is {@code null} */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable concatDelayError(@NonNull Publisher<@NonNull ? extends Publisher> sources) { return concatDelayError(sources, bufferSize(), true); } /** * Concatenates the {@link Publisher} sequence of {@code Publisher}s into a single sequence by subscribing to each inner {@code Publisher}, * one after the other, one at a time and delays any errors till the all inner and the outer {@code Publisher}s terminate. * *
*
Backpressure:
*
{@code concatDelayError} fully supports backpressure.
*
Scheduler:
*
{@code concatDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources the {@code Publisher} sequence of {@code Publisher}s * @param prefetch the number of elements to prefetch from the outer {@code Publisher} * @param tillTheEnd if {@code true}, exceptions from the outer and all inner {@code Publisher}s are delayed to the end * if {@code false}, exception from the outer {@code Publisher} is delayed till the current inner {@code Publisher} terminates * @return the new {@code Flowable} with the concatenating behavior * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is {@code null} */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable concatDelayError(@NonNull Publisher<@NonNull ? extends Publisher> sources, int prefetch, boolean tillTheEnd) { return fromPublisher(sources).concatMapDelayError((Function)Functions.identity(), tillTheEnd, prefetch); } /** * Concatenates a sequence of {@link Publisher}s eagerly into a single stream of values. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * source {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them * in order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream and the inner {@code Publisher}s are * expected to support backpressure. Violating this assumption, the operator will * signal {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources a sequence of {@code Publisher}s that need to be eagerly concatenated * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable concatEager(@NonNull Iterable<@NonNull ? extends Publisher> sources) { return concatEager(sources, bufferSize(), bufferSize()); } /** * Concatenates a sequence of {@link Publisher}s eagerly into a single stream of values and * runs a limited number of inner sequences at once. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * source {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them * in order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream and both the outer and inner {@code Publisher}s are * expected to support backpressure. Violating this assumption, the operator will * signal {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources a sequence of {@code Publisher}s that need to be eagerly concatenated * @param maxConcurrency the maximum number of concurrently running inner {@code Publisher}s; {@link Integer#MAX_VALUE} * is interpreted as all inner {@code Publisher}s can be active at the same time * @param prefetch the number of elements to prefetch from each inner {@code Publisher} source * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code prefetch} is non-positive * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SuppressWarnings({ "rawtypes", "unchecked" }) public static <@NonNull T> Flowable concatEager(@NonNull Iterable<@NonNull ? extends Publisher> sources, int maxConcurrency, int prefetch) { Objects.requireNonNull(sources, "sources is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapEager(new FlowableFromIterable(sources), Functions.identity(), maxConcurrency, prefetch, ErrorMode.BOUNDARY)); } /** * Concatenates a {@link Publisher} sequence of {@code Publisher}s eagerly into a single stream of values. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * emitted source {@code Publisher}s as they are observed. The operator buffers the values emitted by these * {@code Publisher}s and then drains them in order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream and both the outer and inner {@code Publisher}s are * expected to support backpressure. Violating this assumption, the operator will * signal {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources a sequence of {@code Publisher}s that need to be eagerly concatenated * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable concatEager(@NonNull Publisher<@NonNull ? extends Publisher> sources) { return concatEager(sources, bufferSize(), bufferSize()); } /** * Concatenates a {@link Publisher} sequence of {@code Publisher}s eagerly into a single stream of values and * runs a limited number of inner sequences at once. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * emitted source {@code Publisher}s as they are observed. The operator buffers the values emitted by these * {@code Publisher}s and then drains them in order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream and both the outer and inner {@code Publisher}s are * expected to support backpressure. Violating this assumption, the operator will * signal {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources a sequence of {@code Publisher}s that need to be eagerly concatenated * @param maxConcurrency the maximum number of concurrently running inner {@code Publisher}s; {@link Integer#MAX_VALUE} * is interpreted as all inner {@code Publisher}s can be active at the same time * @param prefetch the number of elements to prefetch from each inner {@code Publisher} source * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code prefetch} is non-positive * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SuppressWarnings({ "rawtypes", "unchecked" }) public static <@NonNull T> Flowable concatEager(@NonNull Publisher<@NonNull ? extends Publisher> sources, int maxConcurrency, int prefetch) { Objects.requireNonNull(sources, "sources is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapEagerPublisher(sources, Functions.identity(), maxConcurrency, prefetch, ErrorMode.IMMEDIATE)); } /** * Concatenates a sequence of {@link Publisher}s eagerly into a single stream of values, * delaying errors until all the inner sequences terminate. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * source {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them * in order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream and the inner {@code Publisher}s are * expected to support backpressure. Violating this assumption, the operator will * signal {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources a sequence of {@code Publisher}s that need to be eagerly concatenated * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @since 3.0.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable concatEagerDelayError(@NonNull Iterable<@NonNull ? extends Publisher> sources) { return concatEagerDelayError(sources, bufferSize(), bufferSize()); } /** * Concatenates a sequence of {@link Publisher}s eagerly into a single stream of values, * delaying errors until all the inner sequences terminate and runs a limited number * of inner sequences at once. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * source {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them * in order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream and both the outer and inner {@code Publisher}s are * expected to support backpressure. Violating this assumption, the operator will * signal {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources a sequence of {@code Publisher}s that need to be eagerly concatenated * @param maxConcurrency the maximum number of concurrently running inner {@code Publisher}s; {@link Integer#MAX_VALUE} * is interpreted as all inner {@code Publisher}s can be active at the same time * @param prefetch the number of elements to prefetch from each inner {@code Publisher} source * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code prefetch} is non-positive * @since 3.0.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SuppressWarnings({ "rawtypes", "unchecked" }) public static <@NonNull T> Flowable concatEagerDelayError(@NonNull Iterable<@NonNull ? extends Publisher> sources, int maxConcurrency, int prefetch) { Objects.requireNonNull(sources, "sources is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapEager(new FlowableFromIterable(sources), Functions.identity(), maxConcurrency, prefetch, ErrorMode.END)); } /** * Concatenates a {@link Publisher} sequence of {@code Publisher}s eagerly into a single stream of values, * delaying errors until all the inner and the outer sequences terminate. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * emitted source {@code Publisher}s as they are observed. The operator buffers the values emitted by these * {@code Publisher}s and then drains them in order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream and both the outer and inner {@code Publisher}s are * expected to support backpressure. Violating this assumption, the operator will * signal {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources a sequence of {@code Publisher}s that need to be eagerly concatenated * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @since 3.0.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable concatEagerDelayError(@NonNull Publisher<@NonNull ? extends Publisher> sources) { return concatEagerDelayError(sources, bufferSize(), bufferSize()); } /** * Concatenates a {@link Publisher} sequence of {@code Publisher}s eagerly into a single stream of values, * delaying errors until all the inner and outer sequences terminate and runs a limited number of inner * sequences at once. *

* *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * emitted source {@code Publisher}s as they are observed. The operator buffers the values emitted by these * {@code Publisher}s and then drains them in order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream and both the outer and inner {@code Publisher}s are * expected to support backpressure. Violating this assumption, the operator will * signal {@link MissingBackpressureException}.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param sources a sequence of {@code Publisher}s that need to be eagerly concatenated * @param maxConcurrency the maximum number of concurrently running inner {@code Publisher}s; {@link Integer#MAX_VALUE} * is interpreted as all inner {@code Publisher}s can be active at the same time * @param prefetch the number of elements to prefetch from each inner {@code Publisher} source * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code prefetch} is non-positive * @since 3.0.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SuppressWarnings({ "rawtypes", "unchecked" }) public static <@NonNull T> Flowable concatEagerDelayError(@NonNull Publisher<@NonNull ? extends Publisher> sources, int maxConcurrency, int prefetch) { Objects.requireNonNull(sources, "sources is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapEagerPublisher(sources, Functions.identity(), maxConcurrency, prefetch, ErrorMode.END)); } /** * Provides an API (via a cold {@code Flowable}) that bridges the reactive world with the callback-style, * generally non-backpressured world. *

* Example: *


     * Flowable.<Event>create(emitter -> {
     *     Callback listener = new Callback() {
     *         @Override
     *         public void onEvent(Event e) {
     *             emitter.onNext(e);
     *             if (e.isLast()) {
     *                 emitter.onComplete();
     *             }
     *         }
     *
     *         @Override
     *         public void onFailure(Exception e) {
     *             emitter.onError(e);
     *         }
     *     };
     *
     *     AutoCloseable c = api.someMethod(listener);
     *
     *     emitter.setCancellable(c::close);
     *
     * }, BackpressureStrategy.BUFFER);
     * 
*

* Whenever a {@link Subscriber} subscribes to the returned {@code Flowable}, the provided * {@link FlowableOnSubscribe} callback is invoked with a fresh instance of a {@link FlowableEmitter} * that will interact only with that specific {@code Subscriber}. If this {@code Subscriber} * cancels the flow (making {@link FlowableEmitter#isCancelled} return {@code true}), * other observers subscribed to the same returned {@code Flowable} are not affected. *

* You should call the {@link FlowableEmitter#onNext(Object)}, {@link FlowableEmitter#onError(Throwable)} * and {@link FlowableEmitter#onComplete()} methods in a serialized fashion. The * rest of its methods are thread-safe. *

*
Backpressure:
*
The backpressure behavior is determined by the {@code mode} parameter.
*
Scheduler:
*
{@code create} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type * @param source the emitter that is called when a {@code Subscriber} subscribes to the returned {@code Flowable} * @param mode the backpressure mode to apply if the downstream {@code Subscriber} doesn't request (fast) enough * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source} or {@code mode} is {@code null} * @see FlowableOnSubscribe * @see BackpressureStrategy * @see Cancellable */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable create(@NonNull FlowableOnSubscribe source, @NonNull BackpressureStrategy mode) { Objects.requireNonNull(source, "source is null"); Objects.requireNonNull(mode, "mode is null"); return RxJavaPlugins.onAssembly(new FlowableCreate<>(source, mode)); } /** * Returns a {@code Flowable} that calls a {@link Publisher} factory to create a {@code Publisher} for each new {@link Subscriber} * that subscribes. That is, for each subscriber, the actual {@code Publisher} that subscriber observes is * determined by the factory function. *

* *

* The defer {@code Subscriber} allows you to defer or delay emitting items from a {@code Publisher} until such time as a * {@code Subscriber} subscribes to the {@code Publisher}. This allows a {@code Subscriber} to easily obtain updates or a * refreshed version of the sequence. *

*
Backpressure:
*
The operator itself doesn't interfere with backpressure which is determined by the {@code Publisher} * returned by the {@code supplier}.
*
Scheduler:
*
{@code defer} does not operate by default on a particular {@link Scheduler}.
*
* * @param supplier * the {@code Publisher} factory function to invoke for each {@code Subscriber} that subscribes to the * resulting {@code Flowable} * @param * the type of the items emitted by the {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code supplier} is {@code null} * @see ReactiveX operators documentation: Defer */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable defer(@NonNull Supplier> supplier) { Objects.requireNonNull(supplier, "supplier is null"); return RxJavaPlugins.onAssembly(new FlowableDefer<>(supplier)); } /** * Returns a {@code Flowable} that emits no items to the {@link Subscriber} and immediately invokes its * {@link Subscriber#onComplete onComplete} method. *

* *

*
Backpressure:
*
This source doesn't produce any elements and effectively ignores downstream backpressure.
*
Scheduler:
*
{@code empty} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of the items (ostensibly) emitted by the {@link Publisher} * @return the shared {@code Flowable} instance * @see ReactiveX operators documentation: Empty */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @SuppressWarnings("unchecked") @NonNull public static <@NonNull T> Flowable empty() { return RxJavaPlugins.onAssembly((Flowable) FlowableEmpty.INSTANCE); } /** * Returns a {@code Flowable} that invokes a {@link Subscriber}'s {@link Subscriber#onError onError} method when the * {@code Subscriber} subscribes to it. *

* *

*
Backpressure:
*
This source doesn't produce any elements and effectively ignores downstream backpressure.
*
Scheduler:
*
{@code error} does not operate by default on a particular {@link Scheduler}.
*
* * @param supplier * a {@link Supplier} factory to return a {@link Throwable} for each individual {@code Subscriber} * @param * the type of the items (ostensibly) emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code supplier} is {@code null} * @see ReactiveX operators documentation: Throw */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable error(@NonNull Supplier supplier) { Objects.requireNonNull(supplier, "supplier is null"); return RxJavaPlugins.onAssembly(new FlowableError<>(supplier)); } /** * Returns a {@code Flowable} that invokes a {@link Subscriber}'s {@link Subscriber#onError onError} method when the * {@code Subscriber} subscribes to it. *

* *

*
Backpressure:
*
This source doesn't produce any elements and effectively ignores downstream backpressure.
*
Scheduler:
*
{@code error} does not operate by default on a particular {@link Scheduler}.
*
* * @param throwable * the particular {@link Throwable} to pass to {@link Subscriber#onError onError} * @param * the type of the items (ostensibly) emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code throwable} is {@code null} * @see ReactiveX operators documentation: Throw */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable error(@NonNull Throwable throwable) { Objects.requireNonNull(throwable, "throwable is null"); return error(Functions.justSupplier(throwable)); } /** * Returns a {@code Flowable} instance that runs the given {@link Action} for each {@link Subscriber} and * emits either its exception or simply completes. *

* *

*
Backpressure:
*
This source doesn't produce any elements and effectively ignores downstream backpressure.
*
Scheduler:
*
{@code fromAction} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the {@code Action} throws an exception, the respective {@link Throwable} is * delivered to the downstream via {@link Subscriber#onError(Throwable)}, * except when the downstream has canceled the resulting {@code Flowable} source. * In this latter case, the {@code Throwable} is delivered to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} as an {@link io.reactivex.rxjava3.exceptions.UndeliverableException UndeliverableException}. *
*
* @param the target type * @param action the {@code Action} to run for each {@code Subscriber} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code action} is {@code null} * @since 3.0.0 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.PASS_THROUGH) public static <@NonNull T> Flowable fromAction(@NonNull Action action) { Objects.requireNonNull(action, "action is null"); return RxJavaPlugins.onAssembly(new FlowableFromAction<>(action)); } /** * Converts an array into a {@link Publisher} that emits the items in the array. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and iterates the given {@code array} * on demand (i.e., when requested).
*
Scheduler:
*
{@code fromArray} does not operate by default on a particular {@link Scheduler}.
*
* * @param items * the array of elements * @param * the type of items in the array and the type of items to be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code items} is {@code null} * @see ReactiveX operators documentation: From */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs public static <@NonNull T> Flowable fromArray(@NonNull T... items) { Objects.requireNonNull(items, "items is null"); if (items.length == 0) { return empty(); } if (items.length == 1) { return just(items[0]); } return RxJavaPlugins.onAssembly(new FlowableFromArray<>(items)); } /** * Returns a {@code Flowable} that, when a {@link Subscriber} subscribes to it, invokes a function you specify and then * emits the value returned from that function. *

* *

* This allows you to defer the execution of the function you specify until a {@code Subscriber} subscribes to the * {@link Publisher}. That is to say, it makes the function "lazy." *

*
Backpressure:
*
The operator honors backpressure from downstream.
*
Scheduler:
*
{@code fromCallable} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the {@link Callable} throws an exception, the respective {@link Throwable} is * delivered to the downstream via {@link Subscriber#onError(Throwable)}, * except when the downstream has canceled this {@code Flowable} source. * In this latter case, the {@code Throwable} is delivered to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} as an {@link io.reactivex.rxjava3.exceptions.UndeliverableException UndeliverableException}. *
*
* * @param callable * a function, the execution of which should be deferred; {@code fromCallable} will invoke this * function only when a {@code Subscriber} subscribes to the {@code Publisher} that {@code fromCallable} returns * @param * the type of the item emitted by the {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code callable} is {@code null} * @see #defer(Supplier) * @see #fromSupplier(Supplier) * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable fromCallable(@NonNull Callable callable) { Objects.requireNonNull(callable, "callable is null"); return RxJavaPlugins.onAssembly(new FlowableFromCallable<>(callable)); } /** * Wraps a {@link CompletableSource} into a {@code Flowable}. *

* *

*
Backpressure:
*
This source doesn't produce any elements and effectively ignores downstream backpressure.
*
Scheduler:
*
{@code fromCompletable} does not operate by default on a particular {@link Scheduler}.
*
* @param the target type * @param completableSource the {@code CompletableSource} to convert from * @return the new {@code Flowable} instance * @throws NullPointerException if {@code completableSource} is {@code null} */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.PASS_THROUGH) public static <@NonNull T> Flowable fromCompletable(@NonNull CompletableSource completableSource) { Objects.requireNonNull(completableSource, "completableSource is null"); return RxJavaPlugins.onAssembly(new FlowableFromCompletable<>(completableSource)); } /** * Converts a {@link Future} into a {@link Publisher}. *

* *

* The operator calls {@link Future#get()}, which is a blocking method, on the subscription thread. * It is recommended applying {@link #subscribeOn(Scheduler)} to move this blocking wait to a * background thread, and if the {@link Scheduler} supports it, interrupt the wait when the flow * is disposed. *

* Also note that this operator will consume a {@link CompletionStage}-based {@code Future} subclass (such as * {@link CompletableFuture}) in a blocking manner as well. Use the {@link #fromCompletionStage(CompletionStage)} * operator to convert and consume such sources in a non-blocking fashion instead. *

* Unlike 1.x, canceling the {@code Flowable} won't cancel the future. If necessary, one can use composition to achieve the * cancellation effect: {@code futurePublisher.doOnCancel(() -> future.cancel(true));}. *

*
Backpressure:
*
The operator honors backpressure from downstream.
*
Scheduler:
*
{@code fromFuture} does not operate by default on a particular {@code Scheduler}.
*
* * @param future * the source {@code Future} * @param * the type of object that the {@code Future} returns, and also the type of item to be emitted by * the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code future} is {@code null} * @see ReactiveX operators documentation: From * @see #fromCompletionStage(CompletionStage) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable fromFuture(@NonNull Future future) { Objects.requireNonNull(future, "future is null"); return RxJavaPlugins.onAssembly(new FlowableFromFuture<>(future, 0L, null)); } /** * Converts a {@link Future} into a {@link Publisher}, with a timeout on the {@code Future}. *

* *

* The operator calls {@link Future#get(long, TimeUnit)}, which is a blocking method, on the subscription thread. * It is recommended applying {@link #subscribeOn(Scheduler)} to move this blocking wait to a * background thread, and if the {@link Scheduler} supports it, interrupt the wait when the flow * is disposed. *

* Unlike 1.x, canceling the {@code Flowable} won't cancel the future. If necessary, one can use composition to achieve the * cancellation effect: {@code futurePublisher.doOnCancel(() -> future.cancel(true));}. *

* Also note that this operator will consume a {@link CompletionStage}-based {@code Future} subclass (such as * {@link CompletableFuture}) in a blocking manner as well. Use the {@link #fromCompletionStage(CompletionStage)} * operator to convert and consume such sources in a non-blocking fashion instead. *

*
Backpressure:
*
The operator honors backpressure from downstream.
*
Scheduler:
*
{@code fromFuture} does not operate by default on a particular {@code Scheduler}.
*
* * @param future * the source {@code Future} * @param timeout * the maximum time to wait before calling {@code get} * @param unit * the {@link TimeUnit} of the {@code timeout} argument * @param * the type of object that the {@code Future} returns, and also the type of item to be emitted by * the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code future} or {@code unit} is {@code null} * @see ReactiveX operators documentation: From * @see #fromCompletionStage(CompletionStage) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable fromFuture(@NonNull Future future, long timeout, @NonNull TimeUnit unit) { Objects.requireNonNull(future, "future is null"); Objects.requireNonNull(unit, "unit is null"); return RxJavaPlugins.onAssembly(new FlowableFromFuture<>(future, timeout, unit)); } /** * Converts an {@link Iterable} sequence into a {@link Publisher} that emits the items in the sequence. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and iterates the given {@code iterable} * on demand (i.e., when requested).
*
Scheduler:
*
{@code fromIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param source * the source {@code Iterable} sequence * @param * the type of items in the {@code Iterable} sequence and the type of items to be emitted by the * resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source} is {@code null} * @see ReactiveX operators documentation: From * @see #fromStream(Stream) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable fromIterable(@NonNull Iterable source) { Objects.requireNonNull(source, "source is null"); return RxJavaPlugins.onAssembly(new FlowableFromIterable<>(source)); } /** * Returns a {@code Flowable} instance that when subscribed to, subscribes to the {@link MaybeSource} instance and * emits {@code onSuccess} as a single item or forwards any {@code onComplete} or * {@code onError} signal. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream.
*
Scheduler:
*
{@code fromMaybe} does not operate by default on a particular {@link Scheduler}.
*
* @param the value type of the {@code MaybeSource} element * @param maybe the {@code MaybeSource} instance to subscribe to, not {@code null} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code maybe} is {@code null} * @since 3.0.0 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) public static <@NonNull T> Flowable fromMaybe(@NonNull MaybeSource maybe) { Objects.requireNonNull(maybe, "maybe is null"); return RxJavaPlugins.onAssembly(new MaybeToFlowable<>(maybe)); } /** * Converts the given {@link ObservableSource} into a {@code Flowable} by applying the specified backpressure strategy. *

* Marble diagrams for the various backpressure strategies are as follows: *

    *
  • {@link BackpressureStrategy#BUFFER} *

    * *

  • *
  • {@link BackpressureStrategy#DROP} *

    * *

  • *
  • {@link BackpressureStrategy#LATEST} *

    * *

  • *
  • {@link BackpressureStrategy#ERROR} *

    * *

  • *
  • {@link BackpressureStrategy#MISSING} *

    * *

  • *
*
*
Backpressure:
*
The operator applies the chosen backpressure strategy of {@link BackpressureStrategy} enum.
*
Scheduler:
*
{@code fromObservable} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the source and resulting sequence * @param source the {@code ObservableSource} to convert * @param strategy the backpressure strategy to apply * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source} or {@code strategy} is {@code null} */ @BackpressureSupport(BackpressureKind.SPECIAL) @CheckReturnValue @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable fromObservable(@NonNull ObservableSource source, @NonNull BackpressureStrategy strategy) { Objects.requireNonNull(source, "source is null"); Objects.requireNonNull(strategy, "strategy is null"); Flowable f = new FlowableFromObservable<>(source); switch (strategy) { case DROP: return f.onBackpressureDrop(); case LATEST: return f.onBackpressureLatest(); case MISSING: return f; case ERROR: return RxJavaPlugins.onAssembly(new FlowableOnBackpressureError<>(f)); default: return f.onBackpressureBuffer(); } } /** * Converts an arbitrary Reactive Streams {@link Publisher} into a {@code Flowable} if not already a * {@code Flowable}. *

* The {@code Publisher} must follow the * Reactive-Streams specification. * Violating the specification may result in undefined behavior. *

* If possible, use {@link #create(FlowableOnSubscribe, BackpressureStrategy)} to create a * source-like {@code Flowable} instead. *

* Note that even though {@code Publisher} appears to be a functional interface, it * is not recommended to implement it through a lambda as the specification requires * state management that is not achievable with a stateless lambda. *

*
Backpressure:
*
The operator is a pass-through for backpressure and its behavior is determined by the * backpressure behavior of the wrapped publisher.
*
Scheduler:
*
{@code fromPublisher} does not operate by default on a particular {@link Scheduler}.
*
* @param the value type of the flow * @param publisher the {@code Publisher} to convert * @return the new {@code Flowable} instance * @throws NullPointerException if {@code publisher} is {@code null} * @see #create(FlowableOnSubscribe, BackpressureStrategy) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @SuppressWarnings("unchecked") public static <@NonNull T> Flowable fromPublisher(@NonNull Publisher publisher) { if (publisher instanceof Flowable) { return RxJavaPlugins.onAssembly((Flowable)publisher); } Objects.requireNonNull(publisher, "publisher is null"); return RxJavaPlugins.onAssembly(new FlowableFromPublisher<>(publisher)); } /** * Returns a {@code Flowable} instance that runs the given {@link Runnable} for each {@link Subscriber} and * emits either its unchecked exception or simply completes. *

* *

* If the code to be wrapped needs to throw a checked or more broader {@link Throwable} exception, that * exception has to be converted to an unchecked exception by the wrapped code itself. Alternatively, * use the {@link #fromAction(Action)} method which allows the wrapped code to throw any {@code Throwable} * exception and will signal it to observers as-is. *

*
Backpressure:
*
This source doesn't produce any elements and effectively ignores downstream backpressure.
*
Scheduler:
*
{@code fromRunnable} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the {@code Runnable} throws an exception, the respective {@code Throwable} is * delivered to the downstream via {@link Subscriber#onError(Throwable)}, * except when the downstream has canceled the resulting {@code Flowable} source. * In this latter case, the {@code Throwable} is delivered to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} as an {@link io.reactivex.rxjava3.exceptions.UndeliverableException UndeliverableException}. *
*
* @param the target type * @param run the {@code Runnable} to run for each {@code Subscriber} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code run} is {@code null} * @since 3.0.0 * @see #fromAction(Action) */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.PASS_THROUGH) public static <@NonNull T> Flowable fromRunnable(@NonNull Runnable run) { Objects.requireNonNull(run, "run is null"); return RxJavaPlugins.onAssembly(new FlowableFromRunnable<>(run)); } /** * Returns a {@code Flowable} instance that when subscribed to, subscribes to the {@link SingleSource} instance and * emits {@code onSuccess} as a single item or forwards the {@code onError} signal. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream.
*
Scheduler:
*
{@code fromSingle} does not operate by default on a particular {@link Scheduler}.
*
* @param the value type of the {@code SingleSource} element * @param source the {@code SingleSource} instance to subscribe to, not {@code null} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source} is {@code null} * @since 3.0.0 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) public static <@NonNull T> Flowable fromSingle(@NonNull SingleSource source) { Objects.requireNonNull(source, "source is null"); return RxJavaPlugins.onAssembly(new SingleToFlowable<>(source)); } /** * Returns a {@code Flowable} that, when a {@link Subscriber} subscribes to it, invokes a supplier function you specify and then * emits the value returned from that function. *

* *

* This allows you to defer the execution of the function you specify until a {@code Subscriber} subscribes to the * {@link Publisher}. That is to say, it makes the function "lazy." *

*
Backpressure:
*
The operator honors backpressure from downstream.
*
Scheduler:
*
{@code fromSupplier} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the {@link Supplier} throws an exception, the respective {@link Throwable} is * delivered to the downstream via {@link Subscriber#onError(Throwable)}, * except when the downstream has canceled this {@code Flowable} source. * In this latter case, the {@code Throwable} is delivered to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} as an {@link io.reactivex.rxjava3.exceptions.UndeliverableException UndeliverableException}. *
*
* * @param supplier * a function, the execution of which should be deferred; {@code fromSupplier} will invoke this * function only when a {@code Subscriber} subscribes to the {@code Publisher} that {@code fromSupplier} returns * @param * the type of the item emitted by the {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code supplier} is {@code null} * @see #defer(Supplier) * @see #fromCallable(Callable) * @since 3.0.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable fromSupplier(@NonNull Supplier supplier) { Objects.requireNonNull(supplier, "supplier is null"); return RxJavaPlugins.onAssembly(new FlowableFromSupplier<>(supplier)); } /** * Returns a cold, synchronous, stateless and backpressure-aware generator of values. *

* Note that the {@link Emitter#onNext}, {@link Emitter#onError} and * {@link Emitter#onComplete} methods provided to the function via the {@link Emitter} instance should be called synchronously, * never concurrently and only while the function body is executing. Calling them from multiple threads * or outside the function call is not supported and leads to an undefined behavior. *

*
Backpressure:
*
The operator honors downstream backpressure.
*
Scheduler:
*
{@code generate} does not operate by default on a particular {@link Scheduler}.
*
* * @param the generated value type * @param generator the {@link Consumer} called whenever a particular downstream {@link Subscriber} has * requested a value. The callback then should call {@code onNext}, {@code onError} or * {@code onComplete} to signal a value or a terminal event. Signaling multiple {@code onNext} * in a call will make the operator signal {@link IllegalStateException}. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code generator} is {@code null} */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable generate(@NonNull Consumer<@NonNull Emitter> generator) { Objects.requireNonNull(generator, "generator is null"); return generate(Functions.nullSupplier(), FlowableInternalHelper.simpleGenerator(generator), Functions.emptyConsumer()); } /** * Returns a cold, synchronous, stateful and backpressure-aware generator of values. *

* Note that the {@link Emitter#onNext}, {@link Emitter#onError} and * {@link Emitter#onComplete} methods provided to the function via the {@link Emitter} instance should be called synchronously, * never concurrently and only while the function body is executing. Calling them from multiple threads * or outside the function call is not supported and leads to an undefined behavior. *

*
Backpressure:
*
The operator honors downstream backpressure.
*
Scheduler:
*
{@code generate} does not operate by default on a particular {@link Scheduler}.
*
* * @param the type of the per-{@link Subscriber} state * @param the generated value type * @param initialState the {@link Supplier} to generate the initial state for each {@code Subscriber} * @param generator the {@link Consumer} called with the current state whenever a particular downstream {@code Subscriber} has * requested a value. The callback then should call {@code onNext}, {@code onError} or * {@code onComplete} to signal a value or a terminal event. Signaling multiple {@code onNext} * in a call will make the operator signal {@link IllegalStateException}. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code initialState} or {@code generator} is {@code null} */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T, @NonNull S> Flowable generate(@NonNull Supplier initialState, @NonNull BiConsumer> generator) { Objects.requireNonNull(generator, "generator is null"); return generate(initialState, FlowableInternalHelper.simpleBiGenerator(generator), Functions.emptyConsumer()); } /** * Returns a cold, synchronous, stateful and backpressure-aware generator of values. *

* Note that the {@link Emitter#onNext}, {@link Emitter#onError} and * {@link Emitter#onComplete} methods provided to the function via the {@link Emitter} instance should be called synchronously, * never concurrently and only while the function body is executing. Calling them from multiple threads * or outside the function call is not supported and leads to an undefined behavior. *

*
Backpressure:
*
The operator honors downstream backpressure.
*
Scheduler:
*
{@code generate} does not operate by default on a particular {@link Scheduler}.
*
* * @param the type of the per-{@link Subscriber} state * @param the generated value type * @param initialState the {@link Supplier} to generate the initial state for each {@code Subscriber} * @param generator the {@link Consumer} called with the current state whenever a particular downstream {@code Subscriber} has * requested a value. The callback then should call {@code onNext}, {@code onError} or * {@code onComplete} to signal a value or a terminal event. Signaling multiple {@code onNext} * in a call will make the operator signal {@link IllegalStateException}. * @param disposeState the {@code Consumer} that is called with the current state when the generator * terminates the sequence or it gets canceled * @return the new {@code Flowable} instance * @throws NullPointerException if {@code initialState}, {@code generator} or {@code disposeState} is {@code null} */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T, @NonNull S> Flowable generate(@NonNull Supplier initialState, @NonNull BiConsumer> generator, @NonNull Consumer disposeState) { Objects.requireNonNull(generator, "generator is null"); return generate(initialState, FlowableInternalHelper.simpleBiGenerator(generator), disposeState); } /** * Returns a cold, synchronous, stateful and backpressure-aware generator of values. *

* Note that the {@link Emitter#onNext}, {@link Emitter#onError} and * {@link Emitter#onComplete} methods provided to the function via the {@link Emitter} instance should be called synchronously, * never concurrently and only while the function body is executing. Calling them from multiple threads * or outside the function call is not supported and leads to an undefined behavior. *

*
Backpressure:
*
The operator honors downstream backpressure.
*
Scheduler:
*
{@code generate} does not operate by default on a particular {@link Scheduler}.
*
* * @param the type of the per-{@link Subscriber} state * @param the generated value type * @param initialState the {@link Supplier} to generate the initial state for each {@code Subscriber} * @param generator the {@link Function} called with the current state whenever a particular downstream {@code Subscriber} has * requested a value. The callback then should call {@code onNext}, {@code onError} or * {@code onComplete} to signal a value or a terminal event and should return a (new) state for * the next invocation. Signaling multiple {@code onNext} * in a call will make the operator signal {@link IllegalStateException}. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code initialState} or {@code generator} is {@code null} */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T, @NonNull S> Flowable generate(@NonNull Supplier initialState, @NonNull BiFunction, S> generator) { return generate(initialState, generator, Functions.emptyConsumer()); } /** * Returns a cold, synchronous, stateful and backpressure-aware generator of values. *

* Note that the {@link Emitter#onNext}, {@link Emitter#onError} and * {@link Emitter#onComplete} methods provided to the function via the {@link Emitter} instance should be called synchronously, * never concurrently and only while the function body is executing. Calling them from multiple threads * or outside the function call is not supported and leads to an undefined behavior. *

*
Backpressure:
*
The operator honors downstream backpressure.
*
Scheduler:
*
{@code generate} does not operate by default on a particular {@link Scheduler}.
*
* * @param the type of the per-{@link Subscriber} state * @param the generated value type * @param initialState the {@link Supplier} to generate the initial state for each {@code Subscriber} * @param generator the {@link Function} called with the current state whenever a particular downstream {@code Subscriber} has * requested a value. The callback then should call {@code onNext}, {@code onError} or * {@code onComplete} to signal a value or a terminal event and should return a (new) state for * the next invocation. Signaling multiple {@code onNext} * in a call will make the operator signal {@link IllegalStateException}. * @param disposeState the {@link Consumer} that is called with the current state when the generator * terminates the sequence or it gets canceled * @return the new {@code Flowable} instance * @throws NullPointerException if {@code initialState}, {@code generator} or {@code disposeState} is {@code null} */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T, @NonNull S> Flowable generate(@NonNull Supplier initialState, @NonNull BiFunction, S> generator, @NonNull Consumer disposeState) { Objects.requireNonNull(initialState, "initialState is null"); Objects.requireNonNull(generator, "generator is null"); Objects.requireNonNull(disposeState, "disposeState is null"); return RxJavaPlugins.onAssembly(new FlowableGenerate<>(initialState, generator, disposeState)); } /** * Returns a {@code Flowable} that emits a {@code 0L} after the {@code initialDelay} and ever-increasing numbers * after each {@code period} of time thereafter. *

* *

*
Backpressure:
*
The operator generates values based on time and ignores downstream backpressure which * may lead to {@link MissingBackpressureException} at some point in the chain. * Downstream consumers should consider applying one of the {@code onBackpressureXXX} operators as well.
*
Scheduler:
*
{@code interval} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param initialDelay * the initial delay time to wait before emitting the first value of 0L * @param period * the period of time between emissions of the subsequent numbers * @param unit * the time unit for both {@code initialDelay} and {@code period} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Interval * @since 1.0.12 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public static Flowable interval(long initialDelay, long period, @NonNull TimeUnit unit) { return interval(initialDelay, period, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits a {@code 0L} after the {@code initialDelay} and ever-increasing numbers * after each {@code period} of time thereafter, on a specified {@link Scheduler}. *

* *

*
Backpressure:
*
The operator generates values based on time and ignores downstream backpressure which * may lead to {@link MissingBackpressureException} at some point in the chain. * Downstream consumers should consider applying one of the {@code onBackpressureXXX} operators as well.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param initialDelay * the initial delay time to wait before emitting the first value of 0L * @param period * the period of time between emissions of the subsequent numbers * @param unit * the time unit for both {@code initialDelay} and {@code period} * @param scheduler * the {@code Scheduler} on which the waiting happens and items are emitted * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Interval * @since 1.0.12 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public static Flowable interval(long initialDelay, long period, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableInterval(Math.max(0L, initialDelay), Math.max(0L, period), unit, scheduler)); } /** * Returns a {@code Flowable} that emits a sequential number every specified interval of time. *

* *

*
Backpressure:
*
The operator signals a {@link MissingBackpressureException} if the downstream * is not ready to receive the next value.
*
Scheduler:
*
{@code interval} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param period * the period size in time units (see below) * @param unit * time units to use for the interval size * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Interval */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public static Flowable interval(long period, @NonNull TimeUnit unit) { return interval(period, period, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits a sequential number every specified interval of time, on a * specified {@link Scheduler}. *

* *

*
Backpressure:
*
The operator generates values based on time and ignores downstream backpressure which * may lead to {@link MissingBackpressureException} at some point in the chain. * Downstream consumers should consider applying one of the {@code onBackpressureXXX} operators as well.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param period * the period size in time units (see below) * @param unit * time units to use for the interval size * @param scheduler * the {@code Scheduler} to use for scheduling the items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Interval */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public static Flowable interval(long period, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return interval(period, period, unit, scheduler); } /** * Signals a range of long values, the first after some initial delay and the rest periodically after. *

* The sequence completes immediately after the last value {@code (start + count - 1)} has been reached. *

*
Backpressure:
*
The operator signals a {@link MissingBackpressureException} if the downstream can't keep up.
*
Scheduler:
*
{@code intervalRange} by default operates on the {@link Schedulers#computation() computation} {@link Scheduler}.
*
* @param start that start value of the range * @param count the number of values to emit in total, if zero, the operator emits an {@code onComplete} after the initial delay. * @param initialDelay the initial delay before signaling the first value (the start) * @param period the period between subsequent values * @param unit the unit of measure of the {@code initialDelay} and {@code period} amounts * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @throws IllegalArgumentException * if {@code count} is less than zero, or if {@code start} + {@code count} − 1 exceeds * {@link Long#MAX_VALUE} * @see #range(int, int) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) public static Flowable intervalRange(long start, long count, long initialDelay, long period, @NonNull TimeUnit unit) { return intervalRange(start, count, initialDelay, period, unit, Schedulers.computation()); } /** * Signals a range of long values, the first after some initial delay and the rest periodically after. *

* The sequence completes immediately after the last value (start + count - 1) has been reached. *

*
Backpressure:
*
The operator signals a {@link MissingBackpressureException} if the downstream can't keep up.
*
Scheduler:
*
you provide the {@link Scheduler}.
*
* @param start that start value of the range * @param count the number of values to emit in total, if zero, the operator emits an {@code onComplete} after the initial delay. * @param initialDelay the initial delay before signaling the first value (the start) * @param period the period between subsequent values * @param unit the unit of measure of the {@code initialDelay} and {@code period} amounts * @param scheduler the target {@code Scheduler} where the values and terminal signals will be emitted * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException * if {@code count} is less than zero, or if {@code start} + {@code count} − 1 exceeds * {@link Long#MAX_VALUE} */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public static Flowable intervalRange(long start, long count, long initialDelay, long period, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { if (count < 0L) { throw new IllegalArgumentException("count >= 0 required but it was " + count); } if (count == 0L) { return Flowable.empty().delay(initialDelay, unit, scheduler); } long end = start + (count - 1); if (start > 0 && end < 0) { throw new IllegalArgumentException("Overflow! start + count is bigger than Long.MAX_VALUE"); } Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableIntervalRange(start, end, Math.max(0L, initialDelay), Math.max(0L, period), unit, scheduler)); } /** * Returns a {@code Flowable} that signals the given (constant reference) item and then completes. *

* *

* Note that the item is taken and re-emitted as is and not computed by any means by {@code just}. Use {@link #fromCallable(Callable)} * to generate a single item on demand (when {@link Subscriber}s subscribe to it). *

* See the multi-parameter overloads of {@code just} to emit more than one (constant reference) items one after the other. * Use {@link #fromArray(Object...)} to emit an arbitrary number of items that are known upfront. *

* To emit the items of an {@link Iterable} sequence (such as a {@link java.util.List}), use {@link #fromIterable(Iterable)}. *

*
Backpressure:
*
The operator honors backpressure from downstream.
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item * the item to emit * @param * the type of that item * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item} is {@code null} * @see ReactiveX operators documentation: Just * @see #just(Object, Object) * @see #fromCallable(Callable) * @see #fromArray(Object...) * @see #fromIterable(Iterable) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item) { Objects.requireNonNull(item, "item is null"); return RxJavaPlugins.onAssembly(new FlowableJust<>(item)); } /** * Converts two items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1} or {@code item2} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); return fromArray(item1, item2); } /** * Converts three items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param item3 * third item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1}, {@code item2} or {@code item3} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2, T item3) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); Objects.requireNonNull(item3, "item3 is null"); return fromArray(item1, item2, item3); } /** * Converts four items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param item3 * third item * @param item4 * fourth item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1}, {@code item2}, {@code item3}, * or {@code item4} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2, T item3, T item4) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); Objects.requireNonNull(item3, "item3 is null"); Objects.requireNonNull(item4, "item4 is null"); return fromArray(item1, item2, item3, item4); } /** * Converts five items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param item3 * third item * @param item4 * fourth item * @param item5 * fifth item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1}, {@code item2}, {@code item3}, * {@code item4} or {@code item5} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2, T item3, T item4, T item5) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); Objects.requireNonNull(item3, "item3 is null"); Objects.requireNonNull(item4, "item4 is null"); Objects.requireNonNull(item5, "item5 is null"); return fromArray(item1, item2, item3, item4, item5); } /** * Converts six items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param item3 * third item * @param item4 * fourth item * @param item5 * fifth item * @param item6 * sixth item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1}, {@code item2}, {@code item3}, * {@code item4}, {@code item5} or {@code item6} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2, T item3, T item4, T item5, T item6) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); Objects.requireNonNull(item3, "item3 is null"); Objects.requireNonNull(item4, "item4 is null"); Objects.requireNonNull(item5, "item5 is null"); Objects.requireNonNull(item6, "item6 is null"); return fromArray(item1, item2, item3, item4, item5, item6); } /** * Converts seven items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param item3 * third item * @param item4 * fourth item * @param item5 * fifth item * @param item6 * sixth item * @param item7 * seventh item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1}, {@code item2}, {@code item3}, * {@code item4}, {@code item5}, {@code item6} * or {@code item7} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2, T item3, T item4, T item5, T item6, T item7) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); Objects.requireNonNull(item3, "item3 is null"); Objects.requireNonNull(item4, "item4 is null"); Objects.requireNonNull(item5, "item5 is null"); Objects.requireNonNull(item6, "item6 is null"); Objects.requireNonNull(item7, "item7 is null"); return fromArray(item1, item2, item3, item4, item5, item6, item7); } /** * Converts eight items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param item3 * third item * @param item4 * fourth item * @param item5 * fifth item * @param item6 * sixth item * @param item7 * seventh item * @param item8 * eighth item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1}, {@code item2}, {@code item3}, * {@code item4}, {@code item5}, {@code item6}, * {@code item7} or {@code item8} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2, T item3, T item4, T item5, T item6, T item7, T item8) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); Objects.requireNonNull(item3, "item3 is null"); Objects.requireNonNull(item4, "item4 is null"); Objects.requireNonNull(item5, "item5 is null"); Objects.requireNonNull(item6, "item6 is null"); Objects.requireNonNull(item7, "item7 is null"); Objects.requireNonNull(item8, "item8 is null"); return fromArray(item1, item2, item3, item4, item5, item6, item7, item8); } /** * Converts nine items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param item3 * third item * @param item4 * fourth item * @param item5 * fifth item * @param item6 * sixth item * @param item7 * seventh item * @param item8 * eighth item * @param item9 * ninth item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1}, {@code item2}, {@code item3}, * {@code item4}, {@code item5}, {@code item6}, * {@code item7}, {@code item8} or {@code item9} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2, T item3, T item4, T item5, T item6, T item7, T item8, T item9) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); Objects.requireNonNull(item3, "item3 is null"); Objects.requireNonNull(item4, "item4 is null"); Objects.requireNonNull(item5, "item5 is null"); Objects.requireNonNull(item6, "item6 is null"); Objects.requireNonNull(item7, "item7 is null"); Objects.requireNonNull(item8, "item8 is null"); Objects.requireNonNull(item9, "item9 is null"); return fromArray(item1, item2, item3, item4, item5, item6, item7, item8, item9); } /** * Converts ten items into a {@link Publisher} that emits those items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals each value on-demand (i.e., when requested).
*
Scheduler:
*
{@code just} does not operate by default on a particular {@link Scheduler}.
*
* * @param item1 * first item * @param item2 * second item * @param item3 * third item * @param item4 * fourth item * @param item5 * fifth item * @param item6 * sixth item * @param item7 * seventh item * @param item8 * eighth item * @param item9 * ninth item * @param item10 * tenth item * @param * the type of these items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item1}, {@code item2}, {@code item3}, * {@code item4}, {@code item5}, {@code item6}, * {@code item7}, {@code item8}, {@code item9}, * or {@code item10} is {@code null} * @see ReactiveX operators documentation: Just */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable just(T item1, T item2, T item3, T item4, T item5, T item6, T item7, T item8, T item9, T item10) { Objects.requireNonNull(item1, "item1 is null"); Objects.requireNonNull(item2, "item2 is null"); Objects.requireNonNull(item3, "item3 is null"); Objects.requireNonNull(item4, "item4 is null"); Objects.requireNonNull(item5, "item5 is null"); Objects.requireNonNull(item6, "item6 is null"); Objects.requireNonNull(item7, "item7 is null"); Objects.requireNonNull(item8, "item8 is null"); Objects.requireNonNull(item9, "item9 is null"); Objects.requireNonNull(item10, "item10 is null"); return fromArray(item1, item2, item3, item4, item5, item6, item7, item8, item9, item10); } /** * Flattens an {@link Iterable} of {@link Publisher}s into one {@code Publisher}, without any transformation, while limiting the * number of concurrent subscriptions to these {@code Publisher}s. *

* *

* You can combine the items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code merge} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeDelayError(Iterable, int, int)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param sources * the {@code Iterable} of {@code Publisher}s * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @param bufferSize * the number of items to prefetch from each inner {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException * if {@code maxConcurrency} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Merge * @see #mergeDelayError(Iterable, int, int) */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable merge(@NonNull Iterable<@NonNull ? extends Publisher> sources, int maxConcurrency, int bufferSize) { return fromIterable(sources).flatMap((Function)Functions.identity(), false, maxConcurrency, bufferSize); } /** * Flattens an array of {@link Publisher}s into one {@code Publisher}, without any transformation, while limiting the * number of concurrent subscriptions to these {@code Publisher}s. *

* *

* You can combine the items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeArray} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeArrayDelayError(int, int, Publisher[])} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param sources * the array of {@code Publisher}s * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @param bufferSize * the number of items to prefetch from each inner {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException * if {@code maxConcurrency} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Merge * @see #mergeArrayDelayError(int, int, Publisher...) */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs @NonNull public static <@NonNull T> Flowable mergeArray(int maxConcurrency, int bufferSize, @NonNull Publisher... sources) { return fromArray(sources).flatMap((Function)Functions.identity(), false, maxConcurrency, bufferSize); } /** * Flattens an {@link Iterable} of {@link Publisher}s into one {@code Publisher}, without any transformation. *

* *

* You can combine the items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code merge} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeDelayError(Iterable)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param sources * the {@code Iterable} of {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Merge * @see #mergeDelayError(Iterable) */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable merge(@NonNull Iterable<@NonNull ? extends Publisher> sources) { return fromIterable(sources).flatMap((Function)Functions.identity()); } /** * Flattens an {@link Iterable} of {@link Publisher}s into one {@code Publisher}, without any transformation, while limiting the * number of concurrent subscriptions to these {@code Publisher}s. *

* *

* You can combine the items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code merge} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeDelayError(Iterable, int)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param sources * the {@code Iterable} of {@code Publisher}s * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException * if {@code maxConcurrency} is less than or equal to 0 * @see ReactiveX operators documentation: Merge * @see #mergeDelayError(Iterable, int) */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable merge(@NonNull Iterable<@NonNull ? extends Publisher> sources, int maxConcurrency) { return fromIterable(sources).flatMap((Function)Functions.identity(), maxConcurrency); } /** * Flattens a {@link Publisher} that emits {@code Publisher}s into a single {@code Publisher} that emits the items emitted by * thos {@code Publisher}s , without any transformation. *

* *

* You can combine the items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed * in unbounded mode (i.e., no backpressure is applied to it). The inner {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code merge} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeDelayError(Publisher)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param sources * a {@code Publisher} that emits {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Merge * @see #mergeDelayError(Publisher) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable merge(@NonNull Publisher<@NonNull ? extends Publisher> sources) { return merge(sources, bufferSize()); } /** * Flattens a {@link Publisher} that emits {@code Publisher}s into a single {@code Publisher} that emits the items emitted by * those {@code Publisher}s, without any transformation, while limiting the maximum number of concurrent * subscriptions to these {@code Publisher}s. *

* *

* You can combine the items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. Both the outer and inner {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code merge} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeDelayError(Publisher, int)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param sources * a {@code Publisher} that emits {@code Publisher}s * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException * if {@code maxConcurrency} is less than or equal to 0 * @see ReactiveX operators documentation: Merge * @see #mergeDelayError(Publisher, int) * @since 1.1.0 */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable merge(@NonNull Publisher<@NonNull ? extends Publisher> sources, int maxConcurrency) { return fromPublisher(sources).flatMap((Function)Functions.identity(), maxConcurrency); } /** * Flattens an array of {@link Publisher}s into one {@code Publisher}, without any transformation. *

* *

* You can combine items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeArray} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeArrayDelayError(Publisher...)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param sources * the array of {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Merge * @see #mergeArrayDelayError(Publisher...) */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs @NonNull public static <@NonNull T> Flowable mergeArray(@NonNull Publisher... sources) { return fromArray(sources).flatMap((Function)Functions.identity(), sources.length); } /** * Flattens two {@link Publisher}s into a single {@code Publisher}, without any transformation. *

* *

* You can combine items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code merge} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeDelayError(Publisher, Publisher)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be merged * @param source2 * a {@code Publisher} to be merged * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1} or {@code source2} is {@code null} * @see ReactiveX operators documentation: Merge * @see #mergeDelayError(Publisher, Publisher) */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable merge(@NonNull Publisher source1, @NonNull Publisher source2) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); return fromArray(source1, source2).flatMap((Function)Functions.identity(), false, 2); } /** * Flattens three {@link Publisher}s into a single {@code Publisher}, without any transformation. *

* *

* You can combine items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code merge} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeDelayError(Publisher, Publisher, Publisher)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be merged * @param source2 * a {@code Publisher} to be merged * @param source3 * a {@code Publisher} to be merged * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code source3} is {@code null} * @see ReactiveX operators documentation: Merge * @see #mergeDelayError(Publisher, Publisher, Publisher) */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable merge(@NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); return fromArray(source1, source2, source3).flatMap((Function)Functions.identity(), false, 3); } /** * Flattens four {@link Publisher}s into a single {@code Publisher}, without any transformation. *

* *

* You can combine items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code merge} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code merge} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If any of the source {@code Publisher}s signal a {@link Throwable} via {@code onError}, the resulting * {@code Flowable} terminates with that {@code Throwable} and all other source {@code Publisher}s are canceled. * If more than one {@code Publisher} signals an error, the resulting {@code Flowable} may terminate with the * first one's error or, depending on the concurrency of the sources, may terminate with a * {@link CompositeException} containing two or more of the various error signals. * {@code Throwable}s that didn't make into the composite will be sent (individually) to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. Similarly, {@code Throwable}s * signaled by source(s) after the returned {@code Flowable} has been canceled or terminated with a * (composite) error will be sent to the same global error handler. * Use {@link #mergeDelayError(Publisher, Publisher, Publisher, Publisher)} to merge sources and terminate only when all source {@code Publisher}s * have completed or failed with an error. *
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be merged * @param source2 * a {@code Publisher} to be merged * @param source3 * a {@code Publisher} to be merged * @param source4 * a {@code Publisher} to be merged * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3} or {@code source4} is {@code null} * @see ReactiveX operators documentation: Merge * @see #mergeDelayError(Publisher, Publisher, Publisher, Publisher) */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable merge( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); return fromArray(source1, source2, source3, source4).flatMap((Function)Functions.identity(), false, 4); } /** * Flattens an {@link Iterable} of {@link Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to receive all * successfully emitted items from each of the source {@code Publisher}s without being interrupted by an error * notification from one of them. *

* This behaves like {@link #merge(Publisher)} except that if any of the merged {@code Publisher}s notify of an * error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain from propagating that * error notification until all of the merged {@code Publisher}s have finished emitting items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. All inner {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * the {@code Iterable} of {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Merge */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable mergeDelayError(@NonNull Iterable<@NonNull ? extends Publisher> sources) { return fromIterable(sources).flatMap((Function)Functions.identity(), true); } /** * Flattens an {@link Iterable} of {@link Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to receive all * successfully emitted items from each of the source {@code Publisher}s without being interrupted by an error * notification from one of them, while limiting the number of concurrent subscriptions to these {@code Publisher}s. *

* This behaves like {@link #merge(Publisher)} except that if any of the merged {@code Publisher}s notify of an * error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain from propagating that * error notification until all of the merged {@code Publisher}s have finished emitting items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. All inner {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * the {@code Iterable} of {@code Publisher}s * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @param bufferSize * the number of items to prefetch from each inner {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Merge */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable mergeDelayError(@NonNull Iterable<@NonNull ? extends Publisher> sources, int maxConcurrency, int bufferSize) { return fromIterable(sources).flatMap((Function)Functions.identity(), true, maxConcurrency, bufferSize); } /** * Flattens an array of {@link Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to receive all * successfully emitted items from each of the source {@code Publisher}s without being interrupted by an error * notification from one of them, while limiting the number of concurrent subscriptions to these {@code Publisher}s. *

* This behaves like {@link #merge(Publisher)} except that if any of the merged {@code Publisher}s notify of an * error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain from propagating that * error notification until all of the merged {@code Publisher}s have finished emitting items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. All source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeArrayDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * the array of {@code Publisher}s * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @param bufferSize * the number of items to prefetch from each inner {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Merge */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs @NonNull public static <@NonNull T> Flowable mergeArrayDelayError(int maxConcurrency, int bufferSize, @NonNull Publisher... sources) { return fromArray(sources).flatMap((Function)Functions.identity(), true, maxConcurrency, bufferSize); } /** * Flattens an {@link Iterable} of {@link Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to receive all * successfully emitted items from each of the source {@code Publisher}s without being interrupted by an error * notification from one of them, while limiting the number of concurrent subscriptions to these {@code Publisher}s. *

* This behaves like {@link #merge(Publisher)} except that if any of the merged {@code Publisher}s notify of an * error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain from propagating that * error notification until all of the merged {@code Publisher}s have finished emitting items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. All inner {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * the {@code Iterable} of {@code Publisher}s * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive * @see ReactiveX operators documentation: Merge */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable mergeDelayError(@NonNull Iterable<@NonNull ? extends Publisher> sources, int maxConcurrency) { return fromIterable(sources).flatMap((Function)Functions.identity(), true, maxConcurrency); } /** * Flattens a {@link Publisher} that emits {@code Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to * receive all successfully emitted items from all of the source {@code Publisher}s without being interrupted by * an error notification from one of them. *

* This behaves like {@link #merge(Publisher)} except that if any of the merged {@code Publisher}s notify of an * error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain from propagating that * error notification until all of the merged {@code Publisher}s have finished emitting items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed * in unbounded mode (i.e., no backpressure is applied to it). The inner {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * a {@code Publisher} that emits {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Merge */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable mergeDelayError(@NonNull Publisher<@NonNull ? extends Publisher> sources) { return mergeDelayError(sources, bufferSize()); } /** * Flattens a {@link Publisher} that emits {@code Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to * receive all successfully emitted items from all of the source {@code Publisher}s without being interrupted by * an error notification from one of them, while limiting the * number of concurrent subscriptions to these {@code Publisher}s. *

* This behaves like {@link #merge(Publisher)} except that if any of the merged {@code Publisher}s notify of an * error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain from propagating that * error notification until all of the merged {@code Publisher}s have finished emitting items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. Both the outer and inner {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * a {@code Publisher} that emits {@code Publisher}s * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive * @see ReactiveX operators documentation: Merge * @since 2.0 */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable mergeDelayError(@NonNull Publisher<@NonNull ? extends Publisher> sources, int maxConcurrency) { return fromPublisher(sources).flatMap((Function)Functions.identity(), true, maxConcurrency); } /** * Flattens an array of {@link Publisher}s into one {@code Flowable}, in a way that allows a {@link Subscriber} to receive all * successfully emitted items from each of the source {@code Publisher}s without being interrupted by an error * notification from one of them. *

* This behaves like {@link #merge(Publisher)} except that if any of the merged {@code Publisher}s notify of an * error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain from propagating that * error notification until all of the merged {@code Publisher}s have finished emitting items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. Both the outer and inner {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeArrayDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param sources * the array of {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Merge */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs @NonNull public static <@NonNull T> Flowable mergeArrayDelayError(@NonNull Publisher... sources) { return fromArray(sources).flatMap((Function)Functions.identity(), true, sources.length); } /** * Flattens two {@link Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to receive all * successfully emitted items from each of the source {@code Publisher}s without being interrupted by an error * notification from one of them. *

* This behaves like {@link #merge(Publisher, Publisher)} except that if any of the merged {@code Publisher}s * notify of an error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain from * propagating that error notification until all of the merged {@code Publisher}s have finished emitting items. *

* *

* Even if both merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be merged * @param source2 * a {@code Publisher} to be merged * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1} or {@code source2} is {@code null} * @see ReactiveX operators documentation: Merge */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable mergeDelayError(@NonNull Publisher source1, @NonNull Publisher source2) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); return fromArray(source1, source2).flatMap((Function)Functions.identity(), true, 2); } /** * Flattens three {@link Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to receive all * successfully emitted items from all of the source {@code Publisher}s without being interrupted by an error * notification from one of them. *

* This behaves like {@link #merge(Publisher, Publisher, Publisher)} except that if any of the merged * {@code Publisher}s notify of an error via {@link Subscriber#onError onError}, {@code mergeDelayError} will refrain * from propagating that error notification until all of the merged {@code Publisher}s have finished emitting * items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be merged * @param source2 * a {@code Publisher} to be merged * @param source3 * a {@code Publisher} to be merged * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code source3} is {@code null} * @see ReactiveX operators documentation: Merge */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable mergeDelayError(@NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); return fromArray(source1, source2, source3).flatMap((Function)Functions.identity(), true, 3); } /** * Flattens four {@link Publisher}s into one {@code Publisher}, in a way that allows a {@link Subscriber} to receive all * successfully emitted items from all of the source {@code Publisher}s without being interrupted by an error * notification from one of them. *

* This behaves like {@link #merge(Publisher, Publisher, Publisher, Publisher)} except that if any of * the merged {@code Publisher}s notify of an error via {@link Subscriber#onError onError}, {@code mergeDelayError} * will refrain from propagating that error notification until all of the merged {@code Publisher}s have finished * emitting items. *

* *

* Even if multiple merged {@code Publisher}s send {@code onError} notifications, {@code mergeDelayError} will only * invoke the {@code onError} method of its {@code Subscriber}s once. *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element base type * @param source1 * a {@code Publisher} to be merged * @param source2 * a {@code Publisher} to be merged * @param source3 * a {@code Publisher} to be merged * @param source4 * a {@code Publisher} to be merged * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3} or {@code source4} is {@code null} * @see ReactiveX operators documentation: Merge */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable mergeDelayError( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); return fromArray(source1, source2, source3, source4).flatMap((Function)Functions.identity(), true, 4); } /** * Returns a {@code Flowable} that never sends any items or notifications to a {@link Subscriber}. *

* *

* This {@link Publisher} is useful primarily for testing purposes. *

*
Backpressure:
*
This source doesn't produce any elements and effectively ignores downstream backpressure.
*
Scheduler:
*
{@code never} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items (not) emitted by the {@code Publisher} * @return the shared {@code Flowable} instance * @see ReactiveX operators documentation: Never */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @SuppressWarnings("unchecked") @NonNull public static <@NonNull T> Flowable never() { return RxJavaPlugins.onAssembly((Flowable) FlowableNever.INSTANCE); } /** * Returns a {@code Flowable} that emits a sequence of {@link Integer}s within a specified range. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals values on-demand (i.e., when requested).
*
Scheduler:
*
{@code range} does not operate by default on a particular {@link Scheduler}.
*
* * @param start * the value of the first {@code Integer} in the sequence * @param count * the number of sequential {@code Integer}s to generate * @return the new {@code Flowable} instance * @throws IllegalArgumentException * if {@code count} is less than zero, or if {@code start} + {@code count} − 1 exceeds * {@link Integer#MAX_VALUE} * @see ReactiveX operators documentation: Range * @see #rangeLong(long, long) * @see #intervalRange(long, long, long, long, TimeUnit) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static Flowable range(int start, int count) { if (count < 0) { throw new IllegalArgumentException("count >= 0 required but it was " + count); } else if (count == 0) { return empty(); } else if (count == 1) { return just(start); } else if ((long)start + (count - 1) > Integer.MAX_VALUE) { throw new IllegalArgumentException("Integer overflow"); } return RxJavaPlugins.onAssembly(new FlowableRange(start, count)); } /** * Returns a {@code Flowable} that emits a sequence of {@link Long}s within a specified range. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and signals values on-demand (i.e., when requested).
*
Scheduler:
*
{@code rangeLong} does not operate by default on a particular {@link Scheduler}.
*
* * @param start * the value of the first {@code Long} in the sequence * @param count * the number of sequential {@code Long}s to generate * @return the new {@code Flowable} instance * @throws IllegalArgumentException * if {@code count} is less than zero, or if {@code start} + {@code count} − 1 exceeds * {@link Long#MAX_VALUE} * @see ReactiveX operators documentation: Range */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static Flowable rangeLong(long start, long count) { if (count < 0) { throw new IllegalArgumentException("count >= 0 required but it was " + count); } if (count == 0) { return empty(); } if (count == 1) { return just(start); } long end = start + (count - 1); if (start > 0 && end < 0) { throw new IllegalArgumentException("Overflow! start + count is bigger than Long.MAX_VALUE"); } return RxJavaPlugins.onAssembly(new FlowableRangeLong(start, count)); } /** * Returns a {@link Single} that emits a {@link Boolean} value that indicates whether two {@link Publisher} sequences are the * same by comparing the items emitted by each {@code Publisher} pairwise. *

* *

*
Backpressure:
*
This operator honors downstream backpressure and expects both of its sources * to honor backpressure as well. If violated, the operator will emit a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code sequenceEqual} does not operate by default on a particular {@link Scheduler}.
*
* * @param source1 * the first {@code Publisher} to compare * @param source2 * the second {@code Publisher} to compare * @param * the type of items emitted by each {@code Publisher} * @return the new {@code Single} instance * @throws NullPointerException if {@code source1} or {@code source2} is {@code null} * @see ReactiveX operators documentation: SequenceEqual */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Single sequenceEqual(@NonNull Publisher source1, @NonNull Publisher source2) { return sequenceEqual(source1, source2, ObjectHelper.equalsPredicate(), bufferSize()); } /** * Returns a {@link Single} that emits a {@link Boolean} value that indicates whether two {@link Publisher} sequences are the * same by comparing the items emitted by each {@code Publisher} pairwise based on the results of a specified * equality function. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator signals a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code sequenceEqual} does not operate by default on a particular {@link Scheduler}.
*
* * @param source1 * the first {@code Publisher} to compare * @param source2 * the second {@code Publisher} to compare * @param isEqual * a function used to compare items emitted by each {@code Publisher} * @param * the type of items emitted by each {@code Publisher} * @return the new {@code Single} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code isEqual} is {@code null} * @see ReactiveX operators documentation: SequenceEqual */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Single sequenceEqual(@NonNull Publisher source1, @NonNull Publisher source2, @NonNull BiPredicate isEqual) { return sequenceEqual(source1, source2, isEqual, bufferSize()); } /** * Returns a {@link Single} that emits a {@link Boolean} value that indicates whether two {@link Publisher} sequences are the * same by comparing the items emitted by each {@code Publisher} pairwise based on the results of a specified * equality function. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The source {@code Publisher}s are expected to honor * backpressure; if violated, the operator signals a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code sequenceEqual} does not operate by default on a particular {@link Scheduler}.
*
* * @param source1 * the first {@code Publisher} to compare * @param source2 * the second {@code Publisher} to compare * @param isEqual * a function used to compare items emitted by each {@code Publisher} * @param bufferSize * the number of items to prefetch from the first and second source {@code Publisher} * @param * the type of items emitted by each {@code Publisher} * @return the new {@code Single} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code isEqual} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: SequenceEqual */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Single sequenceEqual(@NonNull Publisher source1, @NonNull Publisher source2, @NonNull BiPredicate isEqual, int bufferSize) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(isEqual, "isEqual is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableSequenceEqualSingle<>(source1, source2, isEqual, bufferSize)); } /** * Returns a {@link Single} that emits a {@link Boolean} value that indicates whether two {@link Publisher} sequences are the * same by comparing the items emitted by each {@code Publisher} pairwise. *

* *

*
Backpressure:
*
This operator honors downstream backpressure and expects both of its sources * to honor backpressure as well. If violated, the operator will emit a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code sequenceEqual} does not operate by default on a particular {@link Scheduler}.
*
* * @param source1 * the first {@code Publisher} to compare * @param source2 * the second {@code Publisher} to compare * @param bufferSize * the number of items to prefetch from the first and second source {@code Publisher} * @param * the type of items emitted by each {@code Publisher} * @return the new {@code Single} instance * @throws NullPointerException if {@code source1} or {@code source2} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: SequenceEqual */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Single sequenceEqual(@NonNull Publisher source1, @NonNull Publisher source2, int bufferSize) { return sequenceEqual(source1, source2, ObjectHelper.equalsPredicate(), bufferSize); } /** * Converts a {@link Publisher} that emits {@code Publisher}s into a {@code Publisher} that emits the items emitted by the * most recently emitted of those {@code Publisher}s. *

* *

* {@code switchOnNext} subscribes to a {@code Publisher} that emits {@code Publisher}s. Each time it observes one of * these emitted {@code Publisher}s, the {@code Publisher} returned by {@code switchOnNext} begins emitting the items * emitted by that {@code Publisher}. When a new {@code Publisher} is emitted, {@code switchOnNext} stops emitting items * from the earlier-emitted {@code Publisher} and begins emitting items from the new one. *

* The resulting {@code Flowable} completes if both the outer {@code Publisher} and the last inner {@code Publisher}, if any, complete. * If the outer {@code Publisher} signals an {@code onError}, the inner {@code Publisher} is canceled and the error delivered in-sequence. *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed in an * unbounded manner (i.e., without backpressure) and the inner {@code Publisher}s are expected to honor * backpressure but it is not enforced; the operator won't signal a {@link MissingBackpressureException} * but the violation may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code switchOnNext} does not operate by default on a particular {@link Scheduler}.
*
* * @param the item type * @param sources * the source {@code Publisher} that emits {@code Publisher}s * @param bufferSize * the number of items to prefetch from the inner {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Switch */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable switchOnNext(@NonNull Publisher> sources, int bufferSize) { return fromPublisher(sources).switchMap((Function)Functions.identity(), bufferSize); } /** * Converts a {@link Publisher} that emits {@code Publisher}s into a {@code Publisher} that emits the items emitted by the * most recently emitted of those {@code Publisher}s. *

* *

* {@code switchOnNext} subscribes to a {@code Publisher} that emits {@code Publisher}s. Each time it observes one of * these emitted {@code Publisher}s, the {@code Publisher} returned by {@code switchOnNext} begins emitting the items * emitted by that {@code Publisher}. When a new {@code Publisher} is emitted, {@code switchOnNext} stops emitting items * from the earlier-emitted {@code Publisher} and begins emitting items from the new one. *

* The resulting {@code Flowable} completes if both the outer {@code Publisher} and the last inner {@code Publisher}, if any, complete. * If the outer {@code Publisher} signals an {@code onError}, the inner {@code Publisher} is canceled and the error delivered in-sequence. *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed in an * unbounded manner (i.e., without backpressure) and the inner {@code Publisher}s are expected to honor * backpressure but it is not enforced; the operator won't signal a {@link MissingBackpressureException} * but the violation may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code switchOnNext} does not operate by default on a particular {@link Scheduler}.
*
* * @param the item type * @param sources * the source {@code Publisher} that emits {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Switch */ @SuppressWarnings({ "unchecked", "rawtypes" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable switchOnNext(@NonNull Publisher<@NonNull ? extends Publisher> sources) { return fromPublisher(sources).switchMap((Function)Functions.identity()); } /** * Converts a {@link Publisher} that emits {@code Publisher}s into a {@code Publisher} that emits the items emitted by the * most recently emitted of those {@code Publisher}s and delays any exception until all {@code Publisher}s terminate. *

* *

* {@code switchOnNext} subscribes to a {@code Publisher} that emits {@code Publisher}s. Each time it observes one of * these emitted {@code Publisher}s, the {@code Publisher} returned by {@code switchOnNext} begins emitting the items * emitted by that {@code Publisher}. When a new {@code Publisher} is emitted, {@code switchOnNext} stops emitting items * from the earlier-emitted {@code Publisher} and begins emitting items from the new one. *

* The resulting {@code Flowable} completes if both the main {@code Publisher} and the last inner {@code Publisher}, if any, complete. * If the main {@code Publisher} signals an {@code onError}, the termination of the last inner {@code Publisher} will emit that error as is * or wrapped into a {@link CompositeException} along with the other possible errors the former inner {@code Publisher}s signaled. *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed in an * unbounded manner (i.e., without backpressure) and the inner {@code Publisher}s are expected to honor * backpressure but it is not enforced; the operator won't signal a {@link MissingBackpressureException} * but the violation may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code switchOnNextDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the item type * @param sources * the source {@code Publisher} that emits {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @see ReactiveX operators documentation: Switch * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable switchOnNextDelayError(@NonNull Publisher<@NonNull ? extends Publisher> sources) { return switchOnNextDelayError(sources, bufferSize()); } /** * Converts a {@link Publisher} that emits {@code Publisher}s into a {@code Publisher} that emits the items emitted by the * most recently emitted of those {@code Publisher}s and delays any exception until all {@code Publisher}s terminate. *

* *

* {@code switchOnNext} subscribes to a {@code Publisher} that emits {@code Publisher}s. Each time it observes one of * these emitted {@code Publisher}s, the {@code Publisher} returned by {@code switchOnNext} begins emitting the items * emitted by that {@code Publisher}. When a new {@code Publisher} is emitted, {@code switchOnNext} stops emitting items * from the earlier-emitted {@code Publisher} and begins emitting items from the new one. *

* The resulting {@code Flowable} completes if both the main {@code Publisher} and the last inner {@code Publisher}, if any, complete. * If the main {@code Publisher} signals an {@code onError}, the termination of the last inner {@code Publisher} will emit that error as is * or wrapped into a {@link CompositeException} along with the other possible errors the former inner {@code Publisher}s signaled. *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed in an * unbounded manner (i.e., without backpressure) and the inner {@code Publisher}s are expected to honor * backpressure but it is not enforced; the operator won't signal a {@link MissingBackpressureException} * but the violation may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code switchOnNextDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the item type * @param sources * the source {@code Publisher} that emits {@code Publisher}s * @param prefetch * the number of items to prefetch from the inner {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see ReactiveX operators documentation: Switch * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable switchOnNextDelayError(@NonNull Publisher<@NonNull ? extends Publisher> sources, int prefetch) { return fromPublisher(sources).switchMapDelayError(Functions.>identity(), prefetch); } /** * Returns a {@code Flowable} that emits {@code 0L} after a specified delay, and then completes. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. If the downstream needs a slower rate * it should slow the timer or use something like {@link #onBackpressureDrop}.
*
Scheduler:
*
{@code timer} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param delay * the initial delay before emitting a single {@code 0L} * @param unit * time units to use for {@code delay} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Timer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public static Flowable timer(long delay, @NonNull TimeUnit unit) { return timer(delay, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits {@code 0L} after a specified delay, on a specified {@link Scheduler}, and then * completes. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. If the downstream needs a slower rate * it should slow the timer or use something like {@link #onBackpressureDrop}.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param delay * the initial delay before emitting a single 0L * @param unit * time units to use for {@code delay} * @param scheduler * the {@code Scheduler} to use for scheduling the item * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Timer */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public static Flowable timer(long delay, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableTimer(Math.max(0L, delay), unit, scheduler)); } /** * Create a {@code Flowable} by wrapping a {@link Publisher} which has to be implemented according * to the Reactive Streams specification by handling backpressure and * cancellation correctly; no safeguards are provided by the {@code Flowable} itself. *
*
Backpressure:
*
This operator is a pass-through for backpressure and the behavior is determined by the * provided {@code Publisher} implementation.
*
Scheduler:
*
{@code unsafeCreate} by default doesn't operate on any particular {@link Scheduler}.
*
* @param the value type emitted * @param onSubscribe the {@code Publisher} instance to wrap * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onSubscribe} is {@code null} * @throws IllegalArgumentException if {@code onSubscribe} is a subclass of {@code Flowable}; such * instances don't need conversion and is possibly a port remnant from 1.x or one should use {@link #hide()} * instead. */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.NONE) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T> Flowable unsafeCreate(@NonNull Publisher onSubscribe) { Objects.requireNonNull(onSubscribe, "onSubscribe is null"); if (onSubscribe instanceof Flowable) { throw new IllegalArgumentException("unsafeCreate(Flowable) should be upgraded"); } return RxJavaPlugins.onAssembly(new FlowableFromPublisher<>(onSubscribe)); } /** * Constructs a {@code Flowable} that creates a dependent resource object, a {@link Publisher} with * that resource and calls the provided {@code resourceDisposer} function if this inner source terminates or the * downstream cancels the flow. *

* *

*
Backpressure:
*
The operator is a pass-through for backpressure and otherwise depends on the * backpressure support of the {@code Publisher} returned by the {@code resourceFactory}.
*
Scheduler:
*
{@code using} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the generated {@code Publisher} * @param the type of the resource associated with the output sequence * @param resourceSupplier * the factory function to create a resource object that depends on the {@code Publisher} * @param sourceSupplier * the factory function to create a {@code Publisher} * @param resourceCleanup * the function that will dispose of the resource * @return the new {@code Flowable} instance * @throws NullPointerException if {@code resourceSupplier}, {@code sourceSupplier} or {@code resourceCleanup} is {@code null} * @see ReactiveX operators documentation: Using */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T, @NonNull D> Flowable using( @NonNull Supplier resourceSupplier, @NonNull Function> sourceSupplier, @NonNull Consumer resourceCleanup) { return using(resourceSupplier, sourceSupplier, resourceCleanup, true); } /** * Constructs a {@code Flowable} that creates a dependent resource object, a {@link Publisher} with * that resource and calls the provided {@code resourceDisposer} function if this inner source terminates or the * downstream disposes the flow; doing it before these end-states have been reached if {@code eager == true}, after otherwise. *

* *

*
Backpressure:
*
The operator is a pass-through for backpressure and otherwise depends on the * backpressure support of the {@code Publisher} returned by the {@code resourceFactory}.
*
Scheduler:
*
{@code using} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the generated {@code Publisher} * @param the type of the resource associated with the output sequence * @param resourceSupplier * the factory function to create a resource object that depends on the {@code Publisher} * @param sourceSupplier * the factory function to create a {@code Publisher} * @param resourceCleanup * the function that will dispose of the resource * @param eager * If {@code true}, the resource disposal will happen either on a {@code cancel()} call before the upstream is disposed * or just before the emission of a terminal event ({@code onComplete} or {@code onError}). * If {@code false} the resource disposal will happen either on a {@code cancel()} call after the upstream is disposed * or just after the emission of a terminal event ({@code onComplete} or {@code onError}). * @return the new {@code Flowable} instance * @throws NullPointerException if {@code resourceSupplier}, {@code sourceSupplier} or {@code resourceCleanup} is {@code null} * @see ReactiveX operators documentation: Using * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T, @NonNull D> Flowable using( @NonNull Supplier resourceSupplier, @NonNull Function> sourceSupplier, @NonNull Consumer resourceCleanup, boolean eager) { Objects.requireNonNull(resourceSupplier, "resourceSupplier is null"); Objects.requireNonNull(sourceSupplier, "sourceSupplier is null"); Objects.requireNonNull(resourceCleanup, "resourceCleanup is null"); return RxJavaPlugins.onAssembly(new FlowableUsing(resourceSupplier, sourceSupplier, resourceCleanup, eager)); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * items emitted, in sequence, by an {@link Iterable} of other {@link Publisher}s. *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by each of the source {@code Publisher}s; * the second item emitted by the new {@code Publisher} will be the result of the function applied to the second * item emitted by each of those {@code Publisher}s; and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@code onNext} as many times as * the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(Arrays.asList(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2)), (a) -> a)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *

* *

*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common value type * @param the zipped result type * @param sources * an {@code Iterable} of source {@code Publisher}s * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T, @NonNull R> Flowable zip(@NonNull Iterable<@NonNull ? extends Publisher> sources, @NonNull Function zipper) { Objects.requireNonNull(zipper, "zipper is null"); Objects.requireNonNull(sources, "sources is null"); return RxJavaPlugins.onAssembly(new FlowableZip<>(null, sources, zipper, bufferSize(), false)); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * items emitted, in sequence, by an {@link Iterable} of other {@link Publisher}s. *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by each of the source {@code Publisher}s; * the second item emitted by the new {@code Publisher} will be the result of the function applied to the second * item emitted by each of those {@code Publisher}s; and so forth. *

* The resulting {@code Floawble} returned from {@code zip} will invoke {@code onNext} as many times as * the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(Arrays.asList(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2)), (a) -> a)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *

* *

*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * * @param sources * an {@code Iterable} of source {@code Publisher}s * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @param delayError * delay errors signaled by any of the source {@code Publisher} until all {@code Publisher}s terminate * @param bufferSize * the number of elements to prefetch from each source {@code Publisher} * @param the common source value type * @param the zipped result type * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code zipper} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T, @NonNull R> Flowable zip(@NonNull Iterable<@NonNull ? extends Publisher> sources, @NonNull Function zipper, boolean delayError, int bufferSize) { Objects.requireNonNull(zipper, "zipper is null"); Objects.requireNonNull(sources, "sources is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableZip<>(null, sources, zipper, bufferSize, delayError)); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * two items emitted, in sequence, by two other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by {@code o1} and the first item * emitted by {@code o2}; the second item emitted by the new {@code Publisher} will be the result of the function * applied to the second item emitted by {@code o1} and the second item emitted by {@code o2}; and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), (a, b) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results * in an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull BiFunction zipper) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), false, bufferSize(), source1, source2); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * two items emitted, in sequence, by two other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by {@code o1} and the first item * emitted by {@code o2}; the second item emitted by the new {@code Publisher} will be the result of the function * applied to the second item emitted by {@code o1} and the second item emitted by {@code o2}; and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), (a, b) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results * in an item that will be emitted by the resulting {@code Flowable} * @param delayError delay errors from any of the source {@code Publisher}s till the other terminates * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull BiFunction zipper, boolean delayError) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), delayError, bufferSize(), source1, source2); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * two items emitted, in sequence, by two other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by {@code o1} and the first item * emitted by {@code o2}; the second item emitted by the new {@code Publisher} will be the result of the function * applied to the second item emitted by {@code o1} and the second item emitted by {@code o2}; and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), (a, b) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results * in an item that will be emitted by the resulting {@code Flowable} * @param delayError delay errors from any of the source {@code Publisher}s till the other terminates * @param bufferSize the number of elements to prefetch from each source {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code zipper} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull BiFunction zipper, boolean delayError, int bufferSize) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), delayError, bufferSize, source1, source2); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * three items emitted, in sequence, by three other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by {@code o1}, the first item * emitted by {@code o2}, and the first item emitted by {@code o3}; the second item emitted by the new * {@code Publisher} will be the result of the function applied to the second item emitted by {@code o1}, the * second item emitted by {@code o2}, and the second item emitted by {@code o3}; and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), ..., (a, b, c) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the value type of the third source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param source3 * a third source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Function3 zipper) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), false, bufferSize(), source1, source2, source3); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * four items emitted, in sequence, by four other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by {@code o1}, the first item * emitted by {@code o2}, the first item emitted by {@code o3}, and the first item emitted by {@code 04}; * the second item emitted by the new {@code Publisher} will be the result of the function applied to the second * item emitted by each of those {@code Publisher}s; and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), ..., (a, b, c, d) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the value type of the third source * @param the value type of the fourth source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param source3 * a third source {@code Publisher} * @param source4 * a fourth source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Function4 zipper) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), false, bufferSize(), source1, source2, source3, source4); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * five items emitted, in sequence, by five other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by {@code o1}, the first item * emitted by {@code o2}, the first item emitted by {@code o3}, the first item emitted by {@code o4}, and * the first item emitted by {@code o5}; the second item emitted by the new {@code Publisher} will be the result of * the function applied to the second item emitted by each of those {@code Publisher}s; and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), ..., (a, b, c, d, e) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the value type of the third source * @param the value type of the fourth source * @param the value type of the fifth source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param source3 * a third source {@code Publisher} * @param source4 * a fourth source {@code Publisher} * @param source5 * a fifth source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Function5 zipper) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), false, bufferSize(), source1, source2, source3, source4, source5); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * six items emitted, in sequence, by six other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by each source {@code Publisher}, the * second item emitted by the new {@code Publisher} will be the result of the function applied to the second item * emitted by each of those {@code Publisher}s, and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), ..., (a, b, c, d, e, f) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the value type of the third source * @param the value type of the fourth source * @param the value type of the fifth source * @param the value type of the sixth source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param source3 * a third source {@code Publisher} * @param source4 * a fourth source {@code Publisher} * @param source5 * a fifth source {@code Publisher} * @param source6 * a sixth source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5}, {@code source6} * or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull T6, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Publisher source6, @NonNull Function6 zipper) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(source6, "source6 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), false, bufferSize(), source1, source2, source3, source4, source5, source6); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * seven items emitted, in sequence, by seven other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by each source {@code Publisher}, the * second item emitted by the new {@code Publisher} will be the result of the function applied to the second item * emitted by each of those {@code Publisher}s, and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), ..., (a, b, c, d, e, f, g) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the value type of the third source * @param the value type of the fourth source * @param the value type of the fifth source * @param the value type of the sixth source * @param the value type of the seventh source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param source3 * a third source {@code Publisher} * @param source4 * a fourth source {@code Publisher} * @param source5 * a fifth source {@code Publisher} * @param source6 * a sixth source {@code Publisher} * @param source7 * a seventh source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5}, {@code source6}, * {@code source7} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull T6, @NonNull T7, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Publisher source6, @NonNull Publisher source7, @NonNull Function7 zipper) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(source6, "source6 is null"); Objects.requireNonNull(source7, "source7 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), false, bufferSize(), source1, source2, source3, source4, source5, source6, source7); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * eight items emitted, in sequence, by eight other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by each source {@code Publisher}, the * second item emitted by the new {@code Publisher} will be the result of the function applied to the second item * emitted by each of those {@code Publisher}s, and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), ..., (a, b, c, d, e, f, g, h) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the value type of the third source * @param the value type of the fourth source * @param the value type of the fifth source * @param the value type of the sixth source * @param the value type of the seventh source * @param the value type of the eighth source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param source3 * a third source {@code Publisher} * @param source4 * a fourth source {@code Publisher} * @param source5 * a fifth source {@code Publisher} * @param source6 * a sixth source {@code Publisher} * @param source7 * a seventh source {@code Publisher} * @param source8 * an eighth source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5}, {@code source6}, * {@code source7}, {@code source8} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull T6, @NonNull T7, @NonNull T8, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Publisher source6, @NonNull Publisher source7, @NonNull Publisher source8, @NonNull Function8 zipper) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(source6, "source6 is null"); Objects.requireNonNull(source7, "source7 is null"); Objects.requireNonNull(source8, "source8 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), false, bufferSize(), source1, source2, source3, source4, source5, source6, source7, source8); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * nine items emitted, in sequence, by nine other {@link Publisher}s. *

* *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by each source {@code Publisher}, the * second item emitted by the new {@code Publisher} will be the result of the function applied to the second item * emitted by each of those {@code Publisher}s, and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@link Subscriber#onNext onNext} * as many times as the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest * items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2), ..., (a, b, c, d, e, f, g, h, i) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *
*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zip} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the first source * @param the value type of the second source * @param the value type of the third source * @param the value type of the fourth source * @param the value type of the fifth source * @param the value type of the sixth source * @param the value type of the seventh source * @param the value type of the eighth source * @param the value type of the ninth source * @param the zipped result type * @param source1 * the first source {@code Publisher} * @param source2 * a second source {@code Publisher} * @param source3 * a third source {@code Publisher} * @param source4 * a fourth source {@code Publisher} * @param source5 * a fifth source {@code Publisher} * @param source6 * a sixth source {@code Publisher} * @param source7 * a seventh source {@code Publisher} * @param source8 * an eighth source {@code Publisher} * @param source9 * a ninth source {@code Publisher} * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4}, {@code source5}, {@code source6}, * {@code source7}, {@code source8}, {@code source9} * or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public static <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull T5, @NonNull T6, @NonNull T7, @NonNull T8, @NonNull T9, @NonNull R> Flowable zip( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Publisher source5, @NonNull Publisher source6, @NonNull Publisher source7, @NonNull Publisher source8, @NonNull Publisher source9, @NonNull Function9 zipper) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(source5, "source5 is null"); Objects.requireNonNull(source6, "source6 is null"); Objects.requireNonNull(source7, "source7 is null"); Objects.requireNonNull(source8, "source8 is null"); Objects.requireNonNull(source9, "source9 is null"); Objects.requireNonNull(zipper, "zipper is null"); return zipArray(Functions.toFunction(zipper), false, bufferSize(), source1, source2, source3, source4, source5, source6, source7, source8, source9); } /** * Returns a {@code Flowable} that emits the results of a specified combiner function applied to combinations of * items emitted, in sequence, by an array of other {@link Publisher}s. *

* {@code zip} applies this function in strict sequence, so the first item emitted by the new {@code Publisher} * will be the result of the function applied to the first item emitted by each of the source {@code Publisher}s; * the second item emitted by the new {@code Publisher} will be the result of the function applied to the second * item emitted by each of those {@code Publisher}s; and so forth. *

* The resulting {@code Flowable} returned from {@code zip} will invoke {@code onNext} as many times as * the number of {@code onNext} invocations of the source {@code Publisher} that emits the fewest items. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

zip(new Publisher[]{range(1, 5).doOnComplete(action1), range(6, 5).doOnComplete(action2)}, (a) ->
     * a)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *

* *

*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zipArray} does not operate by default on a particular {@link Scheduler}.
*
* * @param the common element type * @param the result type * @param sources * an array of source {@code Publisher}s * @param zipper * a function that, when applied to an item emitted by each of the source {@code Publisher}s, results in * an item that will be emitted by the resulting {@code Flowable} * @param delayError * delay errors signaled by any of the source {@code Publisher} until all {@code Publisher}s terminate * @param bufferSize * the number of elements to prefetch from each source {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sources} or {@code zipper} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs public static <@NonNull T, @NonNull R> Flowable zipArray(@NonNull Function zipper, boolean delayError, int bufferSize, @NonNull Publisher... sources) { Objects.requireNonNull(sources, "sources is null"); if (sources.length == 0) { return empty(); } Objects.requireNonNull(zipper, "zipper is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableZip<>(sources, null, zipper, bufferSize, delayError)); } // *************************************************************************************************** // Instance operators // *************************************************************************************************** /** * Returns a {@link Single} that emits a {@link Boolean} that indicates whether all of the items emitted by the current * {@code Flowable} satisfy a condition. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code all} does not operate by default on a particular {@link Scheduler}.
*
* * @param predicate * a function that evaluates an item and returns a {@code Boolean} * @return the new {@code Single} instance * @throws NullPointerException if {@code predicate} is {@code null} * @see ReactiveX operators documentation: All */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Single all(@NonNull Predicate predicate) { Objects.requireNonNull(predicate, "predicate is null"); return RxJavaPlugins.onAssembly(new FlowableAllSingle<>(this, predicate)); } /** * Mirrors the {@link Publisher} (current or provided) that first either emits an item or sends a termination * notification. *

* *

* When the current {@code Flowable} signals an item or terminates first, the subscription to the other * {@code Publisher} is canceled. If the other {@code Publisher} signals an item or terminates first, * the subscription to the current {@code Flowable} is canceled. *

*
Backpressure:
*
The operator itself doesn't interfere with backpressure which is determined by the winning * {@code Publisher}'s backpressure behavior.
*
Scheduler:
*
{@code ambWith} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
* If the losing {@code Publisher} signals an error, the error is routed to the global * error handler via {@link RxJavaPlugins#onError(Throwable)}. *
*
* * @param other * a {@code Publisher} competing to react first. A subscription to this provided {@code Publisher} will occur after subscribing * to the current {@code Flowable}. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @see ReactiveX operators documentation: Amb */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable ambWith(@NonNull Publisher other) { Objects.requireNonNull(other, "other is null"); return ambArray(this, other); } /** * Returns a {@link Single} that emits {@code true} if any item emitted by the current {@code Flowable} satisfies a * specified condition, otherwise {@code false}. Note: this always emits {@code false} if the * current {@code Flowable} is empty. *

* *

* In Rx.Net this is the {@code any} operator but we renamed it in RxJava to better match Java naming * idioms. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code any} does not operate by default on a particular {@link Scheduler}.
*
* * @param predicate * the condition to test items emitted by the current {@code Flowable} * @return the new {@code Single} instance * @throws NullPointerException if {@code predicate} is {@code null} * @see ReactiveX operators documentation: Contains */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Single any(@NonNull Predicate predicate) { Objects.requireNonNull(predicate, "predicate is null"); return RxJavaPlugins.onAssembly(new FlowableAnySingle<>(this, predicate)); } /** * Returns the first item emitted by this {@code Flowable}, or throws * {@link NoSuchElementException} if it emits no items. *
*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingFirst} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the source signals an error, the operator wraps a checked {@link Exception} * into {@link RuntimeException} and throws that. Otherwise, {@code RuntimeException}s and * {@link Error}s are rethrown as they are.
*
* * @return the new {@code Flowable} instance * @throws NoSuchElementException * if this {@code Flowable} emits no items * @see ReactiveX documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final T blockingFirst() { BlockingFirstSubscriber s = new BlockingFirstSubscriber<>(); subscribe(s); T v = s.blockingGet(); if (v != null) { return v; } throw new NoSuchElementException(); } /** * Returns the first item emitted by this {@code Flowable}, or a default value if it emits no * items. *
*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingFirst} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the source signals an error, the operator wraps a checked {@link Exception} * into {@link RuntimeException} and throws that. Otherwise, {@code RuntimeException}s and * {@link Error}s are rethrown as they are.
*
* * @param defaultItem * a default value to return if this {@code Flowable} emits no items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code defaultItem} is {@code null} * @see ReactiveX documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final T blockingFirst(@NonNull T defaultItem) { Objects.requireNonNull(defaultItem, "defaultItem is null"); BlockingFirstSubscriber s = new BlockingFirstSubscriber<>(); subscribe(s); T v = s.blockingGet(); return v != null ? v : defaultItem; } /** * Consumes the current {@code Flowable} in a blocking fashion and invokes the given * {@link Consumer} with each upstream item on the current thread until the * upstream terminates. *

* *

* Note: the method will only return if the upstream terminates or the current * thread is interrupted. *

* This method executes the {@code Consumer} on the current thread while * {@link #subscribe(Consumer)} executes the consumer on the original caller thread of the * sequence. *

*
Backpressure:
*
The operator requests {@link Flowable#bufferSize()} upfront, then 75% of this * amount when 75% is received.
*
Scheduler:
*
{@code blockingForEach} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the source signals an error, the operator wraps a checked {@link Exception} * into {@link RuntimeException} and throws that. Otherwise, {@code RuntimeException}s and * {@link Error}s are rethrown as they are.
*
* * @param onNext * the {@code Consumer} to invoke for each item emitted by the {@code Flowable} * @throws NullPointerException if {@code onNext} is {@code null} * @throws RuntimeException * if an error occurs; {@code Error}s and {@code RuntimeException}s are rethrown * as they are, checked {@code Exception}s are wrapped into {@code RuntimeException}s * @see ReactiveX documentation: Subscribe * @see #subscribe(Consumer) * @see #blockingForEach(Consumer, int) */ @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingForEach(@NonNull Consumer onNext) { blockingForEach(onNext, bufferSize()); } /** * Consumes the current {@code Flowable} in a blocking fashion and invokes the given * {@link Consumer} with each upstream item on the current thread until the * upstream terminates. *

* *

* Note: the method will only return if the upstream terminates or the current * thread is interrupted. *

* This method executes the {@code Consumer} on the current thread while * {@link #subscribe(Consumer)} executes the consumer on the original caller thread of the * sequence. *

*
Backpressure:
*
The operator requests the given {@code prefetch} amount upfront, then 75% of this * amount when 75% is received.
*
Scheduler:
*
{@code blockingForEach} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the source signals an error, the operator wraps a checked {@link Exception} * into {@link RuntimeException} and throws that. Otherwise, {@code RuntimeException}s and * {@link Error}s are rethrown as they are.
*
* * @param onNext * the {@code Consumer} to invoke for each item emitted by the {@code Flowable} * @param bufferSize * the number of items to prefetch upfront, then 75% of it after 75% received * @throws NullPointerException if {@code onNext} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @throws RuntimeException * if an error occurs; {@code Error}s and {@code RuntimeException}s are rethrown * as they are, checked {@code Exception}s are wrapped into {@code RuntimeException}s * @see ReactiveX documentation: Subscribe * @see #subscribe(Consumer) */ @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingForEach(@NonNull Consumer onNext, int bufferSize) { Objects.requireNonNull(onNext, "onNext is null"); Iterator it = blockingIterable(bufferSize).iterator(); while (it.hasNext()) { try { onNext.accept(it.next()); } catch (Throwable e) { Exceptions.throwIfFatal(e); ((Disposable)it).dispose(); throw ExceptionHelper.wrapOrThrow(e); } } } /** * Converts this {@code Flowable} into an {@link Iterable}. *

* *

*
Backpressure:
*
The operator expects the upstream to honor backpressure otherwise the returned * {@code Iterable}'s iterator will throw a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code blockingIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Iterable} instance * @see ReactiveX documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Iterable blockingIterable() { return blockingIterable(bufferSize()); } /** * Converts this {@code Flowable} into an {@link Iterable}. *

* *

*
Backpressure:
*
The operator expects the upstream to honor backpressure otherwise the returned * {@code Iterable}'s iterator will throw a {@link MissingBackpressureException}. *
*
Scheduler:
*
{@code blockingIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param bufferSize the number of items to prefetch from the current {@code Flowable} * @return the new {@code Iterable} instance * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Iterable blockingIterable(int bufferSize) { ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return new BlockingFlowableIterable<>(this, bufferSize); } /** * Returns the last item emitted by this {@code Flowable}, or throws * {@link NoSuchElementException} if this {@code Flowable} emits no items. *

* *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingLast} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the source signals an error, the operator wraps a checked {@link Exception} * into {@link RuntimeException} and throws that. Otherwise, {@code RuntimeException}s and * {@link Error}s are rethrown as they are.
*
* * @return the new {@code Flowable} instance * @throws NoSuchElementException * if this {@code Flowable} emits no items * @see ReactiveX documentation: Last */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final T blockingLast() { BlockingLastSubscriber s = new BlockingLastSubscriber<>(); subscribe(s); T v = s.blockingGet(); if (v != null) { return v; } throw new NoSuchElementException(); } /** * Returns the last item emitted by this {@code Flowable}, or a default value if it emits no * items. *

* *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingLast} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the source signals an error, the operator wraps a checked {@link Exception} * into {@link RuntimeException} and throws that. Otherwise, {@code RuntimeException}s and * {@link Error}s are rethrown as they are.
*
* * @param defaultItem * a default value to return if this {@code Flowable} emits no items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code defaultItem} is {@code null} * @see ReactiveX documentation: Last */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final T blockingLast(@NonNull T defaultItem) { Objects.requireNonNull(defaultItem, "defaultItem is null"); BlockingLastSubscriber s = new BlockingLastSubscriber<>(); subscribe(s); T v = s.blockingGet(); return v != null ? v : defaultItem; } /** * Returns an {@link Iterable} that returns the latest item emitted by this {@code Flowable}, * waiting if necessary for one to become available. *

* If this {@code Flowable} produces items faster than {@code Iterator.next} takes them, * {@code onNext} events might be skipped, but {@code onError} or {@code onComplete} events are not. *

* Note also that an {@code onNext} directly followed by {@code onComplete} might hide the {@code onNext} * event. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Iterable} instance * @see ReactiveX documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Iterable blockingLatest() { return new BlockingFlowableLatest<>(this); } /** * Returns an {@link Iterable} that always returns the item most recently emitted by this * {@code Flowable}. *

* *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingMostRecent} does not operate by default on a particular {@link Scheduler}.
*
* * @param initialItem * the initial item that the {@code Iterable} sequence will yield if this * {@code Flowable} has not yet emitted an item * @return the new {@code Iterable} instance * @throws NullPointerException if {@code initialItem} is {@code null} * @see ReactiveX documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Iterable blockingMostRecent(@NonNull T initialItem) { Objects.requireNonNull(initialItem, "initialItem is null"); return new BlockingFlowableMostRecent<>(this, initialItem); } /** * Returns an {@link Iterable} that blocks until this {@code Flowable} emits another item, then * returns that item. *

* *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingNext} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Iterable} instance * @see ReactiveX documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Iterable blockingNext() { return new BlockingFlowableNext<>(this); } /** * If this {@code Flowable} completes after emitting a single item, return that item, otherwise * throw a {@link NoSuchElementException}. *

* *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingSingle} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the source signals an error, the operator wraps a checked {@link Exception} * into {@link RuntimeException} and throws that. Otherwise, {@code RuntimeException}s and * {@link Error}s are rethrown as they are.
*
* * @return the new {@code Flowable} instance * @see ReactiveX documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final T blockingSingle() { return singleOrError().blockingGet(); } /** * If this {@code Flowable} completes after emitting a single item, return that item; if it emits * more than one item, throw an {@link IllegalArgumentException}; if it emits no items, return a default * value. *

* *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingSingle} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the source signals an error, the operator wraps a checked {@link Exception} * into {@link RuntimeException} and throws that. Otherwise, {@code RuntimeException}s and * {@link Error}s are rethrown as they are.
*
* * @param defaultItem * a default value to return if this {@code Flowable} emits no items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code defaultItem} is {@code null} * @see ReactiveX documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final T blockingSingle(@NonNull T defaultItem) { return single(defaultItem).blockingGet(); } /** * Returns a {@link Future} representing the only value emitted by this {@code Flowable}. *

* *

* If the {@code Flowable} emits more than one item, {@link java.util.concurrent.Future} will receive an * {@link java.lang.IndexOutOfBoundsException}. If the {@code Flowable} is empty, {@link java.util.concurrent.Future} * will receive a {@link java.util.NoSuchElementException}. The {@code Flowable} source has to terminate in order * for the returned {@code Future} to terminate as well. *

* If the {@code Flowable} may emit more than one item, use {@code Flowable.toList().toFuture()}. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code toFuture} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Future} instance * @see ReactiveX documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Future toFuture() { return subscribeWith(new FutureSubscriber<>()); } /** * Runs the current {@code Flowable} to a terminal event, ignoring any values and rethrowing any exception. *

* Note that calling this method will block the caller thread until the upstream terminates * normally or with an error. Therefore, calling this method from special threads such as the * Android Main Thread or the Swing Event Dispatch Thread is not recommended. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingSubscribe} does not operate by default on a particular {@link Scheduler}.
*
* @since 2.0 * @see #blockingSubscribe(Consumer) * @see #blockingSubscribe(Consumer, Consumer) * @see #blockingSubscribe(Consumer, Consumer, Action) */ @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingSubscribe() { FlowableBlockingSubscribe.subscribe(this); } /** * Subscribes to the source and calls the given callbacks on the current thread. *

* If the {@code Flowable} emits an error, it is wrapped into an * {@link io.reactivex.rxjava3.exceptions.OnErrorNotImplementedException OnErrorNotImplementedException} * and routed to the {@link RxJavaPlugins#onError(Throwable)} handler. * Using the overloads {@link #blockingSubscribe(Consumer, Consumer)} * or {@link #blockingSubscribe(Consumer, Consumer, Action)} instead is recommended. *

* Note that calling this method will block the caller thread until the upstream terminates * normally or with an error. Therefore, calling this method from special threads such as the * Android Main Thread or the Swing Event Dispatch Thread is not recommended. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingSubscribe} does not operate by default on a particular {@link Scheduler}.
*
* @param onNext the callback action for each source value * @throws NullPointerException if {@code onNext} is {@code null} * @since 2.0 * @see #blockingSubscribe(Consumer, Consumer) * @see #blockingSubscribe(Consumer, Consumer, Action) */ @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingSubscribe(@NonNull Consumer onNext) { FlowableBlockingSubscribe.subscribe(this, onNext, Functions.ON_ERROR_MISSING, Functions.EMPTY_ACTION); } /** * Subscribes to the source and calls the given callbacks on the current thread. *

* If the {@code Flowable} emits an error, it is wrapped into an * {@link io.reactivex.rxjava3.exceptions.OnErrorNotImplementedException OnErrorNotImplementedException} * and routed to the {@link RxJavaPlugins#onError(Throwable)} handler. * Using the overloads {@link #blockingSubscribe(Consumer, Consumer)} * or {@link #blockingSubscribe(Consumer, Consumer, Action)} instead is recommended. *

* Note that calling this method will block the caller thread until the upstream terminates * normally or with an error. Therefore, calling this method from special threads such as the * Android Main Thread or the Swing Event Dispatch Thread is not recommended. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an bounded manner (up to bufferSize * outstanding request amount for items).
*
Scheduler:
*
{@code blockingSubscribe} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.15 - experimental * @param onNext the callback action for each source value * @param bufferSize the size of the buffer * @throws NullPointerException if {@code onNext} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see #blockingSubscribe(Consumer, Consumer) * @see #blockingSubscribe(Consumer, Consumer, Action) * @since 2.2 */ @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingSubscribe(@NonNull Consumer onNext, int bufferSize) { FlowableBlockingSubscribe.subscribe(this, onNext, Functions.ON_ERROR_MISSING, Functions.EMPTY_ACTION, bufferSize); } /** * Subscribes to the source and calls the given callbacks on the current thread. *

* Note that calling this method will block the caller thread until the upstream terminates * normally or with an error. Therefore, calling this method from special threads such as the * Android Main Thread or the Swing Event Dispatch Thread is not recommended. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingSubscribe} does not operate by default on a particular {@link Scheduler}.
*
* @param onNext the callback action for each source value * @param onError the callback action for an error event * @throws NullPointerException if {@code onNext} or {@code onError} is {@code null} * @since 2.0 * @see #blockingSubscribe(Consumer, Consumer, Action) */ @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingSubscribe(@NonNull Consumer onNext, @NonNull Consumer onError) { FlowableBlockingSubscribe.subscribe(this, onNext, onError, Functions.EMPTY_ACTION); } /** * Subscribes to the source and calls the given callbacks on the current thread. *

* Note that calling this method will block the caller thread until the upstream terminates * normally or with an error. Therefore, calling this method from special threads such as the * Android Main Thread or the Swing Event Dispatch Thread is not recommended. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an bounded manner (up to bufferSize * outstanding request amount for items).
*
Scheduler:
*
{@code blockingSubscribe} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.15 - experimental * @param onNext the callback action for each source value * @param onError the callback action for an error event * @param bufferSize the size of the buffer * @throws NullPointerException if {@code onNext} or {@code onError} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @since 2.2 * @see #blockingSubscribe(Consumer, Consumer, Action) */ @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingSubscribe(@NonNull Consumer onNext, @NonNull Consumer onError, int bufferSize) { FlowableBlockingSubscribe.subscribe(this, onNext, onError, Functions.EMPTY_ACTION, bufferSize); } /** * Subscribes to the source and calls the given callbacks on the current thread. *

* Note that calling this method will block the caller thread until the upstream terminates * normally or with an error. Therefore, calling this method from special threads such as the * Android Main Thread or the Swing Event Dispatch Thread is not recommended. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner * (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code blockingSubscribe} does not operate by default on a particular {@link Scheduler}.
*
* @param onNext the callback action for each source value * @param onError the callback action for an error event * @param onComplete the callback action for the completion event. * @throws NullPointerException if {@code onNext}, {@code onError} or {@code onComplete} is {@code null} * @since 2.0 */ @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingSubscribe(@NonNull Consumer onNext, @NonNull Consumer onError, @NonNull Action onComplete) { FlowableBlockingSubscribe.subscribe(this, onNext, onError, onComplete); } /** * Subscribes to the source and calls the given callbacks on the current thread. *

* Note that calling this method will block the caller thread until the upstream terminates * normally or with an error. Therefore, calling this method from special threads such as the * Android Main Thread or the Swing Event Dispatch Thread is not recommended. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an bounded manner (up to bufferSize * outstanding request amount for items).
*
Scheduler:
*
{@code blockingSubscribe} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.15 - experimental * @param onNext the callback action for each source value * @param onError the callback action for an error event * @param onComplete the callback action for the completion event. * @param bufferSize the size of the buffer * @throws NullPointerException if {@code onNext}, {@code onError} or {@code onComplete} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @since 2.2 */ @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingSubscribe(@NonNull Consumer onNext, @NonNull Consumer onError, @NonNull Action onComplete, int bufferSize) { FlowableBlockingSubscribe.subscribe(this, onNext, onError, onComplete, bufferSize); } /** * Subscribes to the source and calls the {@link Subscriber} methods on the current thread. *

* Note that calling this method will block the caller thread until the upstream terminates * normally, with an error or the {@code Subscriber} cancels the {@link Subscription} it receives via * {@link Subscriber#onSubscribe(Subscription)}. * Therefore, calling this method from special threads such as the * Android Main Thread or the Swing Event Dispatch Thread is not recommended. *

*
Backpressure:
*
The supplied {@code Subscriber} determines how backpressure is applied.
*
Scheduler:
*
{@code blockingSubscribe} does not operate by default on a particular {@link Scheduler}.
*
* The cancellation and backpressure is composed through. * @param subscriber the subscriber to forward events and calls to in the current thread * @throws NullPointerException if {@code subscriber} is {@code null} * @since 2.0 */ @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public final void blockingSubscribe(@NonNull Subscriber subscriber) { Objects.requireNonNull(subscriber, "subscriber is null"); FlowableBlockingSubscribe.subscribe(this, subscriber); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping buffers, each containing {@code count} items. When the current * {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer and propagates the notification from the * current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} notification the event is passed on * immediately without first emitting the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and expects the current {@code Flowable} to honor it as * well, although not enforced; violation may lead to {@link MissingBackpressureException} somewhere * downstream.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * the maximum number of items in each buffer before it should be emitted * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> buffer(int count) { return buffer(count, count); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits buffers every {@code skip} items, each containing {@code count} items. When the current * {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer and propagates the notification from the * current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} notification the event is passed on * immediately without first emitting the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and expects the current {@code Flowable} to honor it as * well, although not enforced; violation may lead to {@link MissingBackpressureException} somewhere * downstream.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * the maximum size of each buffer before it should be emitted * @param skip * how many items emitted by the current {@code Flowable} should be skipped before starting a new * buffer. Note that when {@code skip} and {@code count} are equal, this is the same operation as * {@link #buffer(int)}. * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code count} or {@code skip} is non-positive * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> buffer(int count, int skip) { return buffer(count, skip, ArrayListSupplier.asSupplier()); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits buffers every {@code skip} items, each containing {@code count} items. When the current * {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer and propagates the notification from the * current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} notification the event is passed on * immediately without first emitting the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and expects the current {@code Flowable} to honor it as * well, although not enforced; violation may lead to {@link MissingBackpressureException} somewhere * downstream.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param the collection subclass type to buffer into * @param count * the maximum size of each buffer before it should be emitted * @param skip * how many items emitted by the current {@code Flowable} should be skipped before starting a new * buffer. Note that when {@code skip} and {@code count} are equal, this is the same operation as * {@link #buffer(int)}. * @param bufferSupplier * a factory function that returns an instance of the collection subclass to be used and returned * as the buffer * @return the new {@code Flowable} instance * @throws NullPointerException if {@code bufferSupplier} is {@code null} * @throws IllegalArgumentException if {@code count} or {@code skip} is non-positive * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U extends Collection> Flowable buffer(int count, int skip, @NonNull Supplier bufferSupplier) { ObjectHelper.verifyPositive(count, "count"); ObjectHelper.verifyPositive(skip, "skip"); Objects.requireNonNull(bufferSupplier, "bufferSupplier is null"); return RxJavaPlugins.onAssembly(new FlowableBuffer<>(this, count, skip, bufferSupplier)); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping buffers, each containing {@code count} items. When the current * {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer and propagates the notification from the * current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} notification the event is passed on * immediately without first emitting the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and expects the current {@code Flowable} to honor it as * well, although not enforced; violation may lead to {@link MissingBackpressureException} somewhere * downstream.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param the collection subclass type to buffer into * @param count * the maximum number of items in each buffer before it should be emitted * @param bufferSupplier * a factory function that returns an instance of the collection subclass to be used and returned * as the buffer * @return the new {@code Flowable} instance * @throws NullPointerException if {@code bufferSupplier} is {@code null} * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U extends Collection> Flowable buffer(int count, @NonNull Supplier bufferSupplier) { return buffer(count, count, bufferSupplier); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} starts a new buffer periodically, as determined by the {@code timeskip} argument. It emits * each buffer after a fixed timespan, specified by the {@code timespan} argument. When the current * {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer and propagates the notification from the * current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} notification the event is passed on * immediately without first emitting the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. It requests {@link Long#MAX_VALUE} * upstream and does not obey downstream requests.
*
Scheduler:
*
This version of {@code buffer} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timespan * the period of time each buffer collects items before it is emitted * @param timeskip * the period of time after which a new buffer will be created * @param unit * the unit of time that applies to the {@code timespan} and {@code timeskip} arguments * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable> buffer(long timespan, long timeskip, @NonNull TimeUnit unit) { return buffer(timespan, timeskip, unit, Schedulers.computation(), ArrayListSupplier.asSupplier()); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} starts a new buffer periodically, as determined by the {@code timeskip} argument, and on the * specified {@code scheduler}. It emits each buffer after a fixed timespan, specified by the * {@code timespan} argument. When the current {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer * and propagates the notification from the current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} * notification the event is passed on immediately without first emitting the buffer it is in the process of * assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. It requests {@link Long#MAX_VALUE} * upstream and does not obey downstream requests.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each buffer collects items before it is emitted * @param timeskip * the period of time after which a new buffer will be created * @param unit * the unit of time that applies to the {@code timespan} and {@code timeskip} arguments * @param scheduler * the {@code Scheduler} to use when determining the end and start of a buffer * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable> buffer(long timespan, long timeskip, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return buffer(timespan, timeskip, unit, scheduler, ArrayListSupplier.asSupplier()); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} starts a new buffer periodically, as determined by the {@code timeskip} argument, and on the * specified {@code scheduler}. It emits each buffer after a fixed timespan, specified by the * {@code timespan} argument. When the current {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer * and propagates the notification from the current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} * notification the event is passed on immediately without first emitting the buffer it is in the process of * assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. It requests {@link Long#MAX_VALUE} * upstream and does not obey downstream requests.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param the collection subclass type to buffer into * @param timespan * the period of time each buffer collects items before it is emitted * @param timeskip * the period of time after which a new buffer will be created * @param unit * the unit of time that applies to the {@code timespan} and {@code timeskip} arguments * @param scheduler * the {@code Scheduler} to use when determining the end and start of a buffer * @param bufferSupplier * a factory function that returns an instance of the collection subclass to be used and returned * as the buffer * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit}, {@code scheduler} or {@code bufferSupplier} is {@code null} * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public final <@NonNull U extends Collection> Flowable buffer(long timespan, long timeskip, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, @NonNull Supplier bufferSupplier) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); Objects.requireNonNull(bufferSupplier, "bufferSupplier is null"); return RxJavaPlugins.onAssembly(new FlowableBufferTimed<>(this, timespan, timeskip, unit, scheduler, bufferSupplier, Integer.MAX_VALUE, false)); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping buffers, each of a fixed duration specified by the * {@code timespan} argument. When the current {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer * and propagates the notification from the current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} * notification the event is passed on immediately without first emitting the buffer it is in the process of * assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. It requests {@link Long#MAX_VALUE} * upstream and does not obey downstream requests.
*
Scheduler:
*
This version of {@code buffer} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timespan * the period of time each buffer collects items before it is emitted and replaced with a new * buffer * @param unit * the unit of time that applies to the {@code timespan} argument * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable> buffer(long timespan, @NonNull TimeUnit unit) { return buffer(timespan, unit, Schedulers.computation(), Integer.MAX_VALUE); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping buffers, each of a fixed duration specified by the * {@code timespan} argument or a maximum size specified by the {@code count} argument (whichever is reached * first). When the current {@code Flowable} completes, the resulting {@code Flowable} emits the current buffer and propagates the * notification from the current {@code Flowable}. Note that if the current {@code Flowable} issues an {@code onError} notification the event * is passed on immediately without first emitting the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. It requests {@link Long#MAX_VALUE} * upstream and does not obey downstream requests.
*
Scheduler:
*
This version of {@code buffer} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timespan * the period of time each buffer collects items before it is emitted and replaced with a new * buffer * @param unit * the unit of time which applies to the {@code timespan} argument * @param count * the maximum size of each buffer before it is emitted * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable> buffer(long timespan, @NonNull TimeUnit unit, int count) { return buffer(timespan, unit, Schedulers.computation(), count); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping buffers, each of a fixed duration specified by the * {@code timespan} argument as measured on the specified {@code scheduler}, or a maximum size specified by * the {@code count} argument (whichever is reached first). When the current {@code Flowable} completes, the resulting * {@code Flowable} emits the current buffer and propagates the notification from the current {@code Flowable}. Note that if the * current {@code Flowable} issues an {@code onError} notification the event is passed on immediately without first emitting the * buffer it is in the process of assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. It requests {@link Long#MAX_VALUE} * upstream and does not obey downstream requests.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each buffer collects items before it is emitted and replaced with a new * buffer * @param unit * the unit of time which applies to the {@code timespan} argument * @param scheduler * the {@code Scheduler} to use when determining the end and start of a buffer * @param count * the maximum size of each buffer before it is emitted * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable> buffer(long timespan, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, int count) { return buffer(timespan, unit, scheduler, count, ArrayListSupplier.asSupplier(), false); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping buffers, each of a fixed duration specified by the * {@code timespan} argument as measured on the specified {@code scheduler}, or a maximum size specified by * the {@code count} argument (whichever is reached first). When the current {@code Flowable} completes, the resulting * {@code Flowable} emits the current buffer and propagates the notification from the current {@code Flowable}. Note that if the * current {@code Flowable} issues an {@code onError} notification the event is passed on immediately without first emitting the * buffer it is in the process of assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. It requests {@link Long#MAX_VALUE} * upstream and does not obey downstream requests.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param the collection subclass type to buffer into * @param timespan * the period of time each buffer collects items before it is emitted and replaced with a new * buffer * @param unit * the unit of time which applies to the {@code timespan} argument * @param scheduler * the {@code Scheduler} to use when determining the end and start of a buffer * @param count * the maximum size of each buffer before it is emitted * @param bufferSupplier * a factory function that returns an instance of the collection subclass to be used and returned * as the buffer * @param restartTimerOnMaxSize if {@code true}, the time window is restarted when the max capacity of the current buffer * is reached * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit}, {@code scheduler} or {@code bufferSupplier} is {@code null} * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final <@NonNull U extends Collection> Flowable buffer( long timespan, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, int count, @NonNull Supplier bufferSupplier, boolean restartTimerOnMaxSize) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); Objects.requireNonNull(bufferSupplier, "bufferSupplier is null"); ObjectHelper.verifyPositive(count, "count"); return RxJavaPlugins.onAssembly(new FlowableBufferTimed<>(this, timespan, timespan, unit, scheduler, bufferSupplier, count, restartTimerOnMaxSize)); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping buffers, each of a fixed duration specified by the * {@code timespan} argument and on the specified {@code scheduler}. When the current {@code Flowable} completes, the * resulting {@code Flowable} emits the current buffer and propagates the notification from the current {@code Flowable}. Note that * if the current {@code Flowable} issues an {@code onError} notification the event is passed on immediately without first emitting * the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time. It requests {@link Long#MAX_VALUE} * upstream and does not obey downstream requests.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each buffer collects items before it is emitted and replaced with a new * buffer * @param unit * the unit of time which applies to the {@code timespan} argument * @param scheduler * the {@code Scheduler} to use when determining the end and start of a buffer * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable> buffer(long timespan, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return buffer(timespan, unit, scheduler, Integer.MAX_VALUE, ArrayListSupplier.asSupplier(), false); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits buffers that it creates when the specified {@code openingIndicator} {@link Publisher} emits an * item, and closes when the {@code Publisher} returned from {@code closingIndicator} emits an item. If any of the current * {@code PFlowable}, {@code openingIndicator} or {@code closingIndicator} issues an {@code onError} notification the event is passed * on immediately without first emitting the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it is instead controlled by the given {@code Publisher}s and * buffers data. It requests {@link Long#MAX_VALUE} upstream and does not obey downstream requests.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the buffer-opening {@code Publisher} * @param the element type of the individual buffer-closing {@code Publisher}s * @param openingIndicator * the {@code Publisher} that, when it emits an item, causes a new buffer to be created * @param closingIndicator * the {@link Function} that is used to produce a {@code Publisher} for every buffer created. When this * {@code Publisher} emits an item, the associated buffer is emitted. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code openingIndicator} or {@code closingIndicator} is {@code null} * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull TOpening, @NonNull TClosing> Flowable> buffer( @NonNull Publisher openingIndicator, @NonNull Function> closingIndicator) { return buffer(openingIndicator, closingIndicator, ArrayListSupplier.asSupplier()); } /** * Returns a {@code Flowable} that emits buffers of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits buffers that it creates when the specified {@code openingIndicator} {@link Publisher} emits an * item, and closes when the {@code Publisher} returned from {@code closingIndicator} emits an item. If any of the current * {@code Flowable}, {@code openingIndicator} or {@code closingIndicator} issues an {@code onError} notification the event is passed * on immediately without first emitting the buffer it is in the process of assembling. *

* *

*
Backpressure:
*
This operator does not support backpressure as it is instead controlled by the given {@code Publisher}s and * buffers data. It requests {@link Long#MAX_VALUE} upstream and does not obey downstream requests.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param the collection subclass type to buffer into * @param the element type of the buffer-opening {@code Publisher} * @param the element type of the individual buffer-closing {@code Publisher}s * @param openingIndicator * the {@code Publisher} that, when it emits an item, causes a new buffer to be created * @param closingIndicator * the {@link Function} that is used to produce a {@code Publisher} for every buffer created. When this * {@code Publisher} emits an item, the associated buffer is emitted. * @param bufferSupplier * a factory function that returns an instance of the collection subclass to be used and returned * as the buffer * @return the new {@code Flowable} instance * @throws NullPointerException if {@code openingIndicator}, {@code closingIndicator} or {@code bufferSupplier} is {@code null} * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull TOpening, @NonNull TClosing, @NonNull U extends Collection> Flowable buffer( @NonNull Publisher openingIndicator, @NonNull Function> closingIndicator, @NonNull Supplier bufferSupplier) { Objects.requireNonNull(openingIndicator, "openingIndicator is null"); Objects.requireNonNull(closingIndicator, "closingIndicator is null"); Objects.requireNonNull(bufferSupplier, "bufferSupplier is null"); return RxJavaPlugins.onAssembly(new FlowableBufferBoundary(this, openingIndicator, closingIndicator, bufferSupplier)); } /** * Returns a {@code Flowable} that emits non-overlapping buffered items from the current {@code Flowable} each time the * specified boundary {@link Publisher} emits an item. *

* *

* Completion of either the source or the boundary {@code Publisher} causes the returned {@code Publisher} to emit the * latest buffer and complete. If either the current {@code Flowable} or the boundary {@code Publisher} issues an {@code onError} notification * the event is passed on immediately without first emitting the buffer it is in the process of assembling. *

*
Backpressure:
*
This operator does not support backpressure as it is instead controlled by the {@code Publisher} * {@code boundary} and buffers data. It requests {@link Long#MAX_VALUE} upstream and does not obey * downstream requests.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the boundary value type (ignored) * @param boundaryIndicator * the boundary {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code boundaryIndicator} is {@code null} * @see #buffer(Publisher, int) * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull B> Flowable> buffer(@NonNull Publisher boundaryIndicator) { return buffer(boundaryIndicator, ArrayListSupplier.asSupplier()); } /** * Returns a {@code Flowable} that emits non-overlapping buffered items from the current {@code Flowable} each time the * specified boundary {@link Publisher} emits an item. *

* *

* Completion of either the source or the boundary {@code Publisher} causes the returned {@code Publisher} to emit the * latest buffer and complete. If either the current {@code Flowable} or the boundary {@code Publisher} issues an {@code onError} notification * the event is passed on immediately without first emitting the buffer it is in the process of assembling. *

*
Backpressure:
*
This operator does not support backpressure as it is instead controlled by the {@code Publisher} * {@code boundary} and buffers data. It requests {@link Long#MAX_VALUE} upstream and does not obey * downstream requests.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the boundary value type (ignored) * @param boundaryIndicator * the boundary {@code Publisher} * @param initialCapacity * the initial capacity of each buffer chunk * @return the new {@code Flowable} instance * @throws NullPointerException if {@code boundaryIndicator} is {@code null} * @throws IllegalArgumentException if {@code initialCapacity} is non-positive * @see ReactiveX operators documentation: Buffer * @see #buffer(Publisher) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull B> Flowable> buffer(@NonNull Publisher boundaryIndicator, int initialCapacity) { ObjectHelper.verifyPositive(initialCapacity, "initialCapacity"); return buffer(boundaryIndicator, Functions.createArrayList(initialCapacity)); } /** * Returns a {@code Flowable} that emits non-overlapping buffered items from the current {@code Flowable} each time the * specified boundary {@link Publisher} emits an item. *

* *

* Completion of either the source or the boundary {@code Publisher} causes the returned {@code Publisher} to emit the * latest buffer and complete. If either the current {@code Flowable} or the boundary {@code Publisher} issues an {@code onError} notification * the event is passed on immediately without first emitting the buffer it is in the process of assembling. *

*
Backpressure:
*
This operator does not support backpressure as it is instead controlled by the {@code Publisher} * {@code boundary} and buffers data. It requests {@link Long#MAX_VALUE} upstream and does not obey * downstream requests.
*
Scheduler:
*
This version of {@code buffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param the collection subclass type to buffer into * @param * the boundary value type (ignored) * @param boundaryIndicator * the boundary {@code Publisher} * @param bufferSupplier * a factory function that returns an instance of the collection subclass to be used and returned * as the buffer * @return the new {@code Flowable} instance * @throws NullPointerException if {@code boundaryIndicator} or {@code bufferSupplier} is {@code null} * @see #buffer(Publisher, int) * @see ReactiveX operators documentation: Buffer */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull B, @NonNull U extends Collection> Flowable buffer(@NonNull Publisher boundaryIndicator, @NonNull Supplier bufferSupplier) { Objects.requireNonNull(boundaryIndicator, "boundaryIndicator is null"); Objects.requireNonNull(bufferSupplier, "bufferSupplier is null"); return RxJavaPlugins.onAssembly(new FlowableBufferExactBoundary<>(this, boundaryIndicator, bufferSupplier)); } /** * Returns a {@code Flowable} that subscribes to this {@link Publisher} lazily, caches all of its events * and replays them, in the same order as received, to all the downstream subscribers. *

* *

* This is useful when you want a {@code Publisher} to cache responses and you can't control the * subscribe/cancel behavior of all the {@link Subscriber}s. *

* The operator subscribes only when the first downstream subscriber subscribes and maintains * a single subscription towards this {@code Publisher}. In contrast, the operator family of {@link #replay()} * that return a {@link ConnectableFlowable} require an explicit call to {@link ConnectableFlowable#connect()}. *

* Note: You sacrifice the ability to cancel the origin when you use the {@code cache} * operator so be careful not to use this operator on {@code Publisher}s that emit an infinite or very large number * of items that will use up memory. * A possible workaround is to apply {@link #takeUntil(Publisher)} with a predicate or * another source before (and perhaps after) the application of {@code cache()}. *


     * AtomicBoolean shouldStop = new AtomicBoolean();
     *
     * source.takeUntil(v -> shouldStop.get())
     *       .cache()
     *       .takeUntil(v -> shouldStop.get())
     *       .subscribe(...);
     * 
* Since the operator doesn't allow clearing the cached values either, the possible workaround is * to forget all references to it via {@link #onTerminateDetach()} applied along with the previous * workaround: *

     * AtomicBoolean shouldStop = new AtomicBoolean();
     *
     * source.takeUntil(v -> shouldStop.get())
     *       .onTerminateDetach()
     *       .cache()
     *       .takeUntil(v -> shouldStop.get())
     *       .onTerminateDetach()
     *       .subscribe(...);
     * 
*
*
Backpressure:
*
The operator consumes this {@code Publisher} in an unbounded fashion but respects the backpressure * of each downstream {@code Subscriber} individually.
*
Scheduler:
*
{@code cache} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: Replay * @see #takeUntil(Predicate) * @see #takeUntil(Publisher) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable cache() { return cacheWithInitialCapacity(16); } /** * Returns a {@code Flowable} that subscribes to this {@link Publisher} lazily, caches all of its events * and replays them, in the same order as received, to all the downstream subscribers. *

* *

* This is useful when you want a {@code Publisher} to cache responses and you can't control the * subscribe/cancel behavior of all the {@link Subscriber}s. *

* The operator subscribes only when the first downstream subscriber subscribes and maintains * a single subscription towards this {@code Publisher}. In contrast, the operator family of {@link #replay()} * that return a {@link ConnectableFlowable} require an explicit call to {@link ConnectableFlowable#connect()}. *

* Note: You sacrifice the ability to cancel the origin when you use the {@code cache} * operator so be careful not to use this operator on {@code Publisher}s that emit an infinite or very large number * of items that will use up memory. * A possible workaround is to apply {@link #takeUntil(Publisher)} with a predicate or * another source before (and perhaps after) the application of {@code cacheWithInitialCapacity()}. *


     * AtomicBoolean shouldStop = new AtomicBoolean();
     *
     * source.takeUntil(v -> shouldStop.get())
     *       .cache()
     *       .takeUntil(v -> shouldStop.get())
     *       .subscribe(...);
     * 
* Since the operator doesn't allow clearing the cached values either, the possible workaround is * to forget all references to it via {@link #onTerminateDetach()} applied along with the previous * workaround: *

     * AtomicBoolean shouldStop = new AtomicBoolean();
     *
     * source.takeUntil(v -> shouldStop.get())
     *       .onTerminateDetach()
     *       .cache()
     *       .takeUntil(v -> shouldStop.get())
     *       .onTerminateDetach()
     *       .subscribe(...);
     * 
*
*
Backpressure:
*
The operator consumes this {@code Publisher} in an unbounded fashion but respects the backpressure * of each downstream {@code Subscriber} individually.
*
Scheduler:
*
{@code cacheWithInitialCapacity} does not operate by default on a particular {@link Scheduler}.
*
*

* Note: The capacity hint is not an upper bound on cache size. For that, consider * {@link #replay(int)} in combination with {@link ConnectableFlowable#autoConnect()} or similar. * * @param initialCapacity hint for number of items to cache (for optimizing underlying data structure) * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code initialCapacity} is non-positive * @see ReactiveX operators documentation: Replay * @see #takeUntil(Predicate) * @see #takeUntil(Publisher) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable cacheWithInitialCapacity(int initialCapacity) { ObjectHelper.verifyPositive(initialCapacity, "initialCapacity"); return RxJavaPlugins.onAssembly(new FlowableCache<>(this, initialCapacity)); } /** * Returns a {@code Flowable} that emits the items emitted by the current {@code Flowable}, converted to the specified * type. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code cast} does not operate by default on a particular {@link Scheduler}.
*
* * @param the output value type cast to * @param clazz * the target class type that {@code cast} will cast the items emitted by the current {@code Flowable} * into before emitting them from the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code clazz} is {@code null} * @see ReactiveX operators documentation: Map */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable cast(@NonNull Class clazz) { Objects.requireNonNull(clazz, "clazz is null"); return map(Functions.castFunction(clazz)); } /** * Collects items emitted by the finite source {@link Publisher} into a single mutable data structure and returns * a {@link Single} that emits this structure. *

* *

* This is a simplified version of {@code reduce} that does not need to return the state on each pass. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulator object to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
This operator does not support backpressure because by intent it will receive all values and reduce * them to a single {@code onNext}.
*
Scheduler:
*
{@code collect} does not operate by default on a particular {@link Scheduler}.
*
* * @param the accumulator and output type * @param initialItemSupplier * the mutable data structure that will collect the items * @param collector * a function that accepts the {@code state} and an emitted item, and modifies {@code state} * accordingly * @return the new {@code Single} instance * @throws NullPointerException if {@code initialItemSupplier} or {@code collector} is {@code null} * @see ReactiveX operators documentation: Reduce * @see #collect(Collector) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Single collect(@NonNull Supplier initialItemSupplier, @NonNull BiConsumer collector) { Objects.requireNonNull(initialItemSupplier, "initialItemSupplier is null"); Objects.requireNonNull(collector, "collector is null"); return RxJavaPlugins.onAssembly(new FlowableCollectSingle<>(this, initialItemSupplier, collector)); } /** * Collects items emitted by the finite source {@link Publisher} into a single mutable data structure and returns * a {@link Single} that emits this structure. *

* *

* This is a simplified version of {@code reduce} that does not need to return the state on each pass. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulator object to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
This operator does not support backpressure because by intent it will receive all values and reduce * them to a single {@code onNext}.
*
Scheduler:
*
{@code collectInto} does not operate by default on a particular {@link Scheduler}.
*
* * @param the accumulator and output type * @param initialItem * the mutable data structure that will collect the items * @param collector * a function that accepts the {@code state} and an emitted item, and modifies {@code state} * accordingly * @return the new {@code Single} instance * @throws NullPointerException if {@code initialItem} or {@code collector} is {@code null} * @see ReactiveX operators documentation: Reduce */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Single collectInto(U initialItem, @NonNull BiConsumer collector) { Objects.requireNonNull(initialItem, "initialItem is null"); return collect(Functions.justSupplier(initialItem), collector); } /** * Transform the current {@code Flowable} by applying a particular {@link FlowableTransformer} function to it. *

* This method operates on the {@code Flowable} itself whereas {@link #lift} operates on the {@code Flowable}'s * {@link Subscriber}s. *

* If the operator you are creating is designed to act on the individual items emitted by a current * {@code Flowable}, use {@link #lift}. If your operator is designed to transform the current {@code Flowable} as a whole * (for instance, by applying a particular set of existing RxJava operators to it) use {@code compose}. *

*
Backpressure:
*
The operator itself doesn't interfere with the backpressure behavior which only depends * on what kind of {@link Publisher} the {@code FlowableTransformer} returns.
*
Scheduler:
*
{@code compose} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the output {@code Publisher} * @param composer implements the function that transforms the current {@code Flowable} * @return the new composed {@code Flowable} instance * @throws NullPointerException if {@code composer} is {@code null} * @see RxJava wiki: Implementing Your Own Operators */ @SuppressWarnings("unchecked") @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable compose(@NonNull FlowableTransformer composer) { return fromPublisher(((FlowableTransformer) Objects.requireNonNull(composer, "composer is null")).apply(this)); } /** * Returns a new {@code Flowable} that emits items resulting from applying a function that you supply to each item * emitted by the current {@code Flowable}, where that function returns a {@link Publisher}, and then emitting the items * that result from concatenating those returned {@code Publisher}s. *

* *

* Note that there is no guarantee where the given {@code mapper} function will be executed; it could be on the subscribing thread, * on the upstream thread signaling the new item to be mapped or on the thread where the inner source terminates. To ensure * the {@code mapper} function is confined to a known thread, use the {@link #concatMap(Function, int, Scheduler)} overload. *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the inner {@code Publisher}s are * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}. If any of the inner {@code Publisher}s doesn't honor * backpressure, that may throw an {@link IllegalStateException} when that * {@code Publisher} completes.
*
Scheduler:
*
{@code concatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the type of the inner {@code Publisher} sources and thus the output type * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMap(@NonNull Function> mapper) { return concatMap(mapper, 2); } /** * Returns a new {@code Flowable} that emits items resulting from applying a function that you supply to each item * emitted by the current {@code Flowable}, where that function returns a {@link Publisher}, and then emitting the items * that result from concatenating those returned {@code Publisher}s. *

* *

* Note that there is no guarantee where the given {@code mapper} function will be executed; it could be on the subscribing thread, * on the upstream thread signaling the new item to be mapped or on the thread where the inner source terminates. To ensure * the {@code mapper} function is confined to a known thread, use the {@link #concatMap(Function, int, Scheduler)} overload. *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the inner {@code Publisher}s are * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}. If any of the inner {@code Publisher}s doesn't honor * backpressure, that may throw an {@link IllegalStateException} when that * {@code Publisher} completes.
*
Scheduler:
*
{@code concatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the type of the inner {@code Publisher} sources and thus the output type * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @param prefetch * the number of elements to prefetch from the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see ReactiveX operators documentation: FlatMap * @see #concatMap(Function, int, Scheduler) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable concatMap(@NonNull Function> mapper, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); if (this instanceof ScalarSupplier) { @SuppressWarnings("unchecked") T v = ((ScalarSupplier)this).get(); if (v == null) { return empty(); } return FlowableScalarXMap.scalarXMap(v, mapper); } return RxJavaPlugins.onAssembly(new FlowableConcatMap<>(this, mapper, prefetch, ErrorMode.IMMEDIATE)); } /** * Returns a new {@code Flowable} that emits items resulting from applying a function (on a designated scheduler) * that you supply to each item emitted by the current {@code Flowable}, where that function returns a {@link Publisher}, and then emitting the items * that result from concatenating those returned {@code Publisher}s. *

* *

* The difference between {@link #concatMap(Function, int)} and this operator is that this operator guarantees the {@code mapper} * function is executed on the specified scheduler. *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the inner {@code Publisher}s are * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}. If any of the inner {@code Publisher}s doesn't honor * backpressure, that may throw an {@link IllegalStateException} when that * {@code Publisher} completes.
*
Scheduler:
*
{@code concatMap} executes the given {@code mapper} function on the provided {@link Scheduler}.
*
* * @param the type of the inner {@code Publisher} sources and thus the output type * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @param prefetch * the number of elements to prefetch from the current {@code Flowable} * @param scheduler * the scheduler where the {@code mapper} function will be executed * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 3.0.0 * @see #concatMap(Function, int) * @see #concatMapDelayError(Function, boolean, int, Scheduler) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final <@NonNull R> Flowable concatMap(@NonNull Function> mapper, int prefetch, @NonNull Scheduler scheduler) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableConcatMapScheduler<>(this, mapper, prefetch, ErrorMode.IMMEDIATE, scheduler)); } /** * Maps the upstream items into {@link CompletableSource}s and subscribes to them one after the * other completes. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapCompletable} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param mapper the function called with the upstream item and should return * a {@code CompletableSource} to become the next source to * be subscribed to * @return the new {@link Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapCompletableDelayError(Function) * @since 2.2 */ @CheckReturnValue @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) @NonNull public final Completable concatMapCompletable(@NonNull Function mapper) { return concatMapCompletable(mapper, 2); } /** * Maps the upstream items into {@link CompletableSource}s and subscribes to them one after the * other completes. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapCompletable} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param mapper the function called with the upstream item and should return * a {@code CompletableSource} to become the next source to * be subscribed to * @param prefetch The number of upstream items to prefetch so that fresh items are * ready to be mapped when a previous {@code CompletableSource} terminates. * The operator replenishes after half of the prefetch amount has been consumed * and turned into {@code CompletableSource}s. * @return the new {@link Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see #concatMapCompletableDelayError(Function, boolean, int) * @since 2.2 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) public final Completable concatMapCompletable(@NonNull Function mapper, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapCompletable<>(this, mapper, ErrorMode.IMMEDIATE, prefetch)); } /** * Maps the upstream items into {@link CompletableSource}s and subscribes to them one after the * other terminates, delaying all errors till both this {@code Flowable} and all * inner {@code CompletableSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapCompletableDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param mapper the function called with the upstream item and should return * a {@code CompletableSource} to become the next source to * be subscribed to * @return the new {@link Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapCompletable(Function, int) * @since 2.2 */ @CheckReturnValue @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) @NonNull public final Completable concatMapCompletableDelayError(@NonNull Function mapper) { return concatMapCompletableDelayError(mapper, true, 2); } /** * Maps the upstream items into {@link CompletableSource}s and subscribes to them one after the * other terminates, optionally delaying all errors till both this {@code Flowable} and all * inner {@code CompletableSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapCompletableDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param mapper the function called with the upstream item and should return * a {@code CompletableSource} to become the next source to * be subscribed to * @param tillTheEnd If {@code true}, errors from this {@code Flowable} or any of the * inner {@code CompletableSource}s are delayed until all * of them terminate. If {@code false}, an error from this * {@code Flowable} is delayed until the current inner * {@code CompletableSource} terminates and only then is * it emitted to the downstream. * @return the new {@link Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapCompletable(Function) * @since 2.2 */ @CheckReturnValue @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) @NonNull public final Completable concatMapCompletableDelayError(@NonNull Function mapper, boolean tillTheEnd) { return concatMapCompletableDelayError(mapper, tillTheEnd, 2); } /** * Maps the upstream items into {@link CompletableSource}s and subscribes to them one after the * other terminates, optionally delaying all errors till both this {@code Flowable} and all * inner {@code CompletableSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapCompletableDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param mapper the function called with the upstream item and should return * a {@code CompletableSource} to become the next source to * be subscribed to * @param tillTheEnd If {@code true}, errors from this {@code Flowable} or any of the * inner {@code CompletableSource}s are delayed until all * of them terminate. If {@code false}, an error from this * {@code Flowable} is delayed until the current inner * {@code CompletableSource} terminates and only then is * it emitted to the downstream. * @param prefetch The number of upstream items to prefetch so that fresh items are * ready to be mapped when a previous {@code CompletableSource} terminates. * The operator replenishes after half of the prefetch amount has been consumed * and turned into {@code CompletableSource}s. * @return the new {@link Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see #concatMapCompletable(Function, int) * @since 2.2 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) public final Completable concatMapCompletableDelayError(@NonNull Function mapper, boolean tillTheEnd, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapCompletable<>(this, mapper, tillTheEnd ? ErrorMode.END : ErrorMode.BOUNDARY, prefetch)); } /** * Maps each of the items into a {@link Publisher}, subscribes to them one after the other, * one at a time and emits their values in order * while delaying any error from either this or any of the inner {@code Publisher}s * till all of them terminate. *

* Note that there is no guarantee where the given {@code mapper} function will be executed; it could be on the subscribing thread, * on the upstream thread signaling the new item to be mapped or on the thread where the inner source terminates. To ensure * the {@code mapper} function is confined to a known thread, use the {@link #concatMapDelayError(Function, boolean, int, Scheduler)} overload. *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the inner {@code Publisher}s are * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}. If any of the inner {@code Publisher}s doesn't honor * backpressure, that may throw an {@link IllegalStateException} when that * {@code Publisher} completes.
*
Scheduler:
*
{@code concatMapDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the result value type * @param mapper the function that maps the items of this {@code Publisher} into the inner {@code Publisher}s. * @return the new {@code Flowable} instance with the concatenation behavior * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapDelayError(Function, boolean, int, Scheduler) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapDelayError(@NonNull Function> mapper) { return concatMapDelayError(mapper, true, 2); } /** * Maps each of the items into a {@link Publisher}, subscribes to them one after the other, * one at a time and emits their values in order * while delaying any error from either this or any of the inner {@code Publisher}s * till all of them terminate. *

* Note that there is no guarantee where the given {@code mapper} function will be executed; it could be on the subscribing thread, * on the upstream thread signaling the new item to be mapped or on the thread where the inner source terminates. To ensure * the {@code mapper} function is confined to a known thread, use the {@link #concatMapDelayError(Function, boolean, int, Scheduler)} overload. * *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the inner {@code Publisher}s are * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}. If any of the inner {@code Publisher}s doesn't honor * backpressure, that may throw an {@link IllegalStateException} when that * {@code Publisher} completes.
*
Scheduler:
*
{@code concatMapDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the result value type * @param mapper the function that maps the items of this {@code Publisher} into the inner {@code Publisher}s. * @param tillTheEnd * if {@code true}, all errors from the outer and inner {@code Publisher} sources are delayed until the end, * if {@code false}, an error from the main source is signaled when the current inner {@code Publisher} source terminates * @param prefetch * the number of elements to prefetch from the current {@code Flowable} * @return the new {@code Flowable} instance with the concatenation behavior * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see #concatMapDelayError(Function, boolean, int, Scheduler) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable concatMapDelayError(@NonNull Function> mapper, boolean tillTheEnd, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); if (this instanceof ScalarSupplier) { @SuppressWarnings("unchecked") T v = ((ScalarSupplier)this).get(); if (v == null) { return empty(); } return FlowableScalarXMap.scalarXMap(v, mapper); } return RxJavaPlugins.onAssembly(new FlowableConcatMap<>(this, mapper, prefetch, tillTheEnd ? ErrorMode.END : ErrorMode.BOUNDARY)); } /** * Maps each of the upstream items into a {@link Publisher}, subscribes to them one after the other, * one at a time and emits their values in order * while executing the mapper function on the designated scheduler, delaying any error from either this or any of the * inner {@code Publisher}s till all of them terminate. *

* The difference between {@link #concatMapDelayError(Function, boolean, int)} and this operator is that this operator guarantees the {@code mapper} * function is executed on the specified scheduler. * *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the inner {@code Publisher}s are * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}. If any of the inner {@code Publisher}s doesn't honor * backpressure, that may throw an {@link IllegalStateException} when that * {@code Publisher} completes.
*
Scheduler:
*
{@code concatMapDelayError} executes the given {@code mapper} function on the provided {@link Scheduler}.
*
* * @param the result value type * @param mapper the function that maps the items of this {@code Publisher} into the inner {@code Publisher}s. * @param tillTheEnd * if {@code true}, all errors from the outer and inner {@code Publisher} sources are delayed until the end, * if {@code false}, an error from the main source is signaled when the current inner {@code Publisher} source terminates * @param prefetch * the number of elements to prefetch from the current {@code Flowable} * @param scheduler * the scheduler where the {@code mapper} function will be executed * @return the new {@code Flowable} instance with the concatenation behavior * @throws NullPointerException if {@code mapper} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see #concatMapDelayError(Function, boolean, int) * @since 3.0.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final <@NonNull R> Flowable concatMapDelayError(@NonNull Function> mapper, boolean tillTheEnd, int prefetch, @NonNull Scheduler scheduler) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableConcatMapScheduler<>(this, mapper, prefetch, tillTheEnd ? ErrorMode.END : ErrorMode.BOUNDARY, scheduler)); } /** * Maps a sequence of values into {@link Publisher}s and concatenates these {@code Publisher}s eagerly into a single * {@code Publisher}. *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * inner {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them in * order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream, however, due to the eagerness requirement, sources * are subscribed to in unbounded mode and their values are queued up in an unbounded buffer.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param mapper the function that maps a sequence of values into a sequence of {@code Publisher}s that will be * eagerly concatenated * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code mapper} is {@code null} * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapEager(@NonNull Function> mapper) { return concatMapEager(mapper, bufferSize(), bufferSize()); } /** * Maps a sequence of values into {@link Publisher}s and concatenates these {@code Publisher}s eagerly into a single * {@code Publisher}. *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * inner {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them in * order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream, however, due to the eagerness requirement, sources * are subscribed to in unbounded mode and their values are queued up in an unbounded buffer.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param mapper the function that maps a sequence of values into a sequence of {@code Publisher}s that will be * eagerly concatenated * @param maxConcurrency the maximum number of concurrent subscribed {@code Publisher}s * @param prefetch hints about the number of expected values from each inner {@code Publisher}, must be positive * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code prefetch} is non-positive * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable concatMapEager(@NonNull Function> mapper, int maxConcurrency, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapEager<>(this, mapper, maxConcurrency, prefetch, ErrorMode.IMMEDIATE)); } /** * Maps a sequence of values into {@link Publisher}s and concatenates these {@code Publisher}s eagerly into a single * {@code Publisher}. *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * inner {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them in * order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream, however, due to the eagerness requirement, sources * are subscribed to in unbounded mode and their values are queued up in an unbounded buffer.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param mapper the function that maps a sequence of values into a sequence of {@code Publisher}s that will be * eagerly concatenated * @param tillTheEnd * if {@code true}, all errors from the outer and inner {@code Publisher} sources are delayed until the end, * if {@code false}, an error from the main source is signaled when the current inner {@code Publisher} source terminates * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code mapper} is {@code null} * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapEagerDelayError(@NonNull Function> mapper, boolean tillTheEnd) { return concatMapEagerDelayError(mapper, tillTheEnd, bufferSize(), bufferSize()); } /** * Maps a sequence of values into {@link Publisher}s and concatenates these {@code Publisher}s eagerly into a single * {@code Flowable} sequence. *

* Eager concatenation means that once a subscriber subscribes, this operator subscribes to all of the * inner {@code Publisher}s. The operator buffers the values emitted by these {@code Publisher}s and then drains them in * order, each one after the previous one completes. *

*
Backpressure:
*
Backpressure is honored towards the downstream, however, due to the eagerness requirement, sources * are subscribed to in unbounded mode and their values are queued up in an unbounded buffer.
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* @param the value type * @param mapper the function that maps a sequence of values into a sequence of {@code Publisher}s that will be * eagerly concatenated * @param tillTheEnd * if {@code true}, exceptions from the current {@code Flowable} and all the inner {@code Publisher}s are delayed until * all of them terminate, if {@code false}, exception from the current {@code Flowable} is delayed until the * currently running {@code Publisher} terminates * @param maxConcurrency the maximum number of concurrent subscribed {@code Publisher}s * @param prefetch * the number of elements to prefetch from each source {@code Publisher} * @return the new {@code Flowable} instance with the specified concatenation behavior * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code prefetch} is non-positive * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable concatMapEagerDelayError(@NonNull Function> mapper, boolean tillTheEnd, int maxConcurrency, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapEager<>(this, mapper, maxConcurrency, prefetch, tillTheEnd ? ErrorMode.END : ErrorMode.BOUNDARY)); } /** * Returns a {@code Flowable} that concatenate each item emitted by the current {@code Flowable} with the values in an * {@link Iterable} corresponding to that item that is generated by a selector. * *
*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable}s is * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of item emitted by the resulting {@code Flowable} * @param mapper * a function that returns an {@code Iterable} sequence of values for when given an item emitted by the * current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U> Flowable concatMapIterable(@NonNull Function> mapper) { return concatMapIterable(mapper, 2); } /** * Returns a {@code Flowable} that concatenate each item emitted by the current {@code Flowable} with the values in an * {@link Iterable} corresponding to that item that is generated by a selector. * *
*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable} is * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of item emitted by the resulting {@code Flowable} * @param mapper * a function that returns an {@code Iterable} sequence of values for when given an item emitted by the * current {@code Flowable} * @param prefetch * the number of elements to prefetch from the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable concatMapIterable(@NonNull Function> mapper, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableFlattenIterable<>(this, mapper, prefetch)); } /** * Maps the upstream items into {@link MaybeSource}s and subscribes to them one after the * other succeeds or completes, emits their success value if available or terminates immediately if * either this {@code Flowable} or the current inner {@code MaybeSource} fail. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapMaybe} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code MaybeSource}s * @param mapper the function called with the upstream item and should return * a {@code MaybeSource} to become the next source to * be subscribed to * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapMaybeDelayError(Function) * @see #concatMapMaybe(Function, int) * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapMaybe(@NonNull Function> mapper) { return concatMapMaybe(mapper, 2); } /** * Maps the upstream items into {@link MaybeSource}s and subscribes to them one after the * other succeeds or completes, emits their success value if available or terminates immediately if * either this {@code Flowable} or the current inner {@code MaybeSource} fail. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapMaybe} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code MaybeSource}s * @param mapper the function called with the upstream item and should return * a {@code MaybeSource} to become the next source to * be subscribed to * @param prefetch The number of upstream items to prefetch so that fresh items are * ready to be mapped when a previous {@code MaybeSource} terminates. * The operator replenishes after half of the prefetch amount has been consumed * and turned into {@code MaybeSource}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see #concatMapMaybe(Function) * @see #concatMapMaybeDelayError(Function, boolean, int) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable concatMapMaybe(@NonNull Function> mapper, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapMaybe<>(this, mapper, ErrorMode.IMMEDIATE, prefetch)); } /** * Maps the upstream items into {@link MaybeSource}s and subscribes to them one after the * other terminates, emits their success value if available and delaying all errors * till both this {@code Flowable} and all inner {@code MaybeSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapMaybeDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code MaybeSource}s * @param mapper the function called with the upstream item and should return * a {@code MaybeSource} to become the next source to * be subscribed to * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapMaybe(Function) * @see #concatMapMaybeDelayError(Function, boolean) * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapMaybeDelayError(@NonNull Function> mapper) { return concatMapMaybeDelayError(mapper, true, 2); } /** * Maps the upstream items into {@link MaybeSource}s and subscribes to them one after the * other terminates, emits their success value if available and optionally delaying all errors * till both this {@code Flowable} and all inner {@code MaybeSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapMaybeDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code MaybeSource}s * @param mapper the function called with the upstream item and should return * a {@code MaybeSource} to become the next source to * be subscribed to * @param tillTheEnd If {@code true}, errors from this {@code Flowable} or any of the * inner {@code MaybeSource}s are delayed until all * of them terminate. If {@code false}, an error from this * {@code Flowable} is delayed until the current inner * {@code MaybeSource} terminates and only then is * it emitted to the downstream. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapMaybe(Function, int) * @see #concatMapMaybeDelayError(Function, boolean, int) * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapMaybeDelayError(@NonNull Function> mapper, boolean tillTheEnd) { return concatMapMaybeDelayError(mapper, tillTheEnd, 2); } /** * Maps the upstream items into {@link MaybeSource}s and subscribes to them one after the * other terminates, emits their success value if available and optionally delaying all errors * till both this {@code Flowable} and all inner {@code MaybeSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapMaybeDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code MaybeSource}s * @param mapper the function called with the upstream item and should return * a {@code MaybeSource} to become the next source to * be subscribed to * @param tillTheEnd If {@code true}, errors from this {@code Flowable} or any of the * inner {@code MaybeSource}s are delayed until all * of them terminate. If {@code false}, an error from this * {@code Flowable} is delayed until the current inner * {@code MaybeSource} terminates and only then is * it emitted to the downstream. * @param prefetch The number of upstream items to prefetch so that fresh items are * ready to be mapped when a previous {@code MaybeSource} terminates. * The operator replenishes after half of the prefetch amount has been consumed * and turned into {@code MaybeSource}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see #concatMapMaybe(Function, int) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable concatMapMaybeDelayError(@NonNull Function> mapper, boolean tillTheEnd, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapMaybe<>(this, mapper, tillTheEnd ? ErrorMode.END : ErrorMode.BOUNDARY, prefetch)); } /** * Maps the upstream items into {@link SingleSource}s and subscribes to them one after the * other succeeds, emits their success values or terminates immediately if * either this {@code Flowable} or the current inner {@code SingleSource} fail. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapSingle} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code SingleSource}s * @param mapper the function called with the upstream item and should return * a {@code SingleSource} to become the next source to * be subscribed to * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapSingleDelayError(Function) * @see #concatMapSingle(Function, int) * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapSingle(@NonNull Function> mapper) { return concatMapSingle(mapper, 2); } /** * Maps the upstream items into {@link SingleSource}s and subscribes to them one after the * other succeeds, emits their success values or terminates immediately if * either this {@code Flowable} or the current inner {@code SingleSource} fail. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapSingle} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code SingleSource}s * @param mapper the function called with the upstream item and should return * a {@code SingleSource} to become the next source to * be subscribed to * @param prefetch The number of upstream items to prefetch so that fresh items are * ready to be mapped when a previous {@code SingleSource} terminates. * The operator replenishes after half of the prefetch amount has been consumed * and turned into {@code SingleSource}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see #concatMapSingle(Function) * @see #concatMapSingleDelayError(Function, boolean, int) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable concatMapSingle(@NonNull Function> mapper, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapSingle<>(this, mapper, ErrorMode.IMMEDIATE, prefetch)); } /** * Maps the upstream items into {@link SingleSource}s and subscribes to them one after the * other succeeds or fails, emits their success values and delays all errors * till both this {@code Flowable} and all inner {@code SingleSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapSingleDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code SingleSource}s * @param mapper the function called with the upstream item and should return * a {@code SingleSource} to become the next source to * be subscribed to * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapSingle(Function) * @see #concatMapSingleDelayError(Function, boolean) * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapSingleDelayError(@NonNull Function> mapper) { return concatMapSingleDelayError(mapper, true, 2); } /** * Maps the upstream items into {@link SingleSource}s and subscribes to them one after the * other succeeds or fails, emits their success values and optionally delays all errors * till both this {@code Flowable} and all inner {@code SingleSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapSingleDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code SingleSource}s * @param mapper the function called with the upstream item and should return * a {@code SingleSource} to become the next source to * be subscribed to * @param tillTheEnd If {@code true}, errors from this {@code Flowable} or any of the * inner {@code SingleSource}s are delayed until all * of them terminate. If {@code false}, an error from this * {@code Flowable} is delayed until the current inner * {@code SingleSource} terminates and only then is * it emitted to the downstream. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #concatMapSingle(Function, int) * @see #concatMapSingleDelayError(Function, boolean, int) * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapSingleDelayError(@NonNull Function> mapper, boolean tillTheEnd) { return concatMapSingleDelayError(mapper, tillTheEnd, 2); } /** * Maps the upstream items into {@link SingleSource}s and subscribes to them one after the * other succeeds or fails, emits their success values and optionally delays errors * till both this {@code Flowable} and all inner {@code SingleSource}s terminate. *

* *

*
Backpressure:
*
The operator expects the upstream to support backpressure and honors * the backpressure from downstream. If this {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code concatMapSingleDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the result type of the inner {@code SingleSource}s * @param mapper the function called with the upstream item and should return * a {@code SingleSource} to become the next source to * be subscribed to * @param tillTheEnd If {@code true}, errors from this {@code Flowable} or any of the * inner {@code SingleSource}s are delayed until all * of them terminate. If {@code false}, an error from this * {@code Flowable} is delayed until the current inner * {@code SingleSource} terminates and only then is * it emitted to the downstream. * @param prefetch The number of upstream items to prefetch so that fresh items are * ready to be mapped when a previous {@code SingleSource} terminates. * The operator replenishes after half of the prefetch amount has been consumed * and turned into {@code SingleSource}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see #concatMapSingle(Function, int) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable concatMapSingleDelayError(@NonNull Function> mapper, boolean tillTheEnd, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableConcatMapSingle<>(this, mapper, tillTheEnd ? ErrorMode.END : ErrorMode.BOUNDARY, prefetch)); } /** * Returns a {@code Flowable} that emits the items emitted from the current {@code Flowable}, then the next, one after * the other, without interleaving them. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the {@code other} {@link Publisher}s * are expected to honor backpressure as well. If any of then violates this rule, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
{@code concatWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param other * a {@code Publisher} to be concatenated after the current * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @see ReactiveX operators documentation: Concat */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable concatWith(@NonNull Publisher other) { Objects.requireNonNull(other, "other is null"); return concat(this, other); } /** * Returns a {@code Flowable} that emits the items from this {@code Flowable} followed by the success item or error event * of the other {@link SingleSource}. *

* *

*
Backpressure:
*
The operator supports backpressure and makes sure the success item of the other {@code SingleSource} * is only emitted when there is a demand for it.
*
Scheduler:
*
{@code concatWith} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.10 - experimental * @param other the {@code SingleSource} whose signal should be emitted after this {@code Flowable} completes normally. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable concatWith(@NonNull SingleSource other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableConcatWithSingle<>(this, other)); } /** * Returns a {@code Flowable} that emits the items from this {@code Flowable} followed by the success item or terminal events * of the other {@link MaybeSource}. *

* *

*
Backpressure:
*
The operator supports backpressure and makes sure the success item of the other {@code MaybeSource} * is only emitted when there is a demand for it.
*
Scheduler:
*
{@code concatWith} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.10 - experimental * @param other the {@code MaybeSource} whose signal should be emitted after this {@code Flowable} completes normally. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable concatWith(@NonNull MaybeSource other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableConcatWithMaybe<>(this, other)); } /** * Returns a {@code Flowable} that emits items from this {@code Flowable} and when it completes normally, the * other {@link CompletableSource} is subscribed to and the returned {@code Flowable} emits its terminal events. *

* *

*
Backpressure:
*
The operator does not interfere with backpressure between the current {@code Flowable} and the * downstream consumer (i.e., acts as pass-through). When the operator switches to the * {@link Completable}, backpressure is no longer present because {@code Completable} doesn't * have items to apply backpressure to.
*
Scheduler:
*
{@code concatWith} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.10 - experimental * @param other the {@code CompletableSource} to subscribe to once the current {@code Flowable} completes normally * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable concatWith(@NonNull CompletableSource other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableConcatWithCompletable<>(this, other)); } /** * Returns a {@link Single} that emits a {@link Boolean} that indicates whether the current {@code Flowable} emitted a * specified item. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code contains} does not operate by default on a particular {@link Scheduler}.
*
* * @param item * the item to search for in the emissions from the current {@code Flowable} * @return the new {@code Single} instance * @throws NullPointerException if {@code item} is {@code null} * @see ReactiveX operators documentation: Contains */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Single contains(@NonNull Object item) { Objects.requireNonNull(item, "item is null"); return any(Functions.equalsWith(item)); } /** * Returns a {@link Single} that counts the total number of items emitted by the current {@code Flowable} and emits * this count as a 64-bit {@link Long}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code count} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Single} instance * @see ReactiveX operators documentation: Count */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single count() { return RxJavaPlugins.onAssembly(new FlowableCountSingle<>(this)); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, except that it drops items emitted by the * current {@code Flowable} that are followed by another item within a computed debounce duration. *

* *

* The delivery of the item happens on the thread of the first {@code onNext} or {@code onComplete} * signal of the generated {@link Publisher} sequence, * which if takes too long, a newer item may arrive from the upstream, causing the * generated sequence to get cancelled, which may also interrupt any downstream blocking operation * (yielding an {@code InterruptedException}). It is recommended processing items * that may take long time to be moved to another thread via {@link #observeOn} applied after * {@code debounce} itself. *

*
Backpressure:
*
This operator does not support backpressure as it uses the {@code debounceSelector} to mark * boundaries.
*
Scheduler:
*
This version of {@code debounce} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the debounce value type (ignored) * @param debounceIndicator * function to retrieve a sequence that indicates the throttle duration for each item * @return the new {@code Flowable} instance * @throws NullPointerException if {@code debounceIndicator} is {@code null} * @see ReactiveX operators documentation: Debounce * @see RxJava wiki: Backpressure */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable debounce(@NonNull Function> debounceIndicator) { Objects.requireNonNull(debounceIndicator, "debounceIndicator is null"); return RxJavaPlugins.onAssembly(new FlowableDebounce<>(this, debounceIndicator)); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, except that it drops items emitted by the * current {@code Flowable} that are followed by newer items before a timeout value expires. The timer resets on * each emission. *

* Note: If items keep being emitted by the current {@code Flowable} faster than the timeout then no items * will be emitted by the resulting {@code Flowable}. *

* *

* Delivery of the item after the grace period happens on the {@code computation} {@link Scheduler}'s * {@code Worker} which if takes too long, a newer item may arrive from the upstream, causing the * {@code Worker}'s task to get disposed, which may also interrupt any downstream blocking operation * (yielding an {@code InterruptedException}). It is recommended processing items * that may take long time to be moved to another thread via {@link #observeOn} applied after * {@code debounce} itself. *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
{@code debounce} operates by default on the {@code computation} {@code Scheduler}.
*
* * @param timeout * the length of the window of time that must pass after the emission of an item from the current * {@code Flowable} in which it emits no items in order for the item to be emitted by the * resulting {@code Flowable} * @param unit * the unit of time for the specified {@code timeout} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Debounce * @see RxJava wiki: Backpressure * @see #throttleWithTimeout(long, TimeUnit) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable debounce(long timeout, @NonNull TimeUnit unit) { return debounce(timeout, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, except that it drops items emitted by the * current {@code Flowable} that are followed by newer items before a timeout value expires on a specified * {@link Scheduler}. The timer resets on each emission. *

* Note: If items keep being emitted by the current {@code Flowable} faster than the timeout then no items * will be emitted by the resulting {@code Flowable}. *

* *

* Delivery of the item after the grace period happens on the given {@code Scheduler}'s * {@code Worker} which if takes too long, a newer item may arrive from the upstream, causing the * {@code Worker}'s task to get disposed, which may also interrupt any downstream blocking operation * (yielding an {@code InterruptedException}). It is recommended processing items * that may take long time to be moved to another thread via {@link #observeOn} applied after * {@code debounce} itself. *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param timeout * the time each item has to be "the most recent" of those emitted by the current {@code Flowable} to * ensure that it's not dropped * @param unit * the unit of time for the specified {@code timeout} * @param scheduler * the {@code Scheduler} to use internally to manage the timers that handle the timeout for each * item * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Debounce * @see RxJava wiki: Backpressure * @see #throttleWithTimeout(long, TimeUnit, Scheduler) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable debounce(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableDebounceTimed<>(this, timeout, unit, scheduler, null)); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, except that it drops items emitted by the * current {@code Flowable} that are followed by newer items before a timeout value expires on a specified * {@link Scheduler}. The timer resets on each emission. *

* Note: If items keep being emitted by the current {@code Flowable} faster than the timeout then no items * will be emitted by the resulting {@code Flowable}. *

* *

* Delivery of the item after the grace period happens on the given {@code Scheduler}'s * {@code Worker} which if takes too long, a newer item may arrive from the upstream, causing the * {@code Worker}'s task to get disposed, which may also interrupt any downstream blocking operation * (yielding an {@code InterruptedException}). It is recommended processing items * that may take long time to be moved to another thread via {@link #observeOn} applied after * {@code debounce} itself. *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param timeout * the time each item has to be "the most recent" of those emitted by the current {@code Flowable} to * ensure that it's not dropped * @param unit * the unit of time for the specified {@code timeout} * @param scheduler * the {@code Scheduler} to use internally to manage the timers that handle the timeout for each * item * @param onDropped * called with the current entry when it has been replaced by a new one * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} or {@code onDropped} is {@code null} * @see ReactiveX operators documentation: Debounce * @see RxJava wiki: Backpressure * @see #throttleWithTimeout(long, TimeUnit, Scheduler, Consumer) * @since 3.1.6 - Experimental */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @Experimental public final Flowable debounce(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, @NonNull Consumer onDropped) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); Objects.requireNonNull(onDropped, "onDropped is null"); return RxJavaPlugins.onAssembly(new FlowableDebounceTimed<>(this, timeout, unit, scheduler, onDropped)); } /** * Returns a {@code Flowable} that emits the items emitted by the current {@code Flowable} or a specified default item * if the current {@code Flowable} is empty. *

* *

*
Backpressure:
*
If the current {@code Flowable} is empty, this operator is guaranteed to honor backpressure from downstream. * If the current {@code Flowable} is non-empty, it is expected to honor backpressure as well; if the rule is violated, * a {@link MissingBackpressureException} may get signaled somewhere downstream. *
*
Scheduler:
*
{@code defaultIfEmpty} does not operate by default on a particular {@link Scheduler}.
*
* * @param defaultItem * the item to emit if the current {@code Flowable} emits no items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code defaultItem} is {@code null} * @see ReactiveX operators documentation: DefaultIfEmpty */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable defaultIfEmpty(@NonNull T defaultItem) { Objects.requireNonNull(defaultItem, "defaultItem is null"); return switchIfEmpty(just(defaultItem)); } /** * Returns a {@code Flowable} that delays the emissions of the current {@code Flowable} via another {@link Publisher} on a * per-item basis. *

* *

* Note: the resulting {@code Flowable} will immediately propagate any {@code onError} notification * from the current {@code Flowable}. *

*
Backpressure:
*
The operator doesn't interfere with the backpressure behavior which is determined by the current {@code Flowable}. * All of the other {@code Publisher}s supplied by the function are consumed * in an unbounded manner (i.e., no backpressure applied to them).
*
Scheduler:
*
This version of {@code delay} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the item delay value type (ignored) * @param itemDelayIndicator * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable}, which is * then used to delay the emission of that item by the resulting {@code Flowable} until the {@code Publisher} * returned from {@code itemDelay} emits an item * @return the new {@code Flowable} instance * @throws NullPointerException if {@code itemDelayIndicator} is {@code null} * @see ReactiveX operators documentation: Delay */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable delay(@NonNull Function> itemDelayIndicator) { Objects.requireNonNull(itemDelayIndicator, "itemDelayIndicator is null"); return flatMap(FlowableInternalHelper.itemDelay(itemDelayIndicator)); } /** * Returns a {@code Flowable} that emits the items emitted by the current {@code Flowable} shifted forward in time by a * specified delay. The {@code onError} notification from the current {@code Flowable} is not delayed. *

* *

*
Backpressure:
*
The operator doesn't interfere with the backpressure behavior which is determined by the current {@code Flowable}.
*
Scheduler:
*
This version of {@code delay} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param time * the delay to shift the source by * @param unit * the {@link TimeUnit} in which {@code period} is defined * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Delay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable delay(long time, @NonNull TimeUnit unit) { return delay(time, unit, Schedulers.computation(), false); } /** * Returns a {@code Flowable} that emits the items emitted by the current {@code Flowable} shifted forward in time by a * specified delay. If {@code delayError} is {@code true}, error notifications will also be delayed. *

* *

*
Backpressure:
*
The operator doesn't interfere with the backpressure behavior which is determined by the current {@code Flowable}.
*
Scheduler:
*
This version of {@code delay} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param time * the delay to shift the source by * @param unit * the {@link TimeUnit} in which {@code period} is defined * @param delayError * if {@code true}, the upstream exception is signaled with the given delay, after all preceding normal elements, * if {@code false}, the upstream exception is signaled immediately * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Delay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable delay(long time, @NonNull TimeUnit unit, boolean delayError) { return delay(time, unit, Schedulers.computation(), delayError); } /** * Returns a {@code Flowable} that emits the items emitted by the current {@code Flowable} shifted forward in time by a * specified delay. The {@code onError} notification from the current {@code Flowable} is not delayed. *

* *

*
Backpressure:
*
The operator doesn't interfere with the backpressure behavior which is determined by the current {@code Flowable}.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param time * the delay to shift the source by * @param unit * the time unit of {@code delay} * @param scheduler * the {@code Scheduler} to use for delaying * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Delay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable delay(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return delay(time, unit, scheduler, false); } /** * Returns a {@code Flowable} that emits the items emitted by the current {@code Flowable} shifted forward in time by a * specified delay. If {@code delayError} is {@code true}, error notifications will also be delayed. *

* *

*
Backpressure:
*
The operator doesn't interfere with the backpressure behavior which is determined by the current {@code Flowable}.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param time * the delay to shift the source by * @param unit * the time unit of {@code delay} * @param scheduler * the {@code Scheduler} to use for delaying * @param delayError * if {@code true}, the upstream exception is signaled with the given delay, after all preceding normal elements, * if {@code false}, the upstream exception is signaled immediately * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Delay */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable delay(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean delayError) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableDelay<>(this, Math.max(0L, time), unit, scheduler, delayError)); } /** * Returns a {@code Flowable} that delays the subscription to and emissions from the current {@code Flowable} via another * {@link Publisher} on a per-item basis. *

* *

* Note: the resulting {@code Flowable} will immediately propagate any {@code onError} notification * from the current {@code Flowable}. *

*
Backpressure:
*
The operator doesn't interfere with the backpressure behavior which is determined by the current {@code Flowable}. * All of the other {@code Publisher}s supplied by the functions are consumed * in an unbounded manner (i.e., no backpressure applied to them).
*
Scheduler:
*
This version of {@code delay} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the subscription delay value type (ignored) * @param * the item delay value type (ignored) * @param subscriptionIndicator * a function that returns a {@code Publisher} that triggers the subscription to the current {@code Flowable} * once it emits any item * @param itemDelayIndicator * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable}, which is * then used to delay the emission of that item by the resulting {@code Flowable} until the {@code Publisher} * returned from {@code itemDelay} emits an item * @return the new {@code Flowable} instance * @throws NullPointerException if {@code subscriptionIndicator} and {@code itemDelayIndicator} is {@code null} * @see ReactiveX operators documentation: Delay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U, @NonNull V> Flowable delay(@NonNull Publisher subscriptionIndicator, @NonNull Function> itemDelayIndicator) { return delaySubscription(subscriptionIndicator).delay(itemDelayIndicator); } /** * Returns a {@code Flowable} that delays the subscription to this {@link Publisher} * until the other {@code Publisher} emits an element or completes normally. *
*
Backpressure:
*
The operator forwards the backpressure requests to this {@code Publisher} once * the subscription happens and requests {@link Long#MAX_VALUE} from the other {@code Publisher}
*
Scheduler:
*
This method does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the other {@code Publisher}, irrelevant * @param subscriptionIndicator the other {@code Publisher} that should trigger the subscription * to this {@code Publisher}. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code subscriptionIndicator} is {@code null} * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable delaySubscription(@NonNull Publisher subscriptionIndicator) { Objects.requireNonNull(subscriptionIndicator, "subscriptionIndicator is null"); return RxJavaPlugins.onAssembly(new FlowableDelaySubscriptionOther<>(this, subscriptionIndicator)); } /** * Returns a {@code Flowable} that delays the subscription to the current {@code Flowable} by a given amount of time. *

* *

*
Backpressure:
*
The operator doesn't interfere with the backpressure behavior which is determined by the current {@code Flowable}.
*
Scheduler:
*
This version of {@code delaySubscription} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param time * the time to delay the subscription * @param unit * the time unit of {@code delay} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Delay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable delaySubscription(long time, @NonNull TimeUnit unit) { return delaySubscription(time, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that delays the subscription to the current {@code Flowable} by a given amount of time, * both waiting and subscribing on a given {@link Scheduler}. *

* *

*
Backpressure:
*
The operator doesn't interfere with the backpressure behavior which is determined by the current {@code Flowable}.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param time * the time to delay the subscription * @param unit * the time unit of {@code delay} * @param scheduler * the {@code Scheduler} on which the waiting and subscription will happen * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Delay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable delaySubscription(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return delaySubscription(timer(time, unit, scheduler)); } /** * Returns a {@code Flowable} that reverses the effect of {@link #materialize materialize} by transforming the * {@link Notification} objects extracted from the source items via a selector function * into their respective {@link Subscriber} signal types. *

* *

* The intended use of the {@code selector} function is to perform a * type-safe identity mapping (see example) on a source that is already of type * {@code Notification}. The Java language doesn't allow * limiting instance methods to a certain generic argument shape, therefore, * a function is used to ensure the conversion remains type safe. *

* When the upstream signals an {@link Notification#createOnError(Throwable) onError} or * {@link Notification#createOnComplete() onComplete} item, the * returned {@code Flowable} cancels of the flow and terminates with that type of terminal event: *


     * Flowable.just(createOnNext(1), createOnComplete(), createOnNext(2))
     * .doOnCancel(() -> System.out.println("Canceled!"));
     * .dematerialize(notification -> notification)
     * .test()
     * .assertResult(1);
     * 
* If the upstream signals {@code onError} or {@code onComplete} directly, the flow is terminated * with the same event. *

     * Flowable.just(createOnNext(1), createOnNext(2))
     * .dematerialize(notification -> notification)
     * .test()
     * .assertResult(1, 2);
     * 
* If this behavior is not desired, the completion can be suppressed by applying {@link #concatWith(Publisher)} * with a {@link #never()} source. *
*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code dematerialize} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.2.4 - experimental * * @param the output value type * @param selector function that returns the upstream item and should return a {@code Notification} to signal * the corresponding {@code Subscriber} event to the downstream. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector} is {@code null} * @see ReactiveX operators documentation: Dematerialize * @since 3.0.0 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.PASS_THROUGH) public final <@NonNull R> Flowable dematerialize(@NonNull Function<@NonNull ? super T, @NonNull Notification> selector) { Objects.requireNonNull(selector, "selector is null"); return RxJavaPlugins.onAssembly(new FlowableDematerialize<>(this, selector)); } /** * Returns a {@code Flowable} that emits all items emitted by the current {@code Flowable} that are distinct * based on {@link Object#equals(Object)} comparison. *

* *

* It is recommended the elements' class {@code T} in the flow overrides the default {@code Object.equals()} and {@link Object#hashCode()} to provide * a meaningful comparison between items as the default Java implementation only considers reference equivalence. *

* By default, {@code distinct()} uses an internal {@link java.util.HashSet} per {@link Subscriber} to remember * previously seen items and uses {@link java.util.Set#add(Object)} returning {@code false} as the * indicator for duplicates. *

* Note that this internal {@link HashSet} may grow unbounded as items won't be removed from it by * the operator. Therefore, using very long or infinite upstream (with very distinct elements) may lead * to {@link OutOfMemoryError}. *

* Customizing the retention policy can happen only by providing a custom {@link java.util.Collection} implementation * to the {@link #distinct(Function, Supplier)} overload. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code distinct} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: Distinct * @see #distinct(Function) * @see #distinct(Function, Supplier) */ @SuppressWarnings({ "rawtypes", "unchecked" }) @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable distinct() { return distinct((Function)Functions.identity(), Functions.createHashSet()); } /** * Returns a {@code Flowable} that emits all items emitted by the current {@code Flowable} that are distinct according * to a key selector function and based on {@link Object#equals(Object)} comparison of the objects * returned by the key selector function. *

* *

* It is recommended the keys' class {@code K} overrides the default {@code Object.equals()} and {@link Object#hashCode()} to provide * a meaningful comparison between the key objects as the default Java implementation only considers reference equivalence. *

* By default, {@code distinct()} uses an internal {@link java.util.HashSet} per {@link Subscriber} to remember * previously seen keys and uses {@link java.util.Set#add(Object)} returning {@code false} as the * indicator for duplicates. *

* Note that this internal {@link HashSet} may grow unbounded as keys won't be removed from it by * the operator. Therefore, using very long or infinite upstream (with very distinct keys) may lead * to {@link OutOfMemoryError}. *

* Customizing the retention policy can happen only by providing a custom {@link java.util.Collection} implementation * to the {@link #distinct(Function, Supplier)} overload. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code distinct} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type * @param keySelector * a function that projects an emitted item to a key value that is used to decide whether an item * is distinct from another one or not * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector} is {@code null} * @see ReactiveX operators documentation: Distinct * @see #distinct(Function, Supplier) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K> Flowable distinct(@NonNull Function keySelector) { return distinct(keySelector, Functions.createHashSet()); } /** * Returns a {@code Flowable} that emits all items emitted by the current {@code Flowable} that are distinct according * to a key selector function and based on {@link Object#equals(Object)} comparison of the objects * returned by the key selector function. *

* *

* It is recommended the keys' class {@code K} overrides the default {@code Object.equals()} and {@link Object#hashCode()} to provide * a meaningful comparison between the key objects as the default Java implementation only considers reference equivalence. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code distinct} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type * @param keySelector * a function that projects an emitted item to a key value that is used to decide whether an item * is distinct from another one or not * @param collectionSupplier * function called for each individual {@link Subscriber} to return a {@link Collection} subtype for holding the extracted * keys and whose add() method's return indicates uniqueness. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector} or {@code collectionSupplier} is {@code null} * @see ReactiveX operators documentation: Distinct */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K> Flowable distinct(@NonNull Function keySelector, @NonNull Supplier> collectionSupplier) { Objects.requireNonNull(keySelector, "keySelector is null"); Objects.requireNonNull(collectionSupplier, "collectionSupplier is null"); return RxJavaPlugins.onAssembly(new FlowableDistinct<>(this, keySelector, collectionSupplier)); } /** * Returns a {@code Flowable} that emits all items emitted by the current {@code Flowable} that are distinct from their * immediate predecessors based on {@link Object#equals(Object)} comparison. *

* *

* It is recommended the elements' class {@code T} in the flow overrides the default {@code Object.equals()} to provide * a meaningful comparison between items as the default Java implementation only considers reference equivalence. * Alternatively, use the {@link #distinctUntilChanged(BiPredicate)} overload and provide a comparison function * in case the class {@code T} can't be overridden with custom {@code equals()} or the comparison itself * should happen on different terms or properties of the class {@code T}. *

* Note that the operator always retains the latest item from upstream regardless of the comparison result * and uses it in the next comparison with the next upstream item. *

* Note that if element type {@code T} in the flow is mutable, the comparison of the previous and current * item may yield unexpected results if the items are mutated externally. Common cases are mutable * {@link CharSequence}s or {@link List}s where the objects will actually have the same * references when they are modified and {@code distinctUntilChanged} will evaluate subsequent items as same. * To avoid such situation, it is recommended that mutable data is converted to an immutable one, * for example using {@code map(CharSequence::toString)} or {@code map(list -> Collections.unmodifiableList(new ArrayList<>(list)))}. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code distinctUntilChanged} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: Distinct * @see #distinctUntilChanged(BiPredicate) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable distinctUntilChanged() { return distinctUntilChanged(Functions.identity()); } /** * Returns a {@code Flowable} that emits all items emitted by the current {@code Flowable} that are distinct from their * immediate predecessors, according to a key selector function and based on {@link Object#equals(Object)} comparison * of those objects returned by the key selector function. *

* *

* It is recommended the keys' class {@code K} overrides the default {@code Object.equals()} to provide * a meaningful comparison between the key objects as the default Java implementation only considers reference equivalence. * Alternatively, use the {@link #distinctUntilChanged(BiPredicate)} overload and provide a comparison function * in case the class {@code K} can't be overridden with custom {@code equals()} or the comparison itself * should happen on different terms or properties of the item class {@code T} (for which the keys can be * derived via a similar selector). *

* Note that the operator always retains the latest key from upstream regardless of the comparison result * and uses it in the next comparison with the next key derived from the next upstream item. *

* Note that if element type {@code T} in the flow is mutable, the comparison of the previous and current * item may yield unexpected results if the items are mutated externally. Common cases are mutable * {@link CharSequence}s or {@link List}s where the objects will actually have the same * references when they are modified and {@code distinctUntilChanged} will evaluate subsequent items as same. * To avoid such situation, it is recommended that mutable data is converted to an immutable one, * for example using {@code map(CharSequence::toString)} or {@code map(list -> Collections.unmodifiableList(new ArrayList<>(list)))}. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code distinctUntilChanged} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type * @param keySelector * a function that projects an emitted item to a key value that is used to decide whether an item * is distinct from another one or not * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector} is {@code null} * @see ReactiveX operators documentation: Distinct */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K> Flowable distinctUntilChanged(@NonNull Function keySelector) { Objects.requireNonNull(keySelector, "keySelector is null"); return RxJavaPlugins.onAssembly(new FlowableDistinctUntilChanged<>(this, keySelector, ObjectHelper.equalsPredicate())); } /** * Returns a {@code Flowable} that emits all items emitted by the current {@code Flowable} that are distinct from their * immediate predecessors when compared with each other via the provided comparator function. *

* *

* Note that the operator always retains the latest item from upstream regardless of the comparison result * and uses it in the next comparison with the next upstream item. *

* Note that if element type {@code T} in the flow is mutable, the comparison of the previous and current * item may yield unexpected results if the items are mutated externally. Common cases are mutable * {@link CharSequence}s or {@link List}s where the objects will actually have the same * references when they are modified and {@code distinctUntilChanged} will evaluate subsequent items as same. * To avoid such situation, it is recommended that mutable data is converted to an immutable one, * for example using {@code map(CharSequence::toString)} or {@code map(list -> Collections.unmodifiableList(new ArrayList<>(list)))}. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code distinctUntilChanged} does not operate by default on a particular {@link Scheduler}.
*
* * @param comparer the function that receives the previous item and the current item and is * expected to return {@code true} if the two are equal, thus skipping the current value. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code comparer} is {@code null} * @see ReactiveX operators documentation: Distinct * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable distinctUntilChanged(@NonNull BiPredicate comparer) { Objects.requireNonNull(comparer, "comparer is null"); return RxJavaPlugins.onAssembly(new FlowableDistinctUntilChanged<>(this, Functions.identity(), comparer)); } /** * Calls the specified action after this {@code Flowable} signals {@code onError} or {@code onComplete} or gets canceled by * the downstream. *

In case of a race between a terminal event and a cancellation, the provided {@code onFinally} action * is executed once per subscription. *

Note that the {@code onFinally} action is shared between subscriptions and as such * should be thread-safe. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code doFinally} does not operate by default on a particular {@link Scheduler}.
*
Operator-fusion:
*
This operator supports normal and conditional {@link Subscriber}s as well as boundary-limited * synchronous or asynchronous queue-fusion.
*
*

History: 2.0.1 - experimental * @param onFinally the action called when this {@code Flowable} terminates or gets canceled * @throws NullPointerException if {@code onFinally} is {@code null} * @return the new {@code Flowable} instance * @since 2.1 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doFinally(@NonNull Action onFinally) { Objects.requireNonNull(onFinally, "onFinally is null"); return RxJavaPlugins.onAssembly(new FlowableDoFinally<>(this, onFinally)); } /** * Calls the specified consumer with the current item after this item has been emitted to the downstream. *

Note that the {@code onAfterNext} action is shared between subscriptions and as such * should be thread-safe. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code doAfterNext} does not operate by default on a particular {@link Scheduler}.
*
Operator-fusion:
*
This operator supports normal and conditional {@link Subscriber}s as well as boundary-limited * synchronous or asynchronous queue-fusion.
*
*

History: 2.0.1 - experimental * @param onAfterNext the {@link Consumer} that will be called after emitting an item from upstream to the downstream * @throws NullPointerException if {@code onAfterNext} is {@code null} * @return the new {@code Flowable} instance * @since 2.1 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doAfterNext(@NonNull Consumer onAfterNext) { Objects.requireNonNull(onAfterNext, "onAfterNext is null"); return RxJavaPlugins.onAssembly(new FlowableDoAfterNext<>(this, onAfterNext)); } /** * Registers an {@link Action} to be called when this {@link Publisher} invokes either * {@link Subscriber#onComplete onComplete} or {@link Subscriber#onError onError}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code doAfterTerminate} does not operate by default on a particular {@link Scheduler}.
*
* * @param onAfterTerminate * an {@code Action} to be invoked when the current {@code Flowable} finishes * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onAfterTerminate} is {@code null} * @see ReactiveX operators documentation: Do * @see #doOnTerminate(Action) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doAfterTerminate(@NonNull Action onAfterTerminate) { return doOnEach(Functions.emptyConsumer(), Functions.emptyConsumer(), Functions.EMPTY_ACTION, onAfterTerminate); } /** * Calls the cancel {@link Action} if the downstream cancels the sequence. *

* *

* The action is shared between subscriptions and thus may be called concurrently from multiple * threads; the action must be thread-safe. *

* If the action throws a runtime exception, that exception is rethrown by the {@code onCancel()} call, * sometimes as a {@link CompositeException} if there were multiple exceptions along the way. *

*
Backpressure:
*
{@code doOnCancel} does not interact with backpressure requests or value delivery; backpressure * behavior is preserved between its upstream and its downstream.
*
Scheduler:
*
{@code doOnCancel} does not operate by default on a particular {@link Scheduler}.
*
* * @param onCancel * the action that gets called when the current {@code Flowable}'s {@link Subscription} is canceled * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onCancel} is {@code null} * @see ReactiveX operators documentation: Do */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doOnCancel(@NonNull Action onCancel) { return doOnLifecycle(Functions.emptyConsumer(), Functions.EMPTY_LONG_CONSUMER, onCancel); } /** * Invokes an {@link Action} just before the current {@code Flowable} calls {@code onComplete}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnComplete} does not operate by default on a particular {@link Scheduler}.
*
* * @param onComplete * the action to invoke when the current {@code Flowable} calls {@code onComplete} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onComplete} is {@code null} * @see ReactiveX operators documentation: Do */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doOnComplete(@NonNull Action onComplete) { return doOnEach(Functions.emptyConsumer(), Functions.emptyConsumer(), onComplete, Functions.EMPTY_ACTION); } /** * Calls the appropriate onXXX consumer (shared between all subscribers) whenever a signal with the same type * passes through, before forwarding them to downstream. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnEach} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext the {@link Consumer} to invoke when the current {@code Flowable} calls {@code onNext} * @param onError the {@code Consumer} to invoke when the current {@code Flowable} calls {@code onError} * @param onComplete the {@link Action} to invoke when the current {@code Flowable} calls {@code onComplete} * @param onAfterTerminate the {@code Action} to invoke when the current {@code Flowable} calls {@code onAfterTerminate} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onNext}, {@code onError}, {@code onComplete} or {@code onAfterTerminate} is {@code null} * @see ReactiveX operators documentation: Do */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) private Flowable doOnEach(@NonNull Consumer onNext, @NonNull Consumer onError, Action onComplete, Action onAfterTerminate) { Objects.requireNonNull(onNext, "onNext is null"); Objects.requireNonNull(onError, "onError is null"); Objects.requireNonNull(onComplete, "onComplete is null"); Objects.requireNonNull(onAfterTerminate, "onAfterTerminate is null"); return RxJavaPlugins.onAssembly(new FlowableDoOnEach<>(this, onNext, onError, onComplete, onAfterTerminate)); } /** * Invokes a {@link Consumer} with a {@link Notification} instances matching the signals emitted by the current {@code Flowable} * before they are forwarded to the downstream. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnEach} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNotification * the action to invoke for each item emitted by the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onNotification} is {@code null} * @see ReactiveX operators documentation: Do */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable doOnEach(@NonNull Consumer<@NonNull ? super Notification> onNotification) { Objects.requireNonNull(onNotification, "onNotification is null"); return doOnEach( Functions.notificationOnNext(onNotification), Functions.notificationOnError(onNotification), Functions.notificationOnComplete(onNotification), Functions.EMPTY_ACTION ); } /** * Calls the appropriate methods of the given {@link Subscriber} when the current {@code Flowable} signals events before forwarding it * to the downstream. *

* In case the {@code onError} of the supplied {@code Subscriber} throws, the downstream will receive a composite * exception containing the original exception and the exception thrown by {@code onError}. If either the * {@code onNext} or the {@code onComplete} method of the supplied {@code Subscriber} throws, the downstream will be * terminated and will receive this thrown exception. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnEach} does not operate by default on a particular {@link Scheduler}.
*
* * @param subscriber * the {@code Subscriber} to be notified about {@code onNext}, {@code onError} and {@code onComplete} events on its * respective methods before the actual downstream {@code Subscriber} gets notified. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code subscriber} is {@code null} * @see ReactiveX operators documentation: Do */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable doOnEach(@NonNull Subscriber subscriber) { Objects.requireNonNull(subscriber, "subscriber is null"); return doOnEach( FlowableInternalHelper.subscriberOnNext(subscriber), FlowableInternalHelper.subscriberOnError(subscriber), FlowableInternalHelper.subscriberOnComplete(subscriber), Functions.EMPTY_ACTION); } /** * Calls the given {@link Consumer} with the error {@link Throwable} if the current {@code Flowable} failed before forwarding it to * the downstream. *

* In case the {@code onError} action throws, the downstream will receive a composite exception containing * the original exception and the exception thrown by {@code onError}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnError} does not operate by default on a particular {@link Scheduler}.
*
* * @param onError * the action to invoke if the current {@code Flowable} calls {@code onError} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onError} is {@code null} * @see ReactiveX operators documentation: Do */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doOnError(@NonNull Consumer onError) { return doOnEach(Functions.emptyConsumer(), onError, Functions.EMPTY_ACTION, Functions.EMPTY_ACTION); } /** * Calls the appropriate {@code onXXX} method (shared between all {@link Subscriber}s) for the lifecycle events of * the sequence (subscription, cancellation, requesting). *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnLifecycle} does not operate by default on a particular {@link Scheduler}.
*
* * @param onSubscribe * a {@link Consumer} called with the {@link Subscription} sent via {@link Subscriber#onSubscribe(Subscription)} * @param onRequest * a {@link LongConsumer} called with the request amount sent via {@link Subscription#request(long)} * @param onCancel * called when the downstream cancels the {@code Subscription} via {@link Subscription#cancel()} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onSubscribe}, {@code onRequest} or {@code onCancel} is {@code null} * @see ReactiveX operators documentation: Do */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable doOnLifecycle(@NonNull Consumer onSubscribe, @NonNull LongConsumer onRequest, @NonNull Action onCancel) { Objects.requireNonNull(onSubscribe, "onSubscribe is null"); Objects.requireNonNull(onRequest, "onRequest is null"); Objects.requireNonNull(onCancel, "onCancel is null"); return RxJavaPlugins.onAssembly(new FlowableDoOnLifecycle<>(this, onSubscribe, onRequest, onCancel)); } /** * Calls the given {@link Consumer} with the value emitted by the current {@code Flowable} before forwarding it to the downstream. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnNext} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext * the action to invoke when the current {@code Flowable} calls {@code onNext} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onNext} is {@code null} * @see ReactiveX operators documentation: Do * @see #doAfterNext(Consumer) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doOnNext(@NonNull Consumer onNext) { return doOnEach(onNext, Functions.emptyConsumer(), Functions.EMPTY_ACTION, Functions.EMPTY_ACTION); } /** * Calls the given {@link LongConsumer} with the request amount from the downstream before forwarding it * to the current {@code Flowable}. *

* Note: This operator is for tracing the internal behavior of back-pressure request * patterns and generally intended for debugging use. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnRequest} does not operate by default on a particular {@link Scheduler}.
*
* * @param onRequest * the action that gets called when a {@link Subscriber} requests items from the current * {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onRequest} is {@code null} * @see ReactiveX operators * documentation: Do * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doOnRequest(@NonNull LongConsumer onRequest) { return doOnLifecycle(Functions.emptyConsumer(), onRequest, Functions.EMPTY_ACTION); } /** * Calls the given {@link Consumer} with the {@link Subscription} provided by the current {@code Flowable} upon * subscription from the downstream before forwarding it to the subscriber's * {@link Subscriber#onSubscribe(Subscription) onSubscribe} method. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnSubscribe} does not operate by default on a particular {@link Scheduler}.
*
* * @param onSubscribe * the {@code Consumer} that gets called when a {@link Subscriber} subscribes to the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onSubscribe} is {@code null} * @see ReactiveX operators documentation: Do */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doOnSubscribe(@NonNull Consumer onSubscribe) { return doOnLifecycle(onSubscribe, Functions.EMPTY_LONG_CONSUMER, Functions.EMPTY_ACTION); } /** * Calls the given {@link Action} when the current {@code Flowable} completes normally or with an error before those signals * are forwarded to the downstream. *

* *

* This differs from {@code doAfterTerminate} in that this happens before the {@code onComplete} or * {@code onError} notification. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s * backpressure behavior.
*
Scheduler:
*
{@code doOnTerminate} does not operate by default on a particular {@link Scheduler}.
*
* * @param onTerminate * the action to invoke when the current {@code Flowable} calls {@code onComplete} or {@code onError} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onTerminate} is {@code null} * @see ReactiveX operators documentation: Do * @see #doAfterTerminate(Action) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable doOnTerminate(@NonNull Action onTerminate) { return doOnEach(Functions.emptyConsumer(), Functions.actionConsumer(onTerminate), onTerminate, Functions.EMPTY_ACTION); } /** * Returns a {@link Maybe} that emits the single item at a specified index in a sequence of emissions from * this {@code Flowable} or completes if this {@code Flowable} sequence has fewer elements than index. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in a bounded manner.
*
Scheduler:
*
{@code elementAt} does not operate by default on a particular {@link Scheduler}.
*
* * @param index * the zero-based index of the item to retrieve * @return the new {@code Maybe} instance * @throws IndexOutOfBoundsException if {@code index} is negative * @see ReactiveX operators documentation: ElementAt */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Maybe elementAt(long index) { if (index < 0) { throw new IndexOutOfBoundsException("index >= 0 required but it was " + index); } return RxJavaPlugins.onAssembly(new FlowableElementAtMaybe<>(this, index)); } /** * Returns a {@link Single} that emits the item found at a specified index in a sequence of emissions from * this {@code Flowable}, or a default item if that index is out of range. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in a bounded manner.
*
Scheduler:
*
{@code elementAt} does not operate by default on a particular {@link Scheduler}.
*
* * @param index * the zero-based index of the item to retrieve * @param defaultItem * the default item * @return the new {@code Single} instance * @throws NullPointerException if {@code defaultItem} is {@code null} * @throws IndexOutOfBoundsException * if {@code index} is negative * @see ReactiveX operators documentation: ElementAt */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Single elementAt(long index, @NonNull T defaultItem) { if (index < 0) { throw new IndexOutOfBoundsException("index >= 0 required but it was " + index); } Objects.requireNonNull(defaultItem, "defaultItem is null"); return RxJavaPlugins.onAssembly(new FlowableElementAtSingle<>(this, index, defaultItem)); } /** * Returns a {@link Single} that emits the item found at a specified index in a sequence of emissions from * this {@code Flowable} or signals a {@link NoSuchElementException} if this {@code Flowable} has fewer elements than index. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in a bounded manner.
*
Scheduler:
*
{@code elementAtOrError} does not operate by default on a particular {@link Scheduler}.
*
* * @param index * the zero-based index of the item to retrieve * @return the new {@code Single} instance * @throws IndexOutOfBoundsException * if {@code index} is less than 0 * @see ReactiveX operators documentation: ElementAt */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single elementAtOrError(long index) { if (index < 0) { throw new IndexOutOfBoundsException("index >= 0 required but it was " + index); } return RxJavaPlugins.onAssembly(new FlowableElementAtSingle<>(this, index, null)); } /** * Filters items emitted by the current {@code Flowable} by only emitting those that satisfy a specified predicate. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code filter} does not operate by default on a particular {@link Scheduler}.
*
* * @param predicate * a function that evaluates each item emitted by the current {@code Flowable}, returning {@code true} * if it passes the filter * @return the new {@code Flowable} instance * @throws NullPointerException if {@code predicate} is {@code null} * @see ReactiveX operators documentation: Filter */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable filter(@NonNull Predicate predicate) { Objects.requireNonNull(predicate, "predicate is null"); return RxJavaPlugins.onAssembly(new FlowableFilter<>(this, predicate)); } /** * Returns a {@link Maybe} that emits only the very first item emitted by this {@code Flowable} or * completes if this {@code Flowable} is empty. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in a bounded manner.
*
Scheduler:
*
{@code firstElement} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Maybe} instance * @see ReactiveX operators documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Maybe firstElement() { return elementAt(0); } /** * Returns a {@link Single} that emits only the very first item emitted by this {@code Flowable}, or a default * item if this {@code Flowable} completes without emitting anything. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in a bounded manner.
*
Scheduler:
*
{@code first} does not operate by default on a particular {@link Scheduler}.
*
* * @param defaultItem * the default item to emit if the current {@code Flowable} doesn't emit anything * @return the new {@code Single} instance * @throws NullPointerException if {@code defaultItem} is {@code null} * @see ReactiveX operators documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single first(@NonNull T defaultItem) { return elementAt(0, defaultItem); } /** * Returns a {@link Single} that emits only the very first item emitted by this {@code Flowable} or * signals a {@link NoSuchElementException} if this {@code Flowable} is empty. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in a bounded manner.
*
Scheduler:
*
{@code firstOrError} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Single} instance * @see ReactiveX operators documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) // take may trigger UNBOUNDED_IN @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single firstOrError() { return elementAtOrError(0); } /** * Returns a {@code Flowable} that emits items based on applying a function that you supply to each item emitted * by the current {@code Flowable}, where that function returns a {@link Publisher}, and then merging those resulting * {@code Publisher}s and emitting the results of this merger. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@link #bufferSize()} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the inner {@code Publisher}s and the output type * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable flatMap(@NonNull Function> mapper) { return flatMap(mapper, false, bufferSize(), bufferSize()); } /** * Returns a {@code Flowable} that emits items based on applying a function that you supply to each item emitted * by the current {@code Flowable}, where that function returns a {@link Publisher}, and then merging those resulting * {@code Publisher}s and emitting the results of this merger. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@link #bufferSize()} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the inner {@code Publisher}s and the output type * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @param delayErrors * if {@code true}, exceptions from the current {@code Flowable} and all inner {@code Publisher}s are delayed until all of them terminate * if {@code false}, the first one signaling an exception will terminate the whole sequence immediately * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable flatMap(@NonNull Function> mapper, boolean delayErrors) { return flatMap(mapper, delayErrors, bufferSize(), bufferSize()); } /** * Returns a {@code Flowable} that emits items based on applying a function that you supply to each item emitted * by the current {@code Flowable}, where that function returns a {@link Publisher}, and then merging those resulting * {@code Publisher}s and emitting the results of this merger, while limiting the maximum number of concurrent * subscriptions to these {@code Publisher}s. * * *
*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@code maxConcurrency} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the inner {@code Publisher}s and the output type * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable flatMap(@NonNull Function> mapper, int maxConcurrency) { return flatMap(mapper, false, maxConcurrency, bufferSize()); } /** * Returns a {@code Flowable} that emits items based on applying a function that you supply to each item emitted * by the current {@code Flowable}, where that function returns a {@link Publisher}, and then merging those resulting * {@code Publisher}s and emitting the results of this merger, while limiting the maximum number of concurrent * subscriptions to these {@code Publisher}s. * * *
*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@code maxConcurrency} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the inner {@code Publisher}s and the output type * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @param delayErrors * if {@code true}, exceptions from the current {@code Flowable} and all inner {@code Publisher}s are delayed until all of them terminate * if {@code false}, the first one signaling an exception will terminate the whole sequence immediately * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable flatMap(@NonNull Function> mapper, boolean delayErrors, int maxConcurrency) { return flatMap(mapper, delayErrors, maxConcurrency, bufferSize()); } /** * Returns a {@code Flowable} that emits items based on applying a function that you supply to each item emitted * by the current {@code Flowable}, where that function returns a {@link Publisher}, and then merging those resulting * {@code Publisher}s and emitting the results of this merger, while limiting the maximum number of concurrent * subscriptions to these {@code Publisher}s. * * *
*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@code maxConcurrency} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the inner {@code Publisher}s and the output type * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @param delayErrors * if {@code true}, exceptions from the current {@code Flowable} and all inner {@code Publisher}s are delayed until all of them terminate * if {@code false}, the first one signaling an exception will terminate the whole sequence immediately * @param bufferSize * the number of elements to prefetch from each inner {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable flatMap(@NonNull Function> mapper, boolean delayErrors, int maxConcurrency, int bufferSize) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); if (this instanceof ScalarSupplier) { @SuppressWarnings("unchecked") T v = ((ScalarSupplier)this).get(); if (v == null) { return empty(); } return FlowableScalarXMap.scalarXMap(v, mapper); } return RxJavaPlugins.onAssembly(new FlowableFlatMap<>(this, mapper, delayErrors, maxConcurrency, bufferSize)); } /** * Returns a {@code Flowable} that applies a function to each item emitted or notification raised by the current * {@code Flowable} and then flattens the {@link Publisher}s returned from these functions and emits the resulting items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@link #bufferSize()} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the result type * @param onNextMapper * a function that returns a {@code Publisher} to merge for each item emitted by the current {@code Flowable} * @param onErrorMapper * a function that returns a {@code Publisher} to merge for an {@code onError} notification from the current * {@code Flowable} * @param onCompleteSupplier * a function that returns a {@code Publisher} to merge for an {@code onComplete} notification from the current * {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onNextMapper}, {@code onErrorMapper} or {@code onCompleteSupplier} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable flatMap( @NonNull Function> onNextMapper, @NonNull Function> onErrorMapper, @NonNull Supplier> onCompleteSupplier) { Objects.requireNonNull(onNextMapper, "onNextMapper is null"); Objects.requireNonNull(onErrorMapper, "onErrorMapper is null"); Objects.requireNonNull(onCompleteSupplier, "onCompleteSupplier is null"); return merge(new FlowableMapNotification<>(this, onNextMapper, onErrorMapper, onCompleteSupplier)); } /** * Returns a {@code Flowable} that applies a function to each item emitted or notification raised by the current * {@code Flowable} and then flattens the {@link Publisher}s returned from these functions and emits the resulting items, * while limiting the maximum number of concurrent subscriptions to these {@code Publisher}s. * * *
*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@code maxConcurrency} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the result type * @param onNextMapper * a function that returns a {@code Publisher} to merge for each item emitted by the current {@code Flowable} * @param onErrorMapper * a function that returns a {@code Publisher} to merge for an {@code onError} notification from the current * {@code Flowable} * @param onCompleteSupplier * a function that returns a {@code Publisher} to merge for an {@code onComplete} notification from the current * {@code Flowable} * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onNextMapper}, {@code onErrorMapper} or {@code onCompleteSupplier} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable flatMap( @NonNull Function> onNextMapper, @NonNull Function> onErrorMapper, @NonNull Supplier> onCompleteSupplier, int maxConcurrency) { Objects.requireNonNull(onNextMapper, "onNextMapper is null"); Objects.requireNonNull(onErrorMapper, "onErrorMapper is null"); Objects.requireNonNull(onCompleteSupplier, "onCompleteSupplier is null"); return merge(new FlowableMapNotification<>( this, onNextMapper, onErrorMapper, onCompleteSupplier), maxConcurrency); } /** * Returns a {@code Flowable} that emits the results of a specified function to the pair of values emitted by the * current {@code Flowable} and a specified collection {@link Publisher}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@code maxConcurrency} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the inner {@code Publisher}s * @param * the type of items emitted by the combiner function * @param mapper * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable} * @param combiner * a function that combines one item emitted by each of the source and collection {@code Publisher}s and * returns an item to be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U, @NonNull R> Flowable flatMap(@NonNull Function> mapper, @NonNull BiFunction combiner) { return flatMap(mapper, combiner, false, bufferSize(), bufferSize()); } /** * Returns a {@code Flowable} that emits the results of a specified function to the pair of values emitted by the * current {@code Flowable} and a specified inner {@link Publisher}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@link #bufferSize()} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the inner {@code Publisher}s * @param * the type of items emitted by the combiner functions * @param mapper * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable} * @param combiner * a function that combines one item emitted by each of the source and collection {@code Publisher}s and * returns an item to be emitted by the resulting {@code Flowable} * @param delayErrors * if {@code true}, exceptions from the current {@code Flowable} and all inner {@code Publisher}s are delayed until all of them terminate * if {@code false}, the first one signaling an exception will terminate the whole sequence immediately * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U, @NonNull R> Flowable flatMap(@NonNull Function> mapper, @NonNull BiFunction combiner, boolean delayErrors) { return flatMap(mapper, combiner, delayErrors, bufferSize(), bufferSize()); } /** * Returns a {@code Flowable} that emits the results of a specified function to the pair of values emitted by the * current {@code Flowable} and a specified collection {@link Publisher}, while limiting the maximum number of concurrent * subscriptions to these {@code Publisher}s. * * *
*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@code maxConcurrency} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the inner {@code Publisher}s * @param * the type of items emitted by the combiner function * @param mapper * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable} * @param combiner * a function that combines one item emitted by each of the source and collection {@code Publisher}s and * returns an item to be emitted by the resulting {@code Flowable} * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @param delayErrors * if {@code true}, exceptions from the current {@code Flowable} and all inner {@code Publisher}s are delayed until all of them terminate * if {@code false}, the first one signaling an exception will terminate the whole sequence immediately * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} or {@code combiner} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U, @NonNull R> Flowable flatMap(@NonNull Function> mapper, @NonNull BiFunction combiner, boolean delayErrors, int maxConcurrency) { return flatMap(mapper, combiner, delayErrors, maxConcurrency, bufferSize()); } /** * Returns a {@code Flowable} that emits the results of a specified function to the pair of values emitted by the * current {@code Flowable} and a specified collection {@link Publisher}, while limiting the maximum number of concurrent * subscriptions to these {@code Publisher}s. * * *
*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@code maxConcurrency} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the inner {@code Publisher}s * @param * the type of items emitted by the combiner function * @param mapper * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable} * @param combiner * a function that combines one item emitted by each of the source and collection {@code Publisher}s and * returns an item to be emitted by the resulting {@code Flowable} * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @param delayErrors * if {@code true}, exceptions from the current {@code Flowable} and all inner {@code Publisher}s are delayed until all of them terminate * if {@code false}, the first one signaling an exception will terminate the whole sequence immediately * @param bufferSize * the number of elements to prefetch from the inner {@code Publisher}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} or {@code combiner} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull R> Flowable flatMap(@NonNull Function> mapper, @NonNull BiFunction combiner, boolean delayErrors, int maxConcurrency, int bufferSize) { Objects.requireNonNull(mapper, "mapper is null"); Objects.requireNonNull(combiner, "combiner is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return flatMap(FlowableInternalHelper.flatMapWithCombiner(mapper, combiner), delayErrors, maxConcurrency, bufferSize); } /** * Returns a {@code Flowable} that emits the results of a specified function to the pair of values emitted by the * current {@code Flowable} and a specified collection {@link Publisher}, while limiting the maximum number of concurrent * subscriptions to these {@code Publisher}s. * * *
*
Backpressure:
*
The operator honors backpressure from downstream. The upstream {@code Flowable} is consumed * in a bounded manner (up to {@link #bufferSize()} outstanding request amount for items). * The inner {@code Publisher}s are expected to honor backpressure; if violated, * the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the inner {@code Publisher}s * @param * the type of items emitted by the combiner function * @param mapper * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable} * @param combiner * a function that combines one item emitted by each of the source and collection {@code Publisher}s and * returns an item to be emitted by the resulting {@code Flowable} * @param maxConcurrency * the maximum number of {@code Publisher}s that may be subscribed to concurrently * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} or {@code combiner} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U, @NonNull R> Flowable flatMap(@NonNull Function> mapper, @NonNull BiFunction combiner, int maxConcurrency) { return flatMap(mapper, combiner, false, maxConcurrency, bufferSize()); } /** * Maps each element of the upstream {@code Flowable} into {@link CompletableSource}s, subscribes to them and * waits until the upstream and all {@code CompletableSource}s complete. *
*
Backpressure:
*
The operator consumes the upstream in an unbounded manner.
*
Scheduler:
*
{@code flatMapCompletable} does not operate by default on a particular {@link Scheduler}.
*
* @param mapper the function that received each source value and transforms them into {@code CompletableSource}s. * @return the new {@link Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Completable flatMapCompletable(@NonNull Function mapper) { return flatMapCompletable(mapper, false, Integer.MAX_VALUE); } /** * Maps each element of the upstream {@code Flowable} into {@link CompletableSource}s, subscribes to them and * waits until the upstream and all {@code CompletableSource}s complete, optionally delaying all errors. *
*
Backpressure:
*
If {@code maxConcurrency == }{@link Integer#MAX_VALUE} the operator consumes the upstream in an unbounded manner. * Otherwise, the operator expects the upstream to honor backpressure. If the upstream doesn't support backpressure * the operator behaves as if {@code maxConcurrency == }{@link Integer#MAX_VALUE} was used.
*
Scheduler:
*
{@code flatMapCompletable} does not operate by default on a particular {@link Scheduler}.
*
* @param mapper the function that received each source value and transforms them into {@code CompletableSource}s. * @param delayErrors if {@code true}, errors from the upstream and inner {@code CompletableSource}s are delayed until each of them * terminates. * @param maxConcurrency the maximum number of active subscriptions to the {@code CompletableSource}s. * @return the new {@link Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Completable flatMapCompletable(@NonNull Function mapper, boolean delayErrors, int maxConcurrency) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); return RxJavaPlugins.onAssembly(new FlowableFlatMapCompletableCompletable<>(this, mapper, delayErrors, maxConcurrency)); } /** * Merges {@link Iterable}s generated by a mapper {@link Function} for each individual item emitted by * the current {@code Flowable} into a single {@code Flowable} sequence. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable}s is * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMapIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the output type and the element type of the {@code Iterable}s * @param mapper * a function that returns an {@code Iterable} sequence of values for when given an item emitted by the * current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U> Flowable flatMapIterable(@NonNull Function> mapper) { return flatMapIterable(mapper, bufferSize()); } /** * Merges {@link Iterable}s generated by a mapper {@link Function} for each individual item emitted by * the current {@code Flowable} into a single {@code Flowable} sequence. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable}s is * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMapIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of item emitted by the resulting {@code Iterable} * @param mapper * a function that returns an {@code Iterable} sequence of values for when given an item emitted by the * current {@code Flowable} * @param bufferSize * the number of elements to prefetch from the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable flatMapIterable(@NonNull Function> mapper, int bufferSize) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableFlattenIterable<>(this, mapper, bufferSize)); } /** * Merges {@link Iterable}s generated by a mapper {@link Function} for each individual item emitted by * the current {@code Flowable} into a single {@code Flowable} sequence where the resulting items will * be the combination of the original item and each inner item of the respective {@code Iterable} as returned * by the {@code resultSelector} {@link BiFunction}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and the current {@code Flowable}s is * consumed in a bounded manner (requesting {@link #bufferSize()} items upfront, then 75% of it after 75% received).
*
Scheduler:
*
{@code flatMapIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the element type of the {@code Iterable}s * @param * the output type as determined by the {@code resultSelector} function * @param mapper * a function that returns an {@code Iterable} sequence of values for each item emitted by the current * {@code Flowable} * @param combiner * a function that returns an item based on the item emitted by the current {@code Flowable} and the * {@code Iterable} returned for that item by the {@code collectionSelector} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} or {@code combiner} is {@code null} * @see ReactiveX operators documentation: FlatMap */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull V> Flowable flatMapIterable(@NonNull Function> mapper, @NonNull BiFunction combiner) { Objects.requireNonNull(mapper, "mapper is null"); Objects.requireNonNull(combiner, "combiner is null"); return flatMap(FlowableInternalHelper.flatMapIntoIterable(mapper), combiner, false, bufferSize(), bufferSize()); } /** * Merges {@link Iterable}s generated by a mapper {@link Function} for each individual item emitted by * the current {@code Flowable} into a single {@code Flowable} sequence where the resulting items will * be the combination of the original item and each inner item of the respective {@code Iterable} as returned * by the {@code resultSelector} {@link BiFunction}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable}s is * expected to honor backpressure as well. If the current {@code Flowable} violates the rule, the operator will * signal a {@link MissingBackpressureException}.
*
Scheduler:
*
{@code flatMapIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the element type of the inner {@code Iterable} sequences * @param * the type of item emitted by the resulting {@code Flowable} * @param mapper * a function that returns an {@code Iterable} sequence of values for when given an item emitted by the * current {@code Flowable} * @param combiner * a function that returns an item based on the item emitted by the current {@code Flowable} and the * {@code Iterable} returned for that item by the {@code collectionSelector} * @param prefetch * the number of elements to prefetch from the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} or {@code combiner} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see ReactiveX operators documentation: FlatMap * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull V> Flowable flatMapIterable(@NonNull Function> mapper, @NonNull BiFunction combiner, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); Objects.requireNonNull(combiner, "combiner is null"); return flatMap(FlowableInternalHelper.flatMapIntoIterable(mapper), combiner, false, bufferSize(), prefetch); } /** * Maps each element of the upstream {@code Flowable} into {@link MaybeSource}s, subscribes to all of them * and merges their {@code onSuccess} values, in no particular order, into a single {@code Flowable} sequence. *
*
Backpressure:
*
The operator consumes the upstream in an unbounded manner.
*
Scheduler:
*
{@code flatMapMaybe} does not operate by default on a particular {@link Scheduler}.
*
* @param the result value type * @param mapper the function that received each source value and transforms them into {@code MaybeSource}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable flatMapMaybe(@NonNull Function> mapper) { return flatMapMaybe(mapper, false, Integer.MAX_VALUE); } /** * Maps each element of the upstream {@code Flowable} into {@link MaybeSource}s, subscribes to at most * {@code maxConcurrency} {@code MaybeSource}s at a time and merges their {@code onSuccess} values, * in no particular order, into a single {@code Flowable} sequence, optionally delaying all errors. *
*
Backpressure:
*
If {@code maxConcurrency == }{@link Integer#MAX_VALUE} the operator consumes the upstream in an unbounded manner. * Otherwise, the operator expects the upstream to honor backpressure. If the upstream doesn't support backpressure * the operator behaves as if {@code maxConcurrency == }{@link Integer#MAX_VALUE} was used.
*
Scheduler:
*
{@code flatMapMaybe} does not operate by default on a particular {@link Scheduler}.
*
* @param the result value type * @param mapper the function that received each source value and transforms them into {@code MaybeSource}s. * @param delayErrors if {@code true}, errors from the upstream and inner {@code MaybeSource}s are delayed until each of them * terminates. * @param maxConcurrency the maximum number of active subscriptions to the {@code MaybeSource}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable flatMapMaybe(@NonNull Function> mapper, boolean delayErrors, int maxConcurrency) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); return RxJavaPlugins.onAssembly(new FlowableFlatMapMaybe<>(this, mapper, delayErrors, maxConcurrency)); } /** * Maps each element of the upstream {@code Flowable} into {@link SingleSource}s, subscribes to all of them * and merges their {@code onSuccess} values, in no particular order, into a single {@code Flowable} sequence. *
*
Backpressure:
*
The operator consumes the upstream in an unbounded manner.
*
Scheduler:
*
{@code flatMapSingle} does not operate by default on a particular {@link Scheduler}.
*
* @param the result value type * @param mapper the function that received each source value and transforms them into {@code SingleSource}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable flatMapSingle(@NonNull Function> mapper) { return flatMapSingle(mapper, false, Integer.MAX_VALUE); } /** * Maps each element of the upstream {@code Flowable} into {@link SingleSource}s, subscribes to at most * {@code maxConcurrency} {@code SingleSource}s at a time and merges their {@code onSuccess} values, * in no particular order, into a single {@code Flowable} sequence, optionally delaying all errors. *
*
Backpressure:
*
If {@code maxConcurrency == }{@link Integer#MAX_VALUE} the operator consumes the upstream in an unbounded manner. * Otherwise, the operator expects the upstream to honor backpressure. If the upstream doesn't support backpressure * the operator behaves as if {@code maxConcurrency == }{@link Integer#MAX_VALUE} was used.
*
Scheduler:
*
{@code flatMapSingle} does not operate by default on a particular {@link Scheduler}.
*
* @param the result value type * @param mapper the function that received each source value and transforms them into {@code SingleSource}s. * @param delayErrors if {@code true}, errors from the upstream and inner {@code SingleSources} are delayed until each of them * terminates. * @param maxConcurrency the maximum number of active subscriptions to the {@code SingleSource}s. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code maxConcurrency} is non-positive */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable flatMapSingle(@NonNull Function> mapper, boolean delayErrors, int maxConcurrency) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(maxConcurrency, "maxConcurrency"); return RxJavaPlugins.onAssembly(new FlowableFlatMapSingle<>(this, mapper, delayErrors, maxConcurrency)); } /** * Subscribes to the current {@code Flowable} and receives notifications for each element. *

* Alias to {@link #subscribe(Consumer)} *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code forEach} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext * {@link Consumer} to execute for each item. * @return * a {@link Disposable} that allows canceling an asynchronous sequence * @throws NullPointerException * if {@code onNext} is {@code null} * @see ReactiveX operators documentation: Subscribe */ @CheckReturnValue @BackpressureSupport(BackpressureKind.NONE) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Disposable forEach(@NonNull Consumer onNext) { return subscribe(onNext); } /** * Subscribes to the current {@code Flowable} and receives notifications for each element until the * {@code onNext} Predicate returns {@code false}. *

* If the {@code Flowable} emits an error, it is wrapped into an * {@link io.reactivex.rxjava3.exceptions.OnErrorNotImplementedException OnErrorNotImplementedException} * and routed to the {@link RxJavaPlugins#onError(Throwable)} handler. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code forEachWhile} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext * {@link Predicate} to execute for each item. * @return * a {@link Disposable} that allows canceling an asynchronous sequence * @throws NullPointerException * if {@code onNext} is {@code null} * @see ReactiveX operators documentation: Subscribe */ @CheckReturnValue @BackpressureSupport(BackpressureKind.NONE) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Disposable forEachWhile(@NonNull Predicate onNext) { return forEachWhile(onNext, Functions.ON_ERROR_MISSING, Functions.EMPTY_ACTION); } /** * Subscribes to the current {@code Flowable} and receives notifications for each element and error events until the * {@code onNext} Predicate returns {@code false}. *
*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code forEachWhile} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext * {@link Predicate} to execute for each item. * @param onError * {@link Consumer} to execute when an error is emitted. * @return * a {@link Disposable} that allows canceling an asynchronous sequence * @throws NullPointerException * if {@code onNext} or {@code onError} is {@code null} * @see ReactiveX operators documentation: Subscribe */ @CheckReturnValue @BackpressureSupport(BackpressureKind.NONE) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Disposable forEachWhile(@NonNull Predicate onNext, @NonNull Consumer onError) { return forEachWhile(onNext, onError, Functions.EMPTY_ACTION); } /** * Subscribes to the current {@code Flowable} and receives notifications for each element and the terminal events until the * {@code onNext} Predicate returns {@code false}. *
*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code forEachWhile} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext * {@link Predicate} to execute for each item. * @param onError * {@link Consumer} to execute when an error is emitted. * @param onComplete * {@link Action} to execute when completion is signaled. * @return * a {@link Disposable} that allows canceling an asynchronous sequence * @throws NullPointerException * if {@code onNext}, {@code onError} or {@code onComplete} is {@code null} * @see ReactiveX operators documentation: Subscribe */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.NONE) @SchedulerSupport(SchedulerSupport.NONE) public final Disposable forEachWhile(@NonNull Predicate onNext, @NonNull Consumer onError, @NonNull Action onComplete) { Objects.requireNonNull(onNext, "onNext is null"); Objects.requireNonNull(onError, "onError is null"); Objects.requireNonNull(onComplete, "onComplete is null"); ForEachWhileSubscriber s = new ForEachWhileSubscriber<>(onNext, onError, onComplete); subscribe(s); return s; } /** * Groups the items emitted by the current {@code Flowable} according to a specified criterion, and emits these * grouped items as {@link GroupedFlowable}s. The emitted {@code GroupedFlowable} allows only a single * {@link Subscriber} during its lifetime and if this {@code Subscriber} cancels before the * source terminates, the next emission by the source having the same key will trigger a new * {@code GroupedFlowable} emission. *

* *

* Note: A {@code GroupedFlowable} will cache the items it is to emit until such time as it * is subscribed to. For this reason, in order to avoid memory leaks, you should not simply ignore those * {@code GroupedFlowable}s that do not concern you. Instead, you can signal to them that they may * discard their buffers by applying an operator like {@link #ignoreElements} to them. *

* Note that the {@code GroupedFlowable}s should be subscribed to as soon as possible, otherwise, * the unconsumed groups may starve other groups due to the internal backpressure * coordination of the {@code groupBy} operator. Such hangs can be usually avoided by using * {@link #flatMap(Function, int)} or {@link #concatMapEager(Function, int, int)} and overriding the default maximum concurrency * value to be greater or equal to the expected number of groups, possibly using * {@link Integer#MAX_VALUE} if the number of expected groups is unknown. *

* Note also that ignoring groups or subscribing later (i.e., on another thread) will result in * so-called group abandonment where a group will only contain one element and the group will be * re-created over and over as new upstream items trigger a new group. The behavior is * a trade-off between no-dataloss, upstream cancellation and excessive group creation. * *

*
Backpressure:
*
The consumer of the returned {@code Flowable} has to be ready to receive new {@code GroupedFlowable}s or else * this operator will signal {@link MissingBackpressureException}. To avoid this exception, make * sure a combining operator (such as {@code flatMap}) has adequate amount of buffering/prefetch configured. * The inner {@code GroupedFlowable}s honor backpressure but due to the single-source multiple consumer * nature of this operator, each group must be consumed so the whole operator can make progress and not hang.
*
Scheduler:
*
{@code groupBy} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the upstream signals or the callback(s) throw an exception, the returned {@code Flowable} and * all active inner {@code GroupedFlowable}s will signal the same exception.
*
* * @param keySelector * a function that extracts the key for each item * @param * the key type * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector} is {@code null} * @see ReactiveX operators documentation: GroupBy * @see #groupBy(Function, boolean) * @see #groupBy(Function, Function) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K> Flowable> groupBy(@NonNull Function keySelector) { return groupBy(keySelector, Functions.identity(), false, bufferSize()); } /** * Groups the items emitted by the current {@code Flowable} according to a specified criterion, and emits these * grouped items as {@link GroupedFlowable}s. The emitted {@code GroupedFlowable} allows only a single * {@link Subscriber} during its lifetime and if this {@code Subscriber} cancels before the * source terminates, the next emission by the source having the same key will trigger a new * {@code GroupedFlowable} emission. *

* *

* Note: A {@code GroupedFlowable} will cache the items it is to emit until such time as it * is subscribed to. For this reason, in order to avoid memory leaks, you should not simply ignore those * {@code GroupedFlowable}s that do not concern you. Instead, you can signal to them that they may * discard their buffers by applying an operator like {@link #ignoreElements} to them. *

* Note that the {@code GroupedFlowable}s should be subscribed to as soon as possible, otherwise, * the unconsumed groups may starve other groups due to the internal backpressure * coordination of the {@code groupBy} operator. Such hangs can be usually avoided by using * {@link #flatMap(Function, int)} or {@link #concatMapEager(Function, int, int)} and overriding the default maximum concurrency * value to be greater or equal to the expected number of groups, possibly using * {@link Integer#MAX_VALUE} if the number of expected groups is unknown. *

* Note also that ignoring groups or subscribing later (i.e., on another thread) will result in * so-called group abandonment where a group will only contain one element and the group will be * re-created over and over as new upstream items trigger a new group. The behavior is * a trade-off between no-dataloss, upstream cancellation and excessive group creation. * *

*
Backpressure:
*
The consumer of the returned {@code Flowable} has to be ready to receive new {@code GroupedFlowable}s or else * this operator will signal {@link MissingBackpressureException}. To avoid this exception, make * sure a combining operator (such as {@code flatMap}) has adequate amount of buffering/prefetch configured. * The inner {@code GroupedFlowable}s honor backpressure but due to the single-source multiple consumer * nature of this operator, each group must be consumed so the whole operator can make progress and not hang.
*
Scheduler:
*
{@code groupBy} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the upstream signals or the callback(s) throw an exception, the returned {@code Flowable} and * all active inner {@code GroupedFlowable}s will signal the same exception.
*
* * @param keySelector * a function that extracts the key for each item * @param * the key type * @param delayError * if {@code true}, the exception from the current {@code Flowable} is delayed in each group until that specific group emitted * the normal values; if {@code false}, the exception bypasses values in the groups and is reported immediately. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector} is {@code null} * @see ReactiveX operators documentation: GroupBy */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K> Flowable> groupBy(@NonNull Function keySelector, boolean delayError) { return groupBy(keySelector, Functions.identity(), delayError, bufferSize()); } /** * Groups the items emitted by the current {@code Flowable} according to a specified criterion, and emits these * grouped items as {@link GroupedFlowable}s. The emitted {@code GroupedFlowable} allows only a single * {@link Subscriber} during its lifetime and if this {@code Subscriber} cancels before the * source terminates, the next emission by the source having the same key will trigger a new * {@code GroupedFlowable} emission. *

* *

* Note: A {@code GroupedFlowable} will cache the items it is to emit until such time as it * is subscribed to. For this reason, in order to avoid memory leaks, you should not simply ignore those * {@code GroupedFlowable}s that do not concern you. Instead, you can signal to them that they may * discard their buffers by applying an operator like {@link #ignoreElements} to them. *

* Note that the {@code GroupedFlowable}s should be subscribed to as soon as possible, otherwise, * the unconsumed groups may starve other groups due to the internal backpressure * coordination of the {@code groupBy} operator. Such hangs can be usually avoided by using * {@link #flatMap(Function, int)} or {@link #concatMapEager(Function, int, int)} and overriding the default maximum concurrency * value to be greater or equal to the expected number of groups, possibly using * {@link Integer#MAX_VALUE} if the number of expected groups is unknown. *

* Note also that ignoring groups or subscribing later (i.e., on another thread) will result in * so-called group abandonment where a group will only contain one element and the group will be * re-created over and over as new upstream items trigger a new group. The behavior is * a trade-off between no-dataloss, upstream cancellation and excessive group creation. * *

*
Backpressure:
*
The consumer of the returned {@code Flowable} has to be ready to receive new {@code GroupedFlowable}s or else * this operator will signal {@link MissingBackpressureException}. To avoid this exception, make * sure a combining operator (such as {@code flatMap}) has adequate amount of buffering/prefetch configured. * The inner {@code GroupedFlowable}s honor backpressure but due to the single-source multiple consumer * nature of this operator, each group must be consumed so the whole operator can make progress and not hang.
*
Scheduler:
*
{@code groupBy} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the upstream signals or the callback(s) throw an exception, the returned {@code Flowable} and * all active inner {@code GroupedFlowable}s will signal the same exception.
*
* * @param keySelector * a function that extracts the key for each item * @param valueSelector * a function that extracts the return element for each item * @param * the key type * @param * the element type * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector} or {@code valueSelector} is {@code null} * @see ReactiveX operators documentation: GroupBy * @see #groupBy(Function, Function, boolean) * @see #groupBy(Function, Function, boolean, int) * @see #groupBy(Function, Function, boolean, int, Function) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K, @NonNull V> Flowable> groupBy(@NonNull Function keySelector, @NonNull Function valueSelector) { return groupBy(keySelector, valueSelector, false, bufferSize()); } /** * Groups the items emitted by the current {@code Flowable} according to a specified criterion, and emits these * grouped items as {@link GroupedFlowable}s. The emitted {@code GroupedFlowable} allows only a single * {@link Subscriber} during its lifetime and if this {@code Subscriber} cancels before the * source terminates, the next emission by the source having the same key will trigger a new * {@code GroupedFlowable} emission. *

* *

* Note: A {@code GroupedFlowable} will cache the items it is to emit until such time as it * is subscribed to. For this reason, in order to avoid memory leaks, you should not simply ignore those * {@code GroupedFlowable}s that do not concern you. Instead, you can signal to them that they may * discard their buffers by applying an operator like {@link #ignoreElements} to them. *

* Note that the {@code GroupedFlowable}s should be subscribed to as soon as possible, otherwise, * the unconsumed groups may starve other groups due to the internal backpressure * coordination of the {@code groupBy} operator. Such hangs can be usually avoided by using * {@link #flatMap(Function, int)} or {@link #concatMapEager(Function, int, int)} and overriding the default maximum concurrency * value to be greater or equal to the expected number of groups, possibly using * {@link Integer#MAX_VALUE} if the number of expected groups is unknown. *

* Note also that ignoring groups or subscribing later (i.e., on another thread) will result in * so-called group abandonment where a group will only contain one element and the group will be * re-created over and over as new upstream items trigger a new group. The behavior is * a trade-off between no-dataloss, upstream cancellation and excessive group creation. * *

*
Backpressure:
*
The consumer of the returned {@code Flowable} has to be ready to receive new {@code GroupedFlowable}s or else * this operator will signal {@link MissingBackpressureException}. To avoid this exception, make * sure a combining operator (such as {@code flatMap}) has adequate amount of buffering/prefetch configured. * The inner {@code GroupedFlowable}s honor backpressure but due to the single-source multiple consumer * nature of this operator, each group must be consumed so the whole operator can make progress and not hang.
*
Scheduler:
*
{@code groupBy} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the upstream signals or the callback(s) throw an exception, the returned {@code Flowable} and * all active inner {@code GroupedFlowable}s will signal the same exception.
*
* * @param keySelector * a function that extracts the key for each item * @param valueSelector * a function that extracts the return element for each item * @param * the key type * @param * the element type * @param delayError * if {@code true}, the exception from the current {@code Flowable} is delayed in each group until that specific group emitted * the normal values; if {@code false}, the exception bypasses values in the groups and is reported immediately. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector} or {@code valueSelector} is {@code null} * @see ReactiveX operators documentation: GroupBy * @see #groupBy(Function, Function, boolean, int) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K, @NonNull V> Flowable> groupBy(@NonNull Function keySelector, @NonNull Function valueSelector, boolean delayError) { return groupBy(keySelector, valueSelector, delayError, bufferSize()); } /** * Groups the items emitted by the current {@code Flowable} according to a specified criterion, and emits these * grouped items as {@link GroupedFlowable}s. The emitted {@code GroupedFlowable} allows only a single * {@link Subscriber} during its lifetime and if this {@code Subscriber} cancels before the * source terminates, the next emission by the source having the same key will trigger a new * {@code GroupedFlowable} emission. *

* *

* Note: A {@code GroupedFlowable} will cache the items it is to emit until such time as it * is subscribed to. For this reason, in order to avoid memory leaks, you should not simply ignore those * {@code GroupedFlowable}s that do not concern you. Instead, you can signal to them that they may * discard their buffers by applying an operator like {@link #ignoreElements} to them. *

* Note that the {@code GroupedFlowable}s should be subscribed to as soon as possible, otherwise, * the unconsumed groups may starve other groups due to the internal backpressure * coordination of the {@code groupBy} operator. Such hangs can be usually avoided by using * {@link #flatMap(Function, int)} or {@link #concatMapEager(Function, int, int)} and overriding the default maximum concurrency * value to be greater or equal to the expected number of groups, possibly using * {@link Integer#MAX_VALUE} if the number of expected groups is unknown. *

* Note also that ignoring groups or subscribing later (i.e., on another thread) will result in * so-called group abandonment where a group will only contain one element and the group will be * re-created over and over as new upstream items trigger a new group. The behavior is * a trade-off between no-dataloss, upstream cancellation and excessive group creation. * *

*
Backpressure:
*
The consumer of the returned {@code Flowable} has to be ready to receive new {@code GroupedFlowable}s or else * this operator will signal {@link MissingBackpressureException}. To avoid this exception, make * sure a combining operator (such as {@code flatMap}) has adequate amount of buffering/prefetch configured. * The inner {@code GroupedFlowable}s honor backpressure but due to the single-source multiple consumer * nature of this operator, each group must be consumed so the whole operator can make progress and not hang.
*
Scheduler:
*
{@code groupBy} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the upstream signals or the callback(s) throw an exception, the returned {@code Flowable} and * all active inner {@code GroupedFlowable}s will signal the same exception.
*
* * @param keySelector * a function that extracts the key for each item * @param valueSelector * a function that extracts the return element for each item * @param delayError * if {@code true}, the exception from the current {@code Flowable} is delayed in each group until that specific group emitted * the normal values; if {@code false}, the exception bypasses values in the groups and is reported immediately. * @param bufferSize * the hint for how many {@code GroupedFlowable}s and element in each {@code GroupedFlowable} should be buffered * @param * the key type * @param * the element type * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector} or {@code valueSelector} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: GroupBy */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull K, @NonNull V> Flowable> groupBy(@NonNull Function keySelector, @NonNull Function valueSelector, boolean delayError, int bufferSize) { Objects.requireNonNull(keySelector, "keySelector is null"); Objects.requireNonNull(valueSelector, "valueSelector is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableGroupBy<>(this, keySelector, valueSelector, bufferSize, delayError, null)); } /** * Groups the items emitted by the current {@code Flowable} according to a specified criterion, and emits these * grouped items as {@link GroupedFlowable}s. The emitted {@code GroupedFlowable} allows only a single * {@link Subscriber} during its lifetime and if this {@code Subscriber} cancels before the * source terminates, the next emission by the source having the same key will trigger a new * {@code GroupedFlowable} emission. The {@code evictingMapFactory} is used to create a map that will * be used to hold the {@code GroupedFlowable}s by key. The evicting map created by this factory must * notify the provided {@code Consumer} with the entry value (not the key!) when an entry in this * map has been evicted. The next source emission will bring about the completion of the evicted * {@code GroupedFlowable}s and the arrival of an item with the same key as a completed {@code GroupedFlowable} * will prompt the creation and emission of a new {@code GroupedFlowable} with that key. * *

A use case for specifying an {@code evictingMapFactory} is where the source is infinite and fast and * over time the number of keys grows enough to be a concern in terms of the memory footprint of the * internal hash map containing the {@code GroupedFlowable}s. * *

The map created by an {@code evictingMapFactory} must be thread-safe. * *

An example of an {@code evictingMapFactory} using CacheBuilder from the Guava library is below: * *


     * Function<Consumer<Object>, Map<Integer, Object>> evictingMapFactory =
     *   notify ->
     *       CacheBuilder
     *         .newBuilder()
     *         .maximumSize(3)
     *         .removalListener(entry -> {
     *              try {
     *                  // emit the value not the key!
     *                  notify.accept(entry.getValue());
     *              } catch (Exception e) {
     *                  throw new RuntimeException(e);
     *              }
     *            })
     *         .<Integer, Object> build()
     *         .asMap();
     *
     * // Emit 1000 items but ensure that the
     * // internal map never has more than 3 items in it
     *   Flowable
     *   .range(1, 1000)
     *   // note that number of keys is 10
     *   .groupBy(x -> x % 10, x -> x, true, 16, evictingMapFactory)
     *   .flatMap(g -> g)
     *   .forEach(System.out::println);
     * 
* *

* *

* Note: A {@code GroupedFlowable} will cache the items it is to emit until such time as it * is subscribed to. For this reason, in order to avoid memory leaks, you should not simply ignore those * {@code GroupedFlowable}s that do not concern you. Instead, you can signal to them that they may * discard their buffers by applying an operator like {@link #ignoreElements} to them. *

* Note that the {@code GroupedFlowable}s should be subscribed to as soon as possible, otherwise, * the unconsumed groups may starve other groups due to the internal backpressure * coordination of the {@code groupBy} operator. Such hangs can be usually avoided by using * {@link #flatMap(Function, int)} or {@link #concatMapEager(Function, int, int)} and overriding the default maximum concurrency * value to be greater or equal to the expected number of groups, possibly using * {@link Integer#MAX_VALUE} if the number of expected groups is unknown. *

* Note also that ignoring groups or subscribing later (i.e., on another thread) will result in * so-called group abandonment where a group will only contain one element and the group will be * re-created over and over as new upstream items trigger a new group. The behavior is * a trade-off between no-dataloss, upstream cancellation and excessive group creation. * *

*
Backpressure:
*
The consumer of the returned {@code Flowable} has to be ready to receive new {@code GroupedFlowable}s or else * this operator will signal {@link MissingBackpressureException}. To avoid this exception, make * sure a combining operator (such as {@code flatMap}) has adequate amount of buffering/prefetch configured. * The inner {@code GroupedFlowable}s honor backpressure but due to the single-source multiple consumer * nature of this operator, each group must be consumed so the whole operator can make progress and not hang.
*
Scheduler:
*
{@code groupBy} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If the upstream signals or the callback(s) throw an exception, the returned {@code Flowable} and * all active inner {@code GroupedFlowable}s will signal the same exception.
*
*

History: 2.1.10 - beta * @param keySelector * a function that extracts the key for each item * @param valueSelector * a function that extracts the return element for each item * @param delayError * if {@code true}, the exception from the current {@code Flowable} is delayed in each group until that specific group emitted * the normal values; if {@code false}, the exception bypasses values in the groups and is reported immediately. * @param bufferSize * the hint for how many {@code GroupedFlowable}s and element in each {@code GroupedFlowable} should be buffered * @param evictingMapFactory * The factory used to create a map that will be used by the implementation to hold the * {@code GroupedFlowable}s. The evicting map created by this factory must * notify the provided {@code Consumer} with the entry value (not the key!) when * an entry in this map has been evicted. The next source emission will bring about the * completion of the evicted {@code GroupedFlowable}s. See example above. * @param * the key type * @param * the element type * @return the new {@code Flowable} instance * @throws NullPointerException if {@code keySelector}, {@code valueSelector} or {@code evictingMapFactory} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: GroupBy * * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull K, @NonNull V> Flowable> groupBy(@NonNull Function keySelector, @NonNull Function valueSelector, boolean delayError, int bufferSize, @NonNull Function, ? extends Map> evictingMapFactory) { Objects.requireNonNull(keySelector, "keySelector is null"); Objects.requireNonNull(valueSelector, "valueSelector is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); Objects.requireNonNull(evictingMapFactory, "evictingMapFactory is null"); return RxJavaPlugins.onAssembly(new FlowableGroupBy<>(this, keySelector, valueSelector, bufferSize, delayError, evictingMapFactory)); } /** * Returns a {@code Flowable} that correlates two {@link Publisher}s when they overlap in time and groups the results. *

* There are no guarantees in what order the items get combined when multiple * items from one or both source {@code Publisher}s overlap. *

* *

*
Backpressure:
*
The operator doesn't support backpressure and consumes all participating {@code Publisher}s in * an unbounded mode (i.e., not applying any backpressure to them).
*
Scheduler:
*
{@code groupJoin} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the right {@code Publisher} source * @param the element type of the left duration {@code Publisher}s * @param the element type of the right duration {@code Publisher}s * @param the result type * @param other * the other {@code Publisher} to correlate items from the current {@code Flowable} with * @param leftEnd * a function that returns a {@code Publisher} whose emissions indicate the duration of the values of * the current {@code Flowable} * @param rightEnd * a function that returns a {@code Publisher} whose emissions indicate the duration of the values of * the {@code right} {@code Publisher} * @param resultSelector * a function that takes an item emitted by each {@code Publisher} and returns the value to be emitted * by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other}, {@code leftEnd}, {@code rightEnd} or {@code resultSelector} is {@code null} * @see ReactiveX operators documentation: Join */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull TRight, @NonNull TLeftEnd, @NonNull TRightEnd, @NonNull R> Flowable groupJoin( @NonNull Publisher other, @NonNull Function> leftEnd, @NonNull Function> rightEnd, @NonNull BiFunction, ? extends R> resultSelector) { Objects.requireNonNull(other, "other is null"); Objects.requireNonNull(leftEnd, "leftEnd is null"); Objects.requireNonNull(rightEnd, "rightEnd is null"); Objects.requireNonNull(resultSelector, "resultSelector is null"); return RxJavaPlugins.onAssembly(new FlowableGroupJoin<>( this, other, leftEnd, rightEnd, resultSelector)); } /** * Hides the identity of this {@code Flowable} and its {@link Subscription}. *

Allows hiding extra features such as {@link Processor}'s * {@link Subscriber} methods or preventing certain identity-based * optimizations (fusion). *

*
Backpressure:
*
The operator is a pass-through for backpressure, the behavior is determined by the upstream's * backpressure behavior.
*
Scheduler:
*
{@code hide} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code Flowable} instance * * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable hide() { return RxJavaPlugins.onAssembly(new FlowableHide<>(this)); } /** * Ignores all items emitted by the current {@code Flowable} and only calls {@code onComplete} or {@code onError}. *

* *

*
Backpressure:
*
This operator ignores backpressure as it doesn't emit any elements and consumes the current {@code Flowable} * in an unbounded manner (i.e., no backpressure is applied to it).
*
Scheduler:
*
{@code ignoreElements} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@link Completable} instance * @see ReactiveX operators documentation: IgnoreElements */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Completable ignoreElements() { return RxJavaPlugins.onAssembly(new FlowableIgnoreElementsCompletable<>(this)); } /** * Returns a {@link Single} that emits {@code true} if the current {@code Flowable} is empty, otherwise {@code false}. *

* In Rx.Net this is negated as the {@code any} {@link Subscriber} but we renamed this in RxJava to better match Java * naming idioms. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code isEmpty} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Single} instance * @see ReactiveX operators documentation: Contains */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single isEmpty() { return all(Functions.alwaysFalse()); } /** * Correlates the items emitted by two {@link Publisher}s based on overlapping durations. *

* There are no guarantees in what order the items get combined when multiple * items from one or both source {@code Publisher}s overlap. *

* *

*
Backpressure:
*
The operator doesn't support backpressure and consumes all participating {@code Publisher}s in * an unbounded mode (i.e., not applying any backpressure to them).
*
Scheduler:
*
{@code join} does not operate by default on a particular {@link Scheduler}.
*
* * @param the value type of the right {@code Publisher} source * @param the element type of the left duration {@code Publisher}s * @param the element type of the right duration {@code Publisher}s * @param the result type * @param other * the second {@code Publisher} to join items from * @param leftEnd * a function to select a duration for each item emitted by the current {@code Flowable}, used to * determine overlap * @param rightEnd * a function to select a duration for each item emitted by the {@code right} {@code Publisher}, used to * determine overlap * @param resultSelector * a function that computes an item to be emitted by the resulting {@code Flowable} for any two * overlapping items emitted by the two {@code Publisher}s * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other}, {@code leftEnd}, {@code rightEnd} or {@code resultSelector} is {@code null} * @see ReactiveX operators documentation: Join */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull TRight, @NonNull TLeftEnd, @NonNull TRightEnd, @NonNull R> Flowable join( @NonNull Publisher other, @NonNull Function> leftEnd, @NonNull Function> rightEnd, @NonNull BiFunction resultSelector) { Objects.requireNonNull(other, "other is null"); Objects.requireNonNull(leftEnd, "leftEnd is null"); Objects.requireNonNull(rightEnd, "rightEnd is null"); Objects.requireNonNull(resultSelector, "resultSelector is null"); return RxJavaPlugins.onAssembly(new FlowableJoin( this, other, leftEnd, rightEnd, resultSelector)); } /** * Returns a {@link Maybe} that emits the last item emitted by this {@code Flowable} or completes if * this {@code Flowable} is empty. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code lastElement} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Maybe} instance * @see ReactiveX operators documentation: Last */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Maybe lastElement() { return RxJavaPlugins.onAssembly(new FlowableLastMaybe<>(this)); } /** * Returns a {@link Single} that emits only the last item emitted by this {@code Flowable}, or a default item * if this {@code Flowable} completes without emitting any items. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code last} does not operate by default on a particular {@link Scheduler}.
*
* * @param defaultItem * the default item to emit if the current {@code Flowable} is empty * @return the new {@code Single} instance * @throws NullPointerException if {@code defaultItem} is {@code null} * @see ReactiveX operators documentation: Last */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Single last(@NonNull T defaultItem) { Objects.requireNonNull(defaultItem, "defaultItem is null"); return RxJavaPlugins.onAssembly(new FlowableLastSingle<>(this, defaultItem)); } /** * Returns a {@link Single} that emits only the last item emitted by this {@code Flowable} or signals * a {@link NoSuchElementException} if this {@code Flowable} is empty. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code lastOrError} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Single} instance * @see ReactiveX operators documentation: Last */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single lastOrError() { return RxJavaPlugins.onAssembly(new FlowableLastSingle<>(this, null)); } /** * This method requires advanced knowledge about building operators, please consider * other standard composition methods first; * Returns a {@code Flowable} which, when subscribed to, invokes the {@link FlowableOperator#apply(Subscriber) apply(Subscriber)} method * of the provided {@link FlowableOperator} for each individual downstream {@link Subscriber} and allows the * insertion of a custom operator by accessing the downstream's {@code Subscriber} during this subscription phase * and providing a new {@code Subscriber}, containing the custom operator's intended business logic, that will be * used in the subscription process going further upstream. *

* Generally, such a new {@code Subscriber} will wrap the downstream's {@code Subscriber} and forwards the * {@code onNext}, {@code onError} and {@code onComplete} events from the upstream directly or according to the * emission pattern the custom operator's business logic requires. In addition, such operator can intercept the * flow control calls of {@code cancel} and {@code request} that would have traveled upstream and perform * additional actions depending on the same business logic requirements. *

* Example: *


     * // Step 1: Create the consumer type that will be returned by the FlowableOperator.apply():
     *
     * public final class CustomSubscriber<T> implements FlowableSubscriber<T>, Subscription {
     *
     *     // The downstream's Subscriber that will receive the onXXX events
     *     final Subscriber<? super String> downstream;
     *
     *     // The connection to the upstream source that will call this class' onXXX methods
     *     Subscription upstream;
     *
     *     // The constructor takes the downstream subscriber and usually any other parameters
     *     public CustomSubscriber(Subscriber<? super String> downstream) {
     *         this.downstream = downstream;
     *     }
     *
     *     // In the subscription phase, the upstream sends a Subscription to this class
     *     // and subsequently this class has to send a Subscription to the downstream.
     *     // Note that relaying the upstream's Subscription instance directly is not allowed in RxJava
     *     @Override
     *     public void onSubscribe(Subscription s) {
     *         if (upstream != null) {
     *             s.cancel();
     *         } else {
     *             upstream = s;
     *             downstream.onSubscribe(this);
     *         }
     *     }
     *
     *     // The upstream calls this with the next item and the implementation's
     *     // responsibility is to emit an item to the downstream based on the intended
     *     // business logic, or if it can't do so for the particular item,
     *     // request more from the upstream
     *     @Override
     *     public void onNext(T item) {
     *         String str = item.toString();
     *         if (str.length() < 2) {
     *             downstream.onNext(str);
     *         } else {
     *             upstream.request(1);
     *         }
     *     }
     *
     *     // Some operators may handle the upstream's error while others
     *     // could just forward it to the downstream.
     *     @Override
     *     public void onError(Throwable throwable) {
     *         downstream.onError(throwable);
     *     }
     *
     *     // When the upstream completes, usually the downstream should complete as well.
     *     @Override
     *     public void onComplete() {
     *         downstream.onComplete();
     *     }
     *
     *     // Some operators have to intercept the downstream's request calls to trigger
     *     // the emission of queued items while others can simply forward the request
     *     // amount as is.
     *     @Override
     *     public void request(long n) {
     *         upstream.request(n);
     *     }
     *
     *     // Some operators may use their own resources which should be cleaned up if
     *     // the downstream cancels the flow before it completed. Operators without
     *     // resources can simply forward the cancellation to the upstream.
     *     // In some cases, a canceled flag may be set by this method so that other parts
     *     // of this class may detect the cancellation and stop sending events
     *     // to the downstream.
     *     @Override
     *     public void cancel() {
     *         upstream.cancel();
     *     }
     * }
     *
     * // Step 2: Create a class that implements the FlowableOperator interface and
     * //         returns the custom consumer type from above in its apply() method.
     * //         Such class may define additional parameters to be submitted to
     * //         the custom consumer type.
     *
     * final class CustomOperator<T> implements FlowableOperator<String> {
     *     @Override
     *     public Subscriber<? super String> apply(Subscriber<? super T> upstream) {
     *         return new CustomSubscriber<T>(upstream);
     *     }
     * }
     *
     * // Step 3: Apply the custom operator via lift() in a flow by creating an instance of it
     * //         or reusing an existing one.
     *
     * Flowable.range(5, 10)
     * .lift(new CustomOperator<Integer>())
     * .test()
     * .assertResult("5", "6", "7", "8", "9");
     * 
*

* Creating custom operators can be complicated and it is recommended one consults the * RxJava wiki: Writing operators page about * the tools, requirements, rules, considerations and pitfalls of implementing them. *

* Note that implementing custom operators via this {@code lift()} method adds slightly more overhead by requiring * an additional allocation and indirection per assembled flows. Instead, extending the abstract {@code Flowable} * class and creating a {@link FlowableTransformer} with it is recommended. *

* Note also that it is not possible to stop the subscription phase in {@code lift()} as the {@code apply()} method * requires a non-{@code null} {@code Subscriber} instance to be returned, which is then unconditionally subscribed to * the upstream {@code Flowable}. For example, if the operator decided there is no reason to subscribe to the * upstream source because of some optimization possibility or a failure to prepare the operator, it still has to * return a {@code Subscriber} that should immediately cancel the upstream's {@link Subscription} in its * {@code onSubscribe} method. Again, using a {@code FlowableTransformer} and extending the {@code Flowable} is * a better option as {@link #subscribeActual} can decide to not subscribe to its upstream after all. *

*
Backpressure:
*
The {@code Subscriber} instance returned by the {@code FlowableOperator} is responsible to be * backpressure-aware or document the fact that the consumer of the returned {@link Publisher} has to apply one of * the {@code onBackpressureXXX} operators.
*
Scheduler:
*
{@code lift} does not operate by default on a particular {@link Scheduler}, however, the * {@code FlowableOperator} may use a {@code Scheduler} to support its own asynchronous behavior.
*
* * @param the output value type * @param lifter the {@code FlowableOperator} that receives the downstream's {@code Subscriber} and should return * a {@code Subscriber} with custom behavior to be used as the consumer for the current * {@code Flowable}. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code lifter} is {@code null} * @see RxJava wiki: Writing operators * @see #compose(FlowableTransformer) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable lift(@NonNull FlowableOperator lifter) { Objects.requireNonNull(lifter, "lifter is null"); return RxJavaPlugins.onAssembly(new FlowableLift<>(this, lifter)); } /** * Returns a {@code Flowable} that applies a specified function to each item emitted by the current {@code Flowable} and * emits the results of these function applications. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code map} does not operate by default on a particular {@link Scheduler}.
*
* * @param the output type * @param mapper * a function to apply to each item emitted by the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see ReactiveX operators documentation: Map * @see #mapOptional(Function) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable map(@NonNull Function mapper) { Objects.requireNonNull(mapper, "mapper is null"); return RxJavaPlugins.onAssembly(new FlowableMap<>(this, mapper)); } /** * Returns a {@code Flowable} that represents all of the emissions and notifications from the current * {@code Flowable} into emissions marked with their original types within {@link Notification} objects. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and expects it from the current {@code Flowable}. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code materialize} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: Materialize * @see #dematerialize(Function) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> materialize() { return RxJavaPlugins.onAssembly(new FlowableMaterialize<>(this)); } /** * Flattens this and another {@link Publisher} into a single {@code Publisher}, without any transformation. *

* *

* You can combine items emitted by multiple {@code Publisher}s so that they appear as a single {@code Publisher}, by * using the {@code mergeWith} method. *

*
Backpressure:
*
The operator honors backpressure from downstream. This and the other {@code Publisher}s are expected to honor * backpressure; if violated, the operator may signal {@link MissingBackpressureException}.
*
Scheduler:
*
{@code mergeWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param other * a {@code Publisher} to be merged * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @see ReactiveX operators documentation: Merge */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable mergeWith(@NonNull Publisher other) { Objects.requireNonNull(other, "other is null"); return merge(this, other); } /** * Merges the sequence of items of this {@code Flowable} with the success value of the other {@link SingleSource}. *

* *

* The success value of the other {@code SingleSource} can get interleaved at any point of this * {@code Flowable} sequence. *

*
Backpressure:
*
The operator honors backpressure from downstream and ensures the success item from the * {@code SingleSource} is emitted only when there is a downstream demand.
*
Scheduler:
*
{@code mergeWith} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.10 - experimental * @param other the {@code SingleSource} whose success value to merge with * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable mergeWith(@NonNull SingleSource other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableMergeWithSingle<>(this, other)); } /** * Merges the sequence of items of this {@code Flowable} with the success value of the other {@link MaybeSource} * or waits for both to complete normally if the {@code MaybeSource} is empty. *

* *

* The success value of the other {@code MaybeSource} can get interleaved at any point of this * {@code Flowable} sequence. *

*
Backpressure:
*
The operator honors backpressure from downstream and ensures the success item from the * {@code MaybeSource} is emitted only when there is a downstream demand.
*
Scheduler:
*
{@code mergeWith} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.10 - experimental * @param other the {@code MaybeSource} which provides a success value to merge with or completes * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable mergeWith(@NonNull MaybeSource other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableMergeWithMaybe<>(this, other)); } /** * Relays the items of this {@code Flowable} and completes only when the other {@link CompletableSource} completes * as well. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code mergeWith} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.10 - experimental * @param other the {@code CompletableSource} to await for completion * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable mergeWith(@NonNull CompletableSource other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableMergeWithCompletable<>(this, other)); } /** * Signals the items and terminal signals of the current {@code Flowable} on the specified {@link Scheduler}, * asynchronously with a bounded buffer of {@link #bufferSize()} slots. * *

Note that {@code onError} notifications will cut ahead of {@code onNext} notifications on the emission thread if {@code Scheduler} is truly * asynchronous. If strict event ordering is required, consider using the {@link #observeOn(Scheduler, boolean)} overload. *

* *

* This operator keeps emitting as many signals as it can on the given {@code Scheduler}'s Worker thread, * which may result in a longer than expected occupation of this thread. In other terms, * it does not allow per-signal fairness in case the worker runs on a shared underlying thread. * If such fairness and signal/work interleaving is preferred, use the delay operator with zero time instead. *

*
Backpressure:
*
This operator honors backpressure from downstream and expects it from the current {@code Flowable}. Violating this * expectation will lead to {@link MissingBackpressureException}. This is the most common operator where the exception * pops up; look for sources up the chain that don't support backpressure, * such as {@link #interval(long, TimeUnit)}, {@link #timer(long, TimeUnit)}, * {@link io.reactivex.rxjava3.processors.PublishProcessor PublishProcessor} or * {@link io.reactivex.rxjava3.processors.BehaviorProcessor BehaviorProcessor} and apply any * of the {@code onBackpressureXXX} operators before applying {@code observeOn} itself. * Note also that request amounts are not preserved between the immediate downstream and the * immediate upstream. The operator always requests the default {@link #bufferSize()} amount first, then after * every 75% of that amount delivered, another 75% of this default value. If preserving the request amounts * is to be preferred over potential excess scheduler infrastructure use, consider applying * {@link #delay(long, TimeUnit, Scheduler)} with zero time instead. *
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param scheduler * the {@code Scheduler} to notify {@link Subscriber}s on * @return the new {@code Flowable} instance * @throws NullPointerException if {@code scheduler} is {@code null} * @see ReactiveX operators documentation: ObserveOn * @see RxJava Threading Examples * @see #subscribeOn * @see #observeOn(Scheduler, boolean) * @see #observeOn(Scheduler, boolean, int) * @see #delay(long, TimeUnit, Scheduler) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable observeOn(@NonNull Scheduler scheduler) { return observeOn(scheduler, false, bufferSize()); } /** * Signals the items and terminal signals of the current {@code Flowable} on the specified {@link Scheduler}, * asynchronously with a bounded buffer and optionally delays {@code onError} notifications. *

* *

* This operator keeps emitting as many signals as it can on the given {@code Scheduler}'s Worker thread, * which may result in a longer than expected occupation of this thread. In other terms, * it does not allow per-signal fairness in case the worker runs on a shared underlying thread. * If such fairness and signal/work interleaving is preferred, use the delay operator with zero time instead. *

*
Backpressure:
*
This operator honors backpressure from downstream and expects it from the current {@code Flowable}. Violating this * expectation will lead to {@link MissingBackpressureException}. This is the most common operator where the exception * pops up; look for sources up the chain that don't support backpressure, * such as {@link #interval(long, TimeUnit)}, {@link #timer(long, TimeUnit)}, * {@link io.reactivex.rxjava3.processors.PublishProcessor PublishProcessor} or * {@link io.reactivex.rxjava3.processors.BehaviorProcessor BehaviorProcessor} and apply any * of the {@code onBackpressureXXX} operators before applying {@code observeOn} itself. * Note also that request amounts are not preserved between the immediate downstream and the * immediate upstream. The operator always requests the default {@link #bufferSize()} amount first, then after * every 75% of that amount delivered, another 75% of this default value. If preserving the request amounts * is to be preferred over potential excess scheduler infrastructure use, consider applying * {@link #delay(long, TimeUnit, Scheduler, boolean)} with zero time instead. *
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param scheduler * the {@code Scheduler} to notify {@link Subscriber}s on * @param delayError * indicates if the {@code onError} notification may not cut ahead of {@code onNext} notification on the other side of the * scheduling boundary. If {@code true}, a sequence ending in {@code onError} will be replayed in the same order as was received * from upstream * @return the new {@code Flowable} instance * @throws NullPointerException if {@code scheduler} is {@code null} * @see ReactiveX operators documentation: ObserveOn * @see RxJava Threading Examples * @see #subscribeOn * @see #observeOn(Scheduler) * @see #observeOn(Scheduler, boolean, int) * @see #delay(long, TimeUnit, Scheduler, boolean) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable observeOn(@NonNull Scheduler scheduler, boolean delayError) { return observeOn(scheduler, delayError, bufferSize()); } /** * Signals the items and terminal signals of the current {@code Flowable} on the specified {@link Scheduler}, * asynchronously with a bounded buffer of configurable size and optionally delays {@code onError} notifications. *

* *

* This operator keeps emitting as many signals as it can on the given {@code Scheduler}'s Worker thread, * which may result in a longer than expected occupation of this thread. In other terms, * it does not allow per-signal fairness in case the worker runs on a shared underlying thread. * If such fairness and signal/work interleaving is preferred, use the delay operator with zero time instead. *

*
Backpressure:
*
This operator honors backpressure from downstream and expects it from the current {@code Flowable}. Violating this * expectation will lead to {@link MissingBackpressureException}. This is the most common operator where the exception * pops up; look for sources up the chain that don't support backpressure, * such as {@link #interval(long, TimeUnit)}, {@link #timer(long, TimeUnit)}, * {@link io.reactivex.rxjava3.processors.PublishProcessor PublishProcessor} or * {@link io.reactivex.rxjava3.processors.BehaviorProcessor BehaviorProcessor} and apply any * of the {@code onBackpressureXXX} operators before applying {@code observeOn} itself. * Note also that request amounts are not preserved between the immediate downstream and the * immediate upstream. The operator always requests the specified {@code bufferSize} amount first, then after * every 75% of that amount delivered, another 75% of this specified value. If preserving the request amounts * is to be preferred over potential excess scheduler infrastructure use, consider applying * {@link #delay(long, TimeUnit, Scheduler, boolean)} with zero time instead. *
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param scheduler * the {@code Scheduler} to notify {@link Subscriber}s on * @param delayError * indicates if the {@code onError} notification may not cut ahead of {@code onNext} notification on the other side of the * scheduling boundary. If {@code true}, a sequence ending in {@code onError} will be replayed in the same order as was received * from upstream * @param bufferSize the size of the buffer. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: ObserveOn * @see RxJava Threading Examples * @see #subscribeOn * @see #observeOn(Scheduler) * @see #observeOn(Scheduler, boolean) * @see #delay(long, TimeUnit, Scheduler, boolean) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable observeOn(@NonNull Scheduler scheduler, boolean delayError, int bufferSize) { Objects.requireNonNull(scheduler, "scheduler is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableObserveOn<>(this, scheduler, delayError, bufferSize)); } /** * Filters the items emitted by the current {@code Flowable}, only emitting those of the specified type. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code ofType} does not operate by default on a particular {@link Scheduler}.
*
* * @param the output type * @param clazz * the class type to filter the items emitted by the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code clazz} is {@code null} * @see ReactiveX operators documentation: Filter */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable ofType(@NonNull Class clazz) { Objects.requireNonNull(clazz, "clazz is null"); return filter(Functions.isInstanceOf(clazz)).cast(clazz); } /** * Buffers an unlimited number of items from the current {@code Flowable} and allows it to emit as fast it can while allowing the * downstream to consume the items at its own place. *

* *

* An error from the current {@code Flowable} will cut ahead of any unconsumed item. Use {@link #onBackpressureBuffer(boolean)} * to have the operator keep the original signal order. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureBuffer} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: backpressure operators * @see #onBackpressureBuffer(boolean) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureBuffer() { return onBackpressureBuffer(bufferSize(), false, true); } /** * Buffers an unlimited number of items from the current {@code Flowable} and allows it to emit as fast it can while allowing the * downstream to consume the items at its own place, optionally delaying an error until all buffered items have been consumed. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureBuffer} does not operate by default on a particular {@link Scheduler}.
*
* @param delayError * if {@code true}, an exception from the current {@code Flowable} is delayed until all buffered elements have been * consumed by the downstream; if {@code false}, an exception is immediately signaled to the downstream, skipping * any buffered element * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: backpressure operators */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureBuffer(boolean delayError) { return onBackpressureBuffer(bufferSize(), delayError, true); } /** * Buffers an limited number of items from the current {@code Flowable} and allows it to emit as fast it can while allowing the * downstream to consume the items at its own place, however, the resulting {@code Flowable} will signal a * {@link MissingBackpressureException} via {@code onError} as soon as the buffer's capacity is exceeded, dropping all undelivered * items, and canceling the flow. *

* *

* An error from the current {@code Flowable} will cut ahead of any unconsumed item. Use {@link #onBackpressureBuffer(int, boolean)} * to have the operator keep the original signal order. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureBuffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param capacity number of slots available in the buffer. * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code capacity} is non-positive * @see ReactiveX operators documentation: backpressure operators * @since 1.1.0 * @see #onBackpressureBuffer(long, Action, BackpressureOverflowStrategy) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureBuffer(int capacity) { return onBackpressureBuffer(capacity, false, false); } /** * Buffers an limited number of items from the current {@code Flowable} and allows it to emit as fast it can while allowing the * downstream to consume the items at its own place, however, the resulting {@code Flowable} will signal a * {@link MissingBackpressureException} via {@code onError} as soon as the buffer's capacity is exceeded, dropping all undelivered * items, and canceling the flow. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureBuffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param capacity number of slots available in the buffer. * @param delayError * if {@code true}, an exception from the current {@code Flowable} is delayed until all buffered elements have been * consumed by the downstream; if {@code false}, an exception is immediately signaled to the downstream, skipping * any buffered element * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code capacity} is non-positive * @see ReactiveX operators documentation: backpressure operators * @since 1.1.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureBuffer(int capacity, boolean delayError) { return onBackpressureBuffer(capacity, delayError, false); } /** * Buffers an optionally unlimited number of items from the current {@code Flowable} and allows it to emit as fast it can while allowing the * downstream to consume the items at its own place. * If {@code unbounded} is {@code true}, the resulting {@code Flowable} will signal a * {@link MissingBackpressureException} via {@code onError} as soon as the buffer's capacity is exceeded, dropping all undelivered * items, and canceling the flow. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureBuffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param capacity number of slots available in the buffer. * @param delayError * if {@code true}, an exception from the current {@code Flowable} is delayed until all buffered elements have been * consumed by the downstream; if {@code false}, an exception is immediately signaled to the downstream, skipping * any buffered element * @param unbounded * if {@code true}, the capacity value is interpreted as the internal "island" size of the unbounded buffer * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code capacity} is non-positive * @see ReactiveX operators documentation: backpressure operators * @since 1.1.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureBuffer(int capacity, boolean delayError, boolean unbounded) { ObjectHelper.verifyPositive(capacity, "capacity"); return RxJavaPlugins.onAssembly(new FlowableOnBackpressureBuffer<>(this, capacity, unbounded, delayError, Functions.EMPTY_ACTION)); } /** * Buffers an optionally unlimited number of items from the current {@code Flowable} and allows it to emit as fast it can while allowing the * downstream to consume the items at its own place. * If {@code unbounded} is {@code true}, the resulting {@code Flowable} will signal a * {@link MissingBackpressureException} via {@code onError} as soon as the buffer's capacity is exceeded, dropping all undelivered * items, canceling the flow and calling the {@code onOverflow} action. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureBuffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param capacity number of slots available in the buffer. * @param delayError * if {@code true}, an exception from the current {@code Flowable} is delayed until all buffered elements have been * consumed by the downstream; if {@code false}, an exception is immediately signaled to the downstream, skipping * any buffered element * @param unbounded * if {@code true}, the capacity value is interpreted as the internal "island" size of the unbounded buffer * @param onOverflow action to execute if an item needs to be buffered, but there are no available slots. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onOverflow} is {@code null} * @throws IllegalArgumentException if {@code capacity} is non-positive * @see ReactiveX operators documentation: backpressure operators * @since 1.1.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable onBackpressureBuffer(int capacity, boolean delayError, boolean unbounded, @NonNull Action onOverflow) { Objects.requireNonNull(onOverflow, "onOverflow is null"); ObjectHelper.verifyPositive(capacity, "capacity"); return RxJavaPlugins.onAssembly(new FlowableOnBackpressureBuffer<>(this, capacity, unbounded, delayError, onOverflow)); } /** * Buffers an limited number of items from the current {@code Flowable} and allows it to emit as fast it can while allowing the * downstream to consume the items at its own place, however, the resulting {@code Flowable} will signal a * {@link MissingBackpressureException} via {@code onError} as soon as the buffer's capacity is exceeded, dropping all undelivered * items, canceling the flow and calling the {@code onOverflow} action. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureBuffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param capacity number of slots available in the buffer. * @param onOverflow action to execute if an item needs to be buffered, but there are no available slots. Null is allowed. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onOverflow} is {@code null} * @throws IllegalArgumentException if {@code capacity} is non-positive * @see ReactiveX operators documentation: backpressure operators * @since 1.1.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureBuffer(int capacity, @NonNull Action onOverflow) { return onBackpressureBuffer(capacity, false, false, onOverflow); } /** * Buffers an optionally unlimited number of items from the current {@code Flowable} and allows it to emit as fast it can while allowing the * downstream to consume the items at its own place. * The resulting {@code Flowable} will behave as determined by {@code overflowStrategy} if the buffer capacity is exceeded: *
    *
  • {@link BackpressureOverflowStrategy#ERROR} (default) will call {@code onError} dropping all undelivered items, * canceling the source, and notifying the producer with {@code onOverflow}.
  • *
  • {@link BackpressureOverflowStrategy#DROP_LATEST} will drop any new items emitted by the producer while * the buffer is full, without generating any {@code onError}. Each drop will, however, invoke {@code onOverflow} * to signal the overflow to the producer.
  • *
  • {@link BackpressureOverflowStrategy#DROP_OLDEST} will drop the oldest items in the buffer in order to make * room for newly emitted ones. Overflow will not generate an {@code onError}, but each drop will invoke * {@code onOverflow} to signal the overflow to the producer.
  • *
* *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureBuffer} does not operate by default on a particular {@link Scheduler}.
*
* * @param capacity number of slots available in the buffer. * @param onOverflow action to execute if an item needs to be buffered, but there are no available slots, {@code null} is allowed. * @param overflowStrategy how should the resulting {@code Flowable} react to buffer overflows, {@code null} is not allowed. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onOverflow} or {@code overflowStrategy} is {@code null} * @throws IllegalArgumentException if {@code capacity} is non-positive * @see ReactiveX operators documentation: backpressure operators * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable onBackpressureBuffer(long capacity, @Nullable Action onOverflow, @NonNull BackpressureOverflowStrategy overflowStrategy) { Objects.requireNonNull(overflowStrategy, "overflowStrategy is null"); ObjectHelper.verifyPositive(capacity, "capacity"); return RxJavaPlugins.onAssembly(new FlowableOnBackpressureBufferStrategy<>(this, capacity, onOverflow, overflowStrategy)); } /** * Drops items from the current {@code Flowable} if the downstream is not ready to receive new items (indicated * by a lack of {@link Subscription#request(long)} calls from it). *

* *

* If the downstream request count hits 0 then the resulting {@code Flowable} will refrain from calling {@code onNext} until * the {@link Subscriber} invokes {@code request(n)} again to increase the request count. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureDrop} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: backpressure operators */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureDrop() { return RxJavaPlugins.onAssembly(new FlowableOnBackpressureDrop<>(this)); } /** * Drops items from the current {@code Flowable} if the downstream is not ready to receive new items (indicated * by a lack of {@link Subscription#request(long)} calls from it) and calls the given {@link Consumer} with such * dropped items. *

* *

* If the downstream request count hits 0 then the resulting {@code Flowable} will refrain from calling {@code onNext} until * the {@link Subscriber} invokes {@code request(n)} again to increase the request count. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureDrop} does not operate by default on a particular {@link Scheduler}.
*
* * @param onDrop the action to invoke for each item dropped, should be fast and should never block. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code onDrop} is {@code null} * @see ReactiveX operators documentation: backpressure operators * @since 1.1.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable onBackpressureDrop(@NonNull Consumer onDrop) { Objects.requireNonNull(onDrop, "onDrop is null"); return RxJavaPlugins.onAssembly(new FlowableOnBackpressureDrop<>(this, onDrop)); } /** * Drops all but the latest item emitted by the current {@code Flowable} if the downstream is not ready to receive * new items (indicated by a lack of {@link Subscription#request(long)} calls from it) and emits this latest * item when the downstream becomes ready. *

* *

* Its behavior is logically equivalent to {@code blockingLatest()} with the exception that * the downstream is not blocking while requesting more values. *

* Note that if the current {@code Flowable} does support backpressure, this operator ignores that capability * and doesn't propagate any backpressure requests from downstream. *

* Note that due to the nature of how backpressure requests are propagated through subscribeOn/observeOn, * requesting more than 1 from downstream doesn't guarantee a continuous delivery of {@code onNext} events. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureLatest} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @since 1.1.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureLatest() { return RxJavaPlugins.onAssembly(new FlowableOnBackpressureLatest<>(this)); } /** * Reduces a sequence of two not emitted values via a function into a single value if the downstream is not ready to receive * new items (indicated by a lack of {@link Subscription#request(long)} calls from it) and emits this latest * item when the downstream becomes ready. *

* *

* Note that if the current {@code Flowable} does support backpressure, this operator ignores that capability * and doesn't propagate any backpressure requests from downstream. *

* Note that due to the nature of how backpressure requests are propagated through subscribeOn/observeOn, * requesting more than 1 from downstream doesn't guarantee a continuous delivery of {@code onNext} events. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureReduce} does not operate by default on a particular {@link Scheduler}.
*
*

History: 3.0.9 - experimental * @param reducer the bi-function to call when there is more than one non-emitted value to downstream, * the first argument of the bi-function is previous item and the second one is currently * emitting from upstream * @return the new {@code Flowable} instance * @throws NullPointerException if {@code reducer} is {@code null} * @since 3.1.0 * @see #onBackpressureReduce(Supplier, BiFunction) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onBackpressureReduce(@NonNull BiFunction reducer) { Objects.requireNonNull(reducer, "reducer is null"); return RxJavaPlugins.onAssembly(new FlowableOnBackpressureReduce<>(this, reducer)); } /** * Reduces upstream values into an aggregate value, provided by a supplier and combined via a reducer function, * while the downstream is not ready to receive items, then emits this aggregate value when the downstream becomes ready. *

* *

* Note that even if the downstream is ready to receive an item, the upstream item will always be aggregated into the output type, * calling both the supplier and the reducer to produce the output value. *

* Note that if the current {@code Flowable} does support backpressure, this operator ignores that capability * and doesn't propagate any backpressure requests from downstream. *

* Note that due to the nature of how backpressure requests are propagated through subscribeOn/observeOn, * requesting more than 1 from downstream doesn't guarantee a continuous delivery of {@code onNext} events. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an unbounded * manner (i.e., not applying backpressure to it).
*
Scheduler:
*
{@code onBackpressureReduce} does not operate by default on a particular {@link Scheduler}.
*
*

History: 3.0.9 - experimental * @param the aggregate type emitted when the downstream requests more items * @param supplier the factory to call to create new item of type R to pass it as the first argument to {@code reducer}. * It is called when previous returned value by {@code reducer} already sent to * downstream or the very first update from upstream received. * @param reducer the bi-function to call to reduce excessive updates which downstream is not ready to receive. * The first argument of type R is the object returned by {@code supplier} or result of previous * {@code reducer} invocation. The second argument of type T is the current update from upstream. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code supplier} or {@code reducer} is {@code null} * @see #onBackpressureReduce(BiFunction) * @since 3.1.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable onBackpressureReduce(@NonNull Supplier supplier, @NonNull BiFunction reducer) { Objects.requireNonNull(supplier, "supplier is null"); Objects.requireNonNull(reducer, "reducer is null"); return RxJavaPlugins.onAssembly(new FlowableOnBackpressureReduceWith<>(this, supplier, reducer)); } /** * Returns a {@code Flowable} instance that if the current {@code Flowable} emits an error, it will emit an {@code onComplete} * and swallow the throwable. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code onErrorComplete} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code Flowable} instance * @since 3.0.0 */ @CheckReturnValue @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.PASS_THROUGH) @NonNull public final Flowable onErrorComplete() { return onErrorComplete(Functions.alwaysTrue()); } /** * Returns a {@code Flowable} instance that if the current {@code Flowable} emits an error and the predicate returns * {@code true}, it will emit an {@code onComplete} and swallow the throwable. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code onErrorComplete} does not operate by default on a particular {@link Scheduler}.
*
* @param predicate the predicate to call when an {@link Throwable} is emitted which should return {@code true} * if the {@code Throwable} should be swallowed and replaced with an {@code onComplete}. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code predicate} is {@code null} * @since 3.0.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onErrorComplete(@NonNull Predicate predicate) { Objects.requireNonNull(predicate, "predicate is null"); return RxJavaPlugins.onAssembly(new FlowableOnErrorComplete<>(this, predicate)); } /** * Resumes the flow with a {@link Publisher} returned for the failure {@link Throwable} of the current {@code Flowable} by a * function instead of signaling the error via {@code onError}. *

* *

* By default, when a {@code Publisher} encounters an error that prevents it from emitting the expected item to * its {@link Subscriber}, the {@code Publisher} invokes its {@code Subscriber}'s {@code onError} method, and then quits * without invoking any more of its {@code Subscriber}'s methods. The {@code onErrorResumeNext} method changes this * behavior. If you pass a function that returns a {@code Publisher} ({@code resumeFunction}) to * {@code onErrorResumeNext}, if the original {@code Publisher} encounters an error, instead of invoking its * {@code Subscriber}'s {@code onError} method, it will instead relinquish control to the {@code Publisher} returned from * {@code resumeFunction}, which will invoke the {@code Subscriber}'s {@link Subscriber#onNext onNext} method if it is * able to do so. In such a case, because no {@code Publisher} necessarily invokes {@code onError}, the {@code Subscriber} * may never know that an error happened. *

* You can use this to prevent errors from propagating or to supply fallback data should errors be * encountered. *

*
Backpressure:
*
The operator honors backpressure from downstream. This and the resuming {@code Publisher}s * are expected to honor backpressure as well. * If any of them violate this expectation, the operator may throw an * {@link IllegalStateException} when the current {@code Flowable} completes or * a {@link MissingBackpressureException} is signaled somewhere downstream.
*
Scheduler:
*
{@code onErrorResumeNext} does not operate by default on a particular {@link Scheduler}.
*
* * @param fallbackSupplier * a function that returns a {@code Publisher} that will take over if the current {@code Flowable} encounters * an error * @return the new {@code Flowable} instance * @throws NullPointerException if {@code fallbackSupplier} is {@code null} * @see ReactiveX operators documentation: Catch */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable onErrorResumeNext(@NonNull Function> fallbackSupplier) { Objects.requireNonNull(fallbackSupplier, "fallbackSupplier is null"); return RxJavaPlugins.onAssembly(new FlowableOnErrorNext<>(this, fallbackSupplier)); } /** * Resumes the flow with the given {@link Publisher} when the current {@code Flowable} fails instead of * signaling the error via {@code onError}. *

* *

* By default, when a {@code Publisher} encounters an error that prevents it from emitting the expected item to * its {@link Subscriber}, the {@code Publisher} invokes its {@code Subscriber}'s {@code onError} method, and then quits * without invoking any more of its {@code Subscriber}'s methods. The {@code onErrorResumeWith} method changes this * behavior. If you pass another {@code Publisher} ({@code resumeSequence}) to a {@code Publisher}'s * {@code onErrorResumeWith} method, if the original {@code Publisher} encounters an error, instead of invoking its * {@code Subscriber}'s {@code onError} method, it will instead relinquish control to {@code resumeSequence} which * will invoke the {@code Subscriber}'s {@link Subscriber#onNext onNext} method if it is able to do so. In such a case, * because no {@code Publisher} necessarily invokes {@code onError}, the {@code Subscriber} may never know that an error * happened. *

* You can use this to prevent errors from propagating or to supply fallback data should errors be * encountered. *

*
Backpressure:
*
The operator honors backpressure from downstream. This and the resuming {@code Publisher}s * are expected to honor backpressure as well. * If any of them violate this expectation, the operator may throw an * {@link IllegalStateException} when the current {@code Flowable} completes or * {@link MissingBackpressureException} is signaled somewhere downstream.
*
Scheduler:
*
{@code onErrorResumeWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param fallback * the next {@code Publisher} source that will take over if the current {@code Flowable} encounters * an error * @return the new {@code Flowable} instance * @throws NullPointerException if {@code fallback} is {@code null} * @see ReactiveX operators documentation: Catch */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable onErrorResumeWith(@NonNull Publisher fallback) { Objects.requireNonNull(fallback, "fallback is null"); return onErrorResumeNext(Functions.justFunction(fallback)); } /** * Ends the flow with a last item returned by a function for the {@link Throwable} error signaled by the current * {@code Flowable} instead of signaling the error via {@code onError}. *

* *

* By default, when a {@link Publisher} encounters an error that prevents it from emitting the expected item to * its {@link Subscriber}, the {@code Publisher} invokes its {@code Subscriber}'s {@code onError} method, and then quits * without invoking any more of its {@code Subscriber}'s methods. The {@code onErrorReturn} method changes this * behavior. If you pass a function ({@code resumeFunction}) to a {@code Publisher}'s {@code onErrorReturn} * method, if the original {@code Publisher} encounters an error, instead of invoking its {@code Subscriber}'s * {@code onError} method, it will instead emit the return value of {@code resumeFunction}. *

* You can use this to prevent errors from propagating or to supply fallback data should errors be * encountered. *

*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable} is expected to honor * backpressure as well. If it this expectation is violated, the operator may throw * {@link IllegalStateException} when the current {@code Flowable} completes or * {@link MissingBackpressureException} is signaled somewhere downstream.
*
Scheduler:
*
{@code onErrorReturn} does not operate by default on a particular {@link Scheduler}.
*
* * @param itemSupplier * a function that returns a single value that will be emitted along with a regular {@code onComplete} in case * the current {@code Flowable} signals an {@code onError} event * @return the new {@code Flowable} instance * @throws NullPointerException if {@code itemSupplier} is {@code null} * @see ReactiveX operators documentation: Catch */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable onErrorReturn(@NonNull Function itemSupplier) { Objects.requireNonNull(itemSupplier, "itemSupplier is null"); return RxJavaPlugins.onAssembly(new FlowableOnErrorReturn<>(this, itemSupplier)); } /** * Ends the flow with the given last item when the current {@code Flowable} fails instead of signaling the error via {@code onError}. *

* *

* By default, when a {@link Publisher} encounters an error that prevents it from emitting the expected item to * its {@link Subscriber}, the {@code Publisher} invokes its {@code Subscriber}'s {@code onError} method, and then quits * without invoking any more of its {@code Subscriber}'s methods. The {@code onErrorReturn} method changes this * behavior. If you pass a function ({@code resumeFunction}) to a {@code Publisher}'s {@code onErrorReturn} * method, if the original {@code Publisher} encounters an error, instead of invoking its {@code Subscriber}'s * {@code onError} method, it will instead emit the return value of {@code resumeFunction}. *

* You can use this to prevent errors from propagating or to supply fallback data should errors be * encountered. *

*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable} is expected to honor * backpressure as well. If it this expectation is violated, the operator may throw * {@link IllegalStateException} when the current {@code Flowable} completes or * {@link MissingBackpressureException} is signaled somewhere downstream.
*
Scheduler:
*
{@code onErrorReturnItem} does not operate by default on a particular {@link Scheduler}.
*
* * @param item * the value that is emitted along with a regular {@code onComplete} in case the current * {@code Flowable} signals an exception * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item} is {@code null} * @see ReactiveX operators documentation: Catch */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable onErrorReturnItem(@NonNull T item) { Objects.requireNonNull(item, "item is null"); return onErrorReturn(Functions.justFunction(item)); } /** * Nulls out references to the upstream producer and downstream {@link Subscriber} if * the sequence is terminated or downstream cancels. *
*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code onTerminateDetach} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code Flowable} instance * the sequence is terminated or downstream cancels * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable onTerminateDetach() { return RxJavaPlugins.onAssembly(new FlowableDetach<>(this)); } /** * Parallelizes the flow by creating multiple 'rails' (equal to the number of CPUs) * and dispatches the upstream items to them in a round-robin fashion. *

* Note that the rails don't execute in parallel on their own and one needs to * apply {@link ParallelFlowable#runOn(Scheduler)} to specify the {@link Scheduler} where * each rail will execute. *

* To merge the parallel 'rails' back into a single sequence, use {@link ParallelFlowable#sequential()}. *

* *

*
Backpressure:
*
The operator requires the upstream to honor backpressure and each 'rail' honors backpressure * as well.
*
Scheduler:
*
{@code parallel} does not operate by default on a particular {@code Scheduler}.
*
*

History: 2.0.5 - experimental; 2.1 - beta * @return the new {@link ParallelFlowable} instance * @since 2.2 */ @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @NonNull public final ParallelFlowable parallel() { return ParallelFlowable.from(this); } /** * Parallelizes the flow by creating the specified number of 'rails' * and dispatches the upstream items to them in a round-robin fashion. *

* Note that the rails don't execute in parallel on their own and one needs to * apply {@link ParallelFlowable#runOn(Scheduler)} to specify the {@link Scheduler} where * each rail will execute. *

* To merge the parallel 'rails' back into a single sequence, use {@link ParallelFlowable#sequential()}. *

* *

*
Backpressure:
*
The operator requires the upstream to honor backpressure and each 'rail' honors backpressure * as well.
*
Scheduler:
*
{@code parallel} does not operate by default on a particular {@code Scheduler}.
*
*

History: 2.0.5 - experimental; 2.1 - beta * @param parallelism the number of 'rails' to use * @return the new {@link ParallelFlowable} instance * @throws IllegalArgumentException if {@code parallelism} is non-positive * @since 2.2 */ @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @NonNull public final ParallelFlowable parallel(int parallelism) { return ParallelFlowable.from(this, parallelism); } /** * Parallelizes the flow by creating the specified number of 'rails' * and dispatches the upstream items to them in a round-robin fashion and * uses the defined per-'rail' prefetch amount. *

* Note that the rails don't execute in parallel on their own and one needs to * apply {@link ParallelFlowable#runOn(Scheduler)} to specify the {@link Scheduler} where * each rail will execute. *

* To merge the parallel 'rails' back into a single sequence, use {@link ParallelFlowable#sequential()}. *

* *

*
Backpressure:
*
The operator requires the upstream to honor backpressure and each 'rail' honors backpressure * as well.
*
Scheduler:
*
{@code parallel} does not operate by default on a particular {@code Scheduler}.
*
*

History: 2.0.5 - experimental; 2.1 - beta * @param parallelism the number of 'rails' to use * @param prefetch the number of items each 'rail' should prefetch * @return the new {@link ParallelFlowable} instance * @throws IllegalArgumentException if {@code parallelism} or {@code prefetch} is non-positive * @since 2.2 */ @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @CheckReturnValue @NonNull public final ParallelFlowable parallel(int parallelism, int prefetch) { return ParallelFlowable.from(this, parallelism, prefetch); } /** * Returns a {@link ConnectableFlowable}, which is a variety of {@link Publisher} that waits until its * {@link ConnectableFlowable#connect connect} method is called before it begins emitting items to those * {@link Subscriber}s that have subscribed to it. *

* *

*
Backpressure:
*
The returned {@code ConnectableFlowable} honors backpressure for each of its {@code Subscriber}s * and expects the current {@code Flowable} to honor backpressure as well. If this expectation is violated, * the operator will signal a {@link MissingBackpressureException} to its {@code Subscriber}s and disconnect.
*
Scheduler:
*
{@code publish} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code ConnectableFlowable} instance * @see ReactiveX operators documentation: Publish */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final ConnectableFlowable publish() { return publish(bufferSize()); } /** * Returns a {@code Flowable} that emits the results of invoking a specified selector on items emitted by a * {@link ConnectableFlowable} that shares a single subscription to the underlying sequence. *

* *

*
Backpressure:
*
The operator expects the current {@code Flowable} to honor backpressure and if this expectation is * violated, the operator will signal a {@link MissingBackpressureException} through the {@code Flowable} * provided to the function. Since the {@link Publisher} returned by the {@code selector} may be * independent of the provided {@code Flowable} to the function, the output's backpressure behavior * is determined by this returned {@code Publisher}.
*
Scheduler:
*
{@code publish} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * a function that can use the multicasted source sequence as many times as needed, without * causing multiple subscriptions to the source sequence. {@link Subscriber}s to the given source will * receive all notifications of the source from the time of the subscription forward. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector} is {@code null} * @see ReactiveX operators documentation: Publish */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable publish(@NonNull Function, @NonNull ? extends Publisher> selector) { return publish(selector, bufferSize()); } /** * Returns a {@code Flowable} that emits the results of invoking a specified selector on items emitted by a * {@link ConnectableFlowable} that shares a single subscription to the underlying sequence. *

* *

*
Backpressure:
*
The operator expects the current {@code Flowable} to honor backpressure and if this expectation is * violated, the operator will signal a {@link MissingBackpressureException} through the {@code Flowable} * provided to the function. Since the {@link Publisher} returned by the {@code selector} may be * independent of the provided {@code Flowable} to the function, the output's backpressure behavior * is determined by this returned {@code Publisher}.
*
Scheduler:
*
{@code publish} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * a function that can use the multicasted source sequence as many times as needed, without * causing multiple subscriptions to the source sequence. {@link Subscriber}s to the given source will * receive all notifications of the source from the time of the subscription forward. * @param prefetch * the number of elements to prefetch from the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @see ReactiveX operators documentation: Publish */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable publish(@NonNull Function, @NonNull ? extends Publisher> selector, int prefetch) { Objects.requireNonNull(selector, "selector is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowablePublishMulticast<>(this, selector, prefetch, false)); } /** * Returns a {@link ConnectableFlowable}, which is a variety of {@link Publisher} that waits until its * {@link ConnectableFlowable#connect connect} method is called before it begins emitting items to those * {@link Subscriber}s that have subscribed to it. *

* *

*
Backpressure:
*
The returned {@code ConnectableFlowable} honors backpressure for each of its {@code Subscriber}s * and expects the current {@code Flowable} to honor backpressure as well. If this expectation is violated, * the operator will signal a {@link MissingBackpressureException} to its {@code Subscriber}s and disconnect.
*
Scheduler:
*
{@code publish} does not operate by default on a particular {@link Scheduler}.
*
* * @param bufferSize * the number of elements to prefetch from the current {@code Flowable} * @return the new {@code ConnectableFlowable} instance * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Publish */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final ConnectableFlowable publish(int bufferSize) { ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowablePublish<>(this, bufferSize)); } /** * Requests {@code n} initially from the upstream and then 75% of {@code n} subsequently * after 75% of {@code n} values have been emitted to the downstream. * *

This operator allows preventing the downstream to trigger unbounded mode via {@code request(}{@link Long#MAX_VALUE}{@code )} * or compensate for the per-item overhead of small and frequent requests. * *

*
Backpressure:
*
The operator expects backpressure from upstream and honors backpressure from downstream.
*
Scheduler:
*
{@code rebatchRequests} does not operate by default on a particular {@link Scheduler}.
*
* * @param n the initial request amount, further request will happen after 75% of this value * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code n} is non-positive * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable rebatchRequests(int n) { return observeOn(ImmediateThinScheduler.INSTANCE, true, n); } /** * Returns a {@link Maybe} that applies a specified accumulator function to the first item emitted by the current * {@code Flowable}, then feeds the result of that function along with the second item emitted by the current * {@code Flowable} into the same function, and so on until all items have been emitted by the current and finite {@code Flowable}, * and emits the final result from the final call to your function as its sole item. *

* *

* This technique, which is called "reduce" here, is sometimes called "aggregate," "fold," "accumulate," * "compress," or "inject" in other programming contexts. Groovy, for instance, has an {@code inject} method * that does a similar operation on lists. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulator object to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure of its downstream consumer and consumes the * upstream source in unbounded mode.
*
Scheduler:
*
{@code reduce} does not operate by default on a particular {@link Scheduler}.
*
* * @param reducer * an accumulator function to be invoked on each item emitted by the current {@code Flowable}, whose * result will be used in the next accumulator call * @return the new {@code Maybe} instance * @throws NullPointerException if {@code reducer} is {@code null} * @see ReactiveX operators documentation: Reduce * @see Wikipedia: Fold (higher-order function) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Maybe reduce(@NonNull BiFunction reducer) { Objects.requireNonNull(reducer, "reducer is null"); return RxJavaPlugins.onAssembly(new FlowableReduceMaybe<>(this, reducer)); } /** * Returns a {@link Single} that applies a specified accumulator function to the first item emitted by the current * {@code Flowable} and a specified seed value, then feeds the result of that function along with the second item * emitted by the current {@code Flowable} into the same function, and so on until all items have been emitted by the * current and finite {@code Flowable}, emitting the final result from the final call to your function as its sole item. *

* *

* This technique, which is called "reduce" here, is sometimes called "aggregate," "fold," "accumulate," * "compress," or "inject" in other programming contexts. Groovy, for instance, has an {@code inject} method * that does a similar operation on lists. *

* Note that the {@code seed} is shared among all subscribers to the resulting {@code Flowable} * and may cause problems if it is mutable. To make sure each subscriber gets its own value, defer * the application of this operator via {@link #defer(Supplier)}: *


     * Flowable<T> source = ...
     * Single.defer(() -> source.reduce(new ArrayList<>(), (list, item) -> list.add(item)));
     *
     * // alternatively, by using compose to stay fluent
     *
     * source.compose(o ->
     *     Flowable.defer(() -> o.reduce(new ArrayList<>(), (list, item) -> list.add(item)).toFlowable())
     * ).firstOrError();
     *
     * // or, by using reduceWith instead of reduce
     *
     * source.reduceWith(() -> new ArrayList<>(), (list, item) -> list.add(item)));
     * 
*

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulator object to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure of its downstream consumer and consumes the * upstream source in unbounded mode.
*
Scheduler:
*
{@code reduce} does not operate by default on a particular {@link Scheduler}.
*
* * @param the accumulator and output value type * @param seed * the initial (seed) accumulator value * @param reducer * an accumulator function to be invoked on each item emitted by the current {@code Flowable}, the * result of which will be used in the next accumulator call * @return the new {@code Single} instance * @throws NullPointerException if {@code seed} or {@code reducer} is {@code null} * @see ReactiveX operators documentation: Reduce * @see Wikipedia: Fold (higher-order function) * @see #reduceWith(Supplier, BiFunction) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Single reduce(R seed, @NonNull BiFunction reducer) { Objects.requireNonNull(seed, "seed is null"); Objects.requireNonNull(reducer, "reducer is null"); return RxJavaPlugins.onAssembly(new FlowableReduceSeedSingle<>(this, seed, reducer)); } /** * Returns a {@link Single} that applies a specified accumulator function to the first item emitted by the current * {@code Flowable} and a seed value derived from calling a specified {@code seedSupplier}, then feeds the result * of that function along with the second item emitted by the current {@code Flowable} into the same function, and so on until * all items have been emitted by the current and finite {@code Flowable}, emitting the final result from the final call to your * function as its sole item. *

* *

* This technique, which is called "reduce" here, is sometimes called "aggregate", "fold", "accumulate", * "compress", or "inject" in other programming contexts. Groovy, for instance, has an {@code inject} method * that does a similar operation on lists. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulator object to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure of its downstream consumer and consumes the * upstream source in unbounded mode.
*
Scheduler:
*
{@code reduceWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param the accumulator and output value type * @param seedSupplier * the {@link Supplier} that provides the initial (seed) accumulator value for each individual {@link Subscriber} * @param reducer * an accumulator function to be invoked on each item emitted by the current {@code Flowable}, the * result of which will be used in the next accumulator call * @return the new {@code Single} instance * @throws NullPointerException if {@code seedSupplier} or {@code reducer} is {@code null} * @see ReactiveX operators documentation: Reduce * @see Wikipedia: Fold (higher-order function) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Single reduceWith(@NonNull Supplier seedSupplier, @NonNull BiFunction reducer) { Objects.requireNonNull(seedSupplier, "seedSupplier is null"); Objects.requireNonNull(reducer, "reducer is null"); return RxJavaPlugins.onAssembly(new FlowableReduceWithSingle<>(this, seedSupplier, reducer)); } /** * Returns a {@code Flowable} that repeats the sequence of items emitted by the current {@code Flowable} indefinitely. *

* *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code repeat} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: Repeat */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable repeat() { return repeat(Long.MAX_VALUE); } /** * Returns a {@code Flowable} that repeats the sequence of items emitted by the current {@code Flowable} at most * {@code count} times. *

* *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code repeat} does not operate by default on a particular {@link Scheduler}.
*
* * @param times * the number of times the current {@code Flowable} items are repeated, a count of 0 will yield an empty * sequence * @return the new {@code Flowable} instance * @throws IllegalArgumentException * if {@code times} is less than zero * @see ReactiveX operators documentation: Repeat */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable repeat(long times) { if (times < 0) { throw new IllegalArgumentException("times >= 0 required but it was " + times); } if (times == 0) { return empty(); } return RxJavaPlugins.onAssembly(new FlowableRepeat<>(this, times)); } /** * Returns a {@code Flowable} that repeats the sequence of items emitted by the current {@code Flowable} until * the provided stop function returns {@code true}. *

* *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code repeatUntil} does not operate by default on a particular {@link Scheduler}.
*
* * @param stop * a boolean supplier that is called when the current {@code Flowable} completes and unless it returns * {@code false}, the current {@code Flowable} is resubscribed * @return the new {@code Flowable} instance * @throws NullPointerException * if {@code stop} is {@code null} * @see ReactiveX operators documentation: Repeat */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable repeatUntil(@NonNull BooleanSupplier stop) { Objects.requireNonNull(stop, "stop is null"); return RxJavaPlugins.onAssembly(new FlowableRepeatUntil<>(this, stop)); } /** * Returns a {@code Flowable} that emits the same values as the current {@code Flowable} with the exception of an * {@code onComplete}. An {@code onComplete} notification from the source will result in the emission of * a {@code void} item to the {@code Flowable} provided as an argument to the {@code notificationHandler} * function. If that {@link Publisher} calls {@code onComplete} or {@code onError} then {@code repeatWhen} will * call {@code onComplete} or {@code onError} on the child subscription. Otherwise, this {@code Publisher} will * resubscribe to the current {@code Flowable}. *

* *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code repeatWhen} does not operate by default on a particular {@link Scheduler}.
*
* * @param handler * receives a {@code Publisher} of notifications with which a user can complete or error, aborting the repeat. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code handler} is {@code null} * @see ReactiveX operators documentation: Repeat */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable repeatWhen(@NonNull Function, @NonNull ? extends Publisher<@NonNull ?>> handler) { Objects.requireNonNull(handler, "handler is null"); return RxJavaPlugins.onAssembly(new FlowableRepeatWhen<>(this, handler)); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the underlying {@link Publisher} * that will replay all of its items and notifications to any future {@link Subscriber}. A connectable * {@code Flowable} resembles an ordinary {@code Flowable}, except that it does not begin emitting items when it is * subscribed to, but only when its {@code connect} method is called. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@code Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code ConnectableFlowable} instance * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final ConnectableFlowable replay() { return FlowableReplay.createFrom(this); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on the items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * the selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector} is {@code null} * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector) { Objects.requireNonNull(selector, "selector is null"); return FlowableReplay.multicastSelector(FlowableInternalHelper.replaySupplier(this), selector); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}, * replaying {@code bufferSize} notifications. *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * the selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @param bufferSize * the buffer size that limits the number of items the operator can replay * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay * @see #replay(Function, int, boolean) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector, int bufferSize) { Objects.requireNonNull(selector, "selector is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return FlowableReplay.multicastSelector(FlowableInternalHelper.replaySupplier(this, bufferSize, false), selector); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}, * replaying {@code bufferSize} notifications. *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * the selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @param bufferSize * the buffer size that limits the number of items the operator can replay * @param eagerTruncate * if {@code true}, whenever the internal buffer is truncated to the given bufferSize, the * oldest item will be guaranteed dereferenced, thus avoiding unexpected retention * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector, int bufferSize, boolean eagerTruncate) { Objects.requireNonNull(selector, "selector is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return FlowableReplay.multicastSelector(FlowableInternalHelper.replaySupplier(this, bufferSize, eagerTruncate), selector); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}, * replaying no more than {@code bufferSize} items that were emitted within a specified time window. *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * a selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @param bufferSize * the buffer size that limits the number of items the operator can replay * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector} or {@code unit} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector, int bufferSize, long time, @NonNull TimeUnit unit) { return replay(selector, bufferSize, time, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}, * replaying no more than {@code bufferSize} items that were emitted within a specified time window. *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * a selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @param bufferSize * the buffer size that limits the number of items the operator can replay * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that is the time source for the window * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector}, {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException * if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay * @see #replay(Function, int, long, TimeUnit, Scheduler, boolean) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector, int bufferSize, long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(selector, "selector is null"); Objects.requireNonNull(unit, "unit is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); Objects.requireNonNull(scheduler, "scheduler is null"); return FlowableReplay.multicastSelector( FlowableInternalHelper.replaySupplier(this, bufferSize, time, unit, scheduler, false), selector); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}, * replaying no more than {@code bufferSize} items that were emitted within a specified time window. *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * a selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @param bufferSize * the buffer size that limits the number of items the operator can replay * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that is the time source for the window * @param eagerTruncate * if {@code true}, whenever the internal buffer is truncated to the given bufferSize/age, the * oldest item will be guaranteed dereferenced, thus avoiding unexpected retention * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector}, {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException * if {@code bufferSize} is less than zero * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector, int bufferSize, long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean eagerTruncate) { Objects.requireNonNull(selector, "selector is null"); Objects.requireNonNull(unit, "unit is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); Objects.requireNonNull(scheduler, "scheduler is null"); return FlowableReplay.multicastSelector( FlowableInternalHelper.replaySupplier(this, bufferSize, time, unit, scheduler, eagerTruncate), selector); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}, * replaying all items that were emitted within a specified time window. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * a selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector} or {@code unit} is {@code null} * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector, long time, @NonNull TimeUnit unit) { return replay(selector, time, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}, * replaying all items that were emitted within a specified time window. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * a selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @param scheduler * the scheduler that is the time source for the window * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector}, {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Replay * @see #replay(Function, long, TimeUnit, Scheduler, boolean) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector, long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(selector, "selector is null"); Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return FlowableReplay.multicastSelector(FlowableInternalHelper.replaySupplier(this, time, unit, scheduler, false), selector); } /** * Returns a {@code Flowable} that emits items that are the results of invoking a specified selector on items * emitted by a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable}, * replaying all items that were emitted within a specified time window. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param * the type of items emitted by the resulting {@code Flowable} * @param selector * a selector function, which can use the multicasted sequence as many times as needed, without * causing multiple subscriptions to the current {@code Flowable} * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @param scheduler * the scheduler that is the time source for the window * @param eagerTruncate * if {@code true}, whenever the internal buffer is truncated to the given age, the * oldest item will be guaranteed dereferenced, thus avoiding unexpected retention * @return the new {@code Flowable} instance * @throws NullPointerException if {@code selector}, {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final <@NonNull R> Flowable replay(@NonNull Function, @NonNull ? extends Publisher> selector, long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean eagerTruncate) { Objects.requireNonNull(selector, "selector is null"); Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return FlowableReplay.multicastSelector(FlowableInternalHelper.replaySupplier(this, time, unit, scheduler, eagerTruncate), selector); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable} and * replays at most {@code bufferSize} items to late {@link Subscriber}s. A Connectable {@code Flowable} resembles * an ordinary {@code Flowable}, except that it does not begin emitting items when it is subscribed to, but only * when its {@code connect} method is called. *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. * To ensure no beyond-bufferSize items are referenced, * use the {@link #replay(int, boolean)} overload with {@code eagerTruncate = true}. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@code Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} does not operate by default on a particular {@link Scheduler}.
*
* * @param bufferSize * the buffer size that limits the number of items that can be replayed * @return the new {@code ConnectableFlowable} instance * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay * @see #replay(int, boolean) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final ConnectableFlowable replay(int bufferSize) { ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return FlowableReplay.create(this, bufferSize, false); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable} and * replays at most {@code bufferSize} items to late {@link Subscriber}s. A connectable {@code Flowable} resembles * an ordinary {@code Flowable}, except that it does not begin emitting items when it is subscribed to, but only * when its {@code connect} method is called. *

* *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. * To ensure no beyond-bufferSize items are referenced, set {@code eagerTruncate = true}. *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@code Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} does not operate by default on a particular {@link Scheduler}.
*
* * @param bufferSize * the buffer size that limits the number of items that can be replayed * @param eagerTruncate * if {@code true}, whenever the internal buffer is truncated to the given bufferSize, the * oldest item will be guaranteed dereferenced, thus avoiding unexpected retention * @return the new {@code ConnectableFlowable} instance * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay * @since 3.0.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final ConnectableFlowable replay(int bufferSize, boolean eagerTruncate) { ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return FlowableReplay.create(this, bufferSize, eagerTruncate); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable} and * replays at most {@code bufferSize} items that were emitted during a specified time window. A connectable * {@code Flowable} resembles an ordinary {@code Flowable}, except that it does not begin emitting items when it is * subscribed to, but only when its {@code connect} method is called. *

* *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. * To ensure no out-of-date or beyond-bufferSize items are referenced, * use the {@link #replay(int, long, TimeUnit, Scheduler, boolean)} overload with {@code eagerTruncate = true}. *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@link Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param bufferSize * the buffer size that limits the number of items that can be replayed * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @return the new {@code ConnectableFlowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay * @see #replay(int, long, TimeUnit, Scheduler, boolean) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final ConnectableFlowable replay(int bufferSize, long time, @NonNull TimeUnit unit) { return replay(bufferSize, time, unit, Schedulers.computation()); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable} and * replays a maximum of {@code bufferSize} items that are emitted within a specified time window to late {@link Subscriber}s. A * connectable {@code Flowable} resembles an ordinary {@code Flowable}, except that it does not begin emitting items * when it is subscribed to, but only when its {@code connect} method is called. *

* *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. * To ensure no out-of-date or beyond-bufferSize items are referenced, * use the {@link #replay(int, long, TimeUnit, Scheduler, boolean)} overload with {@code eagerTruncate = true}. *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@code Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param bufferSize * the buffer size that limits the number of items that can be replayed * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @param scheduler * the scheduler that is used as a time source for the window * @return the new {@code ConnectableFlowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay * @see #replay(int, long, TimeUnit, Scheduler, boolean) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final ConnectableFlowable replay(int bufferSize, long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return FlowableReplay.create(this, time, unit, scheduler, bufferSize, false); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable} and * replays a maximum of {@code bufferSize} items that are emitted within a specified time window to late {@link Subscriber}s. A * connectable {@code Flowable} resembles an ordinary {@code Flowable}, except that it does not begin emitting items * when it is subscribed to, but only when its {@code connect} method is called. *

* Note that due to concurrency requirements, {@code replay(bufferSize)} may hold strong references to more than * {@code bufferSize} source emissions. To ensure no out-of-date or beyond-bufferSize items * are referenced, set {@code eagerTruncate = true}. *

* *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@code Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param bufferSize * the buffer size that limits the number of items that can be replayed * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @param scheduler * the scheduler that is used as a time source for the window * @param eagerTruncate * if {@code true}, whenever the internal buffer is truncated to the given bufferSize/age, the * oldest item will be guaranteed dereferenced, thus avoiding unexpected retention * @return the new {@code ConnectableFlowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Replay * @since 3.0.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final ConnectableFlowable replay(int bufferSize, long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean eagerTruncate) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return FlowableReplay.create(this, time, unit, scheduler, bufferSize, eagerTruncate); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable} and * replays all items emitted by it within a specified time window to late {@link Subscriber}s. A connectable {@code Flowable} * resembles an ordinary {@code Flowable}, except that it does not begin emitting items when it is subscribed to, * but only when its {@code connect} method is called. *

* *

* Note that the internal buffer may retain strong references to the oldest item. To ensure no out-of-date items * are referenced, use the {@link #replay(long, TimeUnit, Scheduler, boolean)} overload with {@code eagerTruncate = true}. *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@code Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
This version of {@code replay} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @return the new {@code ConnectableFlowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final ConnectableFlowable replay(long time, @NonNull TimeUnit unit) { return replay(time, unit, Schedulers.computation()); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable} and * replays all items emitted by it within a specified time window to late {@link Subscriber}s. A connectable {@code Flowable} * resembles an ordinary {@code Flowable}, except that it does not begin emitting items when it is subscribed to, * but only when its {@code connect} method is called. *

* *

* Note that the internal buffer may retain strong references to the oldest item. To ensure no out-of-date items * are referenced, use the {@link #replay(long, TimeUnit, Scheduler, boolean)} overload with {@code eagerTruncate = true}. *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@code Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that is the time source for the window * @return the new {@code ConnectableFlowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Replay * @see #replay(long, TimeUnit, Scheduler, boolean) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final ConnectableFlowable replay(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return FlowableReplay.create(this, time, unit, scheduler, false); } /** * Returns a {@link ConnectableFlowable} that shares a single subscription to the current {@code Flowable} and * replays all items emitted by it within a specified time window to late {@link Subscriber}s. A connectable {@code Flowable} * resembles an ordinary {@code Flowable}, except that it does not begin emitting items when it is subscribed to, * but only when its {@code connect} method is called. *

* *

* Note that the internal buffer may retain strong references to the oldest item. To ensure no out-of-date items * are referenced, set {@code eagerTruncate = true}. *

*
Backpressure:
*
This operator supports backpressure. Note that the upstream requests are determined by the child * {@code Subscriber} which requests the largest amount: i.e., two child {@code Subscriber}s with requests of 10 and 100 will * request 100 elements from the current {@code Flowable} sequence.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param time * the duration of the window in which the replayed items must have been emitted * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that is the time source for the window * @param eagerTruncate * if {@code true}, whenever the internal buffer is truncated to the given bufferSize/age, the * oldest item will be guaranteed dereferenced, thus avoiding unexpected retention * @return the new {@code ConnectableFlowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Replay */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final ConnectableFlowable replay(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean eagerTruncate) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return FlowableReplay.create(this, time, unit, scheduler, eagerTruncate); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, resubscribing to it if it calls {@code onError} * (infinite retry count). *

* *

* If the current {@code Flowable} calls {@link Subscriber#onError}, this method will resubscribe to the current * {@code Flowable} rather than propagating the {@code onError} call. *

* Any and all items emitted by the current {@code Flowable} will be emitted by the resulting {@code Flowable}, even * those emitted during failed subscriptions. For example, if the current {@code Flowable} fails at first but emits * {@code [1, 2]} then succeeds the second time and emits {@code [1, 2, 3, 4, 5]} then the complete sequence * of emissions and notifications would be {@code [1, 2, 1, 2, 3, 4, 5, onComplete]}. *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code retry} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: Retry */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable retry() { return retry(Long.MAX_VALUE, Functions.alwaysTrue()); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, resubscribing to it if it calls {@code onError} * and the predicate returns {@code true} for that specific exception and retry count. *

* *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code retry} does not operate by default on a particular {@link Scheduler}.
*
* * @param predicate * the predicate that determines if a resubscription may happen in case of a specific exception * and retry count * @return the new {@code Flowable} instance * @throws NullPointerException if {@code predicate} is {@code null} * @see #retry() * @see ReactiveX operators documentation: Retry */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable retry(@NonNull BiPredicate<@NonNull ? super Integer, @NonNull ? super Throwable> predicate) { Objects.requireNonNull(predicate, "predicate is null"); return RxJavaPlugins.onAssembly(new FlowableRetryBiPredicate<>(this, predicate)); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, resubscribing to it if it calls {@code onError} * up to a specified number of retries. *

* *

* If the current {@code Flowable} calls {@link Subscriber#onError}, this method will resubscribe to the current * {@code Flowable} for a maximum of {@code count} resubscriptions rather than propagating the * {@code onError} call. *

* Any and all items emitted by the current {@code Flowable} will be emitted by the resulting {@code Flowable}, even * those emitted during failed subscriptions. For example, if the current {@code Flowable} fails at first but emits * {@code [1, 2]} then succeeds the second time and emits {@code [1, 2, 3, 4, 5]} then the complete sequence * of emissions and notifications would be {@code [1, 2, 1, 2, 3, 4, 5, onComplete]}. *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code retry} does not operate by default on a particular {@link Scheduler}.
*
* * @param times * the number of times to resubscribe if the current {@code Flowable} fails * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code times} is negative * @see ReactiveX operators documentation: Retry */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable retry(long times) { return retry(times, Functions.alwaysTrue()); } /** * Retries at most times or until the predicate returns {@code false}, whichever happens first. * *
*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code retry} does not operate by default on a particular {@link Scheduler}.
*
* @param times the number of times to resubscribe if the current {@code Flowable} fails * @param predicate the predicate called with the failure {@link Throwable} and should return {@code true} to trigger a retry. * @throws NullPointerException if {@code predicate} is {@code null} * @throws IllegalArgumentException if {@code times} is negative * @return the new {@code Flowable} instance */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable retry(long times, @NonNull Predicate<@NonNull ? super Throwable> predicate) { if (times < 0) { throw new IllegalArgumentException("times >= 0 required but it was " + times); } Objects.requireNonNull(predicate, "predicate is null"); return RxJavaPlugins.onAssembly(new FlowableRetryPredicate<>(this, times, predicate)); } /** * Retries the current {@code Flowable} if the predicate returns {@code true}. *
*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code retry} does not operate by default on a particular {@link Scheduler}.
*
* * @param predicate the predicate that receives the failure {@link Throwable} and should return {@code true} to trigger a retry. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code predicate} is {@code null} */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable retry(@NonNull Predicate<@NonNull ? super Throwable> predicate) { return retry(Long.MAX_VALUE, predicate); } /** * Retries until the given stop function returns {@code true}. *
*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code retryUntil} does not operate by default on a particular {@link Scheduler}.
*
* @param stop the function that should return {@code true} to stop retrying * @return the new {@code Flowable} instance * @throws NullPointerException if {@code stop} is {@code null} */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable retryUntil(@NonNull BooleanSupplier stop) { Objects.requireNonNull(stop, "stop is null"); return retry(Long.MAX_VALUE, Functions.predicateReverseFor(stop)); } /** * Returns a {@code Flowable} that emits the same values as the current {@code Flowable} with the exception of an * {@code onError}. An {@code onError} notification from the source will result in the emission of a * {@link Throwable} item to the {@code Flowable} provided as an argument to the {@code notificationHandler} * function. If that {@link Publisher} calls {@code onComplete} or {@code onError} then {@code retry} will call * {@code onComplete} or {@code onError} on the child subscription. Otherwise, this {@code Publisher} will * resubscribe to the current {@code Flowable}. *

* *

* Example: * * This retries 3 times, each time incrementing the number of seconds it waits. * *


     *  Flowable.create((FlowableEmitter<? super String> s) -> {
     *      System.out.println("subscribing");
     *      s.onError(new RuntimeException("always fails"));
     *  }, BackpressureStrategy.BUFFER).retryWhen(attempts -> {
     *      return attempts.zipWith(Flowable.range(1, 3), (n, i) -> i).flatMap(i -> {
     *          System.out.println("delay retry by " + i + " second(s)");
     *          return Flowable.timer(i, TimeUnit.SECONDS);
     *      });
     *  }).blockingForEach(System.out::println);
     * 
* * Output is: * *
 {@code
     * subscribing
     * delay retry by 1 second(s)
     * subscribing
     * delay retry by 2 second(s)
     * subscribing
     * delay retry by 3 second(s)
     * subscribing
     * } 
*

* Note that the inner {@code Publisher} returned by the handler function should signal * either {@code onNext}, {@code onError} or {@code onComplete} in response to the received * {@code Throwable} to indicate the operator should retry or terminate. If the upstream to * the operator is asynchronous, signaling {@code onNext} followed by {@code onComplete} immediately may * result in the sequence to be completed immediately. Similarly, if this inner * {@code Publisher} signals {@code onError} or {@code onComplete} while the upstream is * active, the sequence is terminated with the same signal immediately. *

* The following example demonstrates how to retry an asynchronous source with a delay: *


     * Flowable.timer(1, TimeUnit.SECONDS)
     *     .doOnSubscribe(s -> System.out.println("subscribing"))
     *     .map(v -> { throw new RuntimeException(); })
     *     .retryWhen(errors -> {
     *         AtomicInteger counter = new AtomicInteger();
     *         return errors
     *                   .takeWhile(e -> counter.getAndIncrement() != 3)
     *                   .flatMap(e -> {
     *                       System.out.println("delay retry by " + counter.get() + " second(s)");
     *                       return Flowable.timer(counter.get(), TimeUnit.SECONDS);
     *                   });
     *     })
     *     .blockingSubscribe(System.out::println, System.out::println);
     * 
*
*
Backpressure:
*
The operator honors downstream backpressure and expects both the source * and inner {@code Publisher}s to honor backpressure as well. * If this expectation is violated, the operator may throw an {@link IllegalStateException}.
*
Scheduler:
*
{@code retryWhen} does not operate by default on a particular {@link Scheduler}.
*
* * @param handler * receives a {@code Publisher} of notifications with which a user can complete or error, aborting the * retry * @return the new {@code Flowable} instance * @throws NullPointerException if {@code handler} is {@code null} * @see ReactiveX operators documentation: Retry */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable retryWhen( @NonNull Function, @NonNull ? extends Publisher<@NonNull ?>> handler) { Objects.requireNonNull(handler, "handler is null"); return RxJavaPlugins.onAssembly(new FlowableRetryWhen<>(this, handler)); } /** * Subscribes to the current {@code Flowable} and wraps the given {@link Subscriber} into a {@link SafeSubscriber} * (if not already a {@code SafeSubscriber}) that * deals with exceptions thrown by a misbehaving {@code Subscriber} (that doesn't follow the * Reactive Streams specification). *
*
Backpressure:
*
This operator leaves the reactive world and the backpressure behavior depends on the {@code Subscriber}'s behavior.
*
Scheduler:
*
{@code safeSubscribe} does not operate by default on a particular {@link Scheduler}.
*
* @param subscriber the incoming {@code Subscriber} instance * @throws NullPointerException if {@code subscriber} is {@code null} */ @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final void safeSubscribe(@NonNull Subscriber subscriber) { Objects.requireNonNull(subscriber, "subscriber is null"); if (subscriber instanceof SafeSubscriber) { subscribe((SafeSubscriber)subscriber); } else { subscribe(new SafeSubscriber<>(subscriber)); } } /** * Returns a {@code Flowable} that emits the most recently emitted item (if any) emitted by the current {@code Flowable} * within periodic time intervals. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
{@code sample} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param period * the sampling rate * @param unit * the {@link TimeUnit} in which {@code period} is defined * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @see #throttleLast(long, TimeUnit) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable sample(long period, @NonNull TimeUnit unit) { return sample(period, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits the most recently emitted item (if any) emitted by the current {@code Flowable} * within periodic time intervals and optionally emit the very last upstream item when the upstream completes. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
{@code sample} operates by default on the {@code computation} {@link Scheduler}.
*
* *

History: 2.0.5 - experimental * @param period * the sampling rate * @param unit * the {@link TimeUnit} in which {@code period} is defined * @param emitLast * if {@code true}, and the upstream completes while there is still an unsampled item available, * that item is emitted to downstream before completion * if {@code false}, an unsampled last item is ignored. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @see #throttleLast(long, TimeUnit) * @since 2.1 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable sample(long period, @NonNull TimeUnit unit, boolean emitLast) { return sample(period, unit, Schedulers.computation(), emitLast); } /** * Returns a {@code Flowable} that emits the most recently emitted item (if any) emitted by the current {@code Flowable} * within periodic time intervals, where the intervals are defined on a particular {@link Scheduler}. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param period * the sampling rate * @param unit * the {@link TimeUnit} in which {@code period} is defined * @param scheduler * the {@code Scheduler} to use when sampling * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @see #throttleLast(long, TimeUnit, Scheduler) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable sample(long period, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableSampleTimed<>(this, period, unit, scheduler, false, null)); } /** * Returns a {@code Flowable} that emits the most recently emitted item (if any) emitted by the current {@code Flowable} * within periodic time intervals, where the intervals are defined on a particular {@link Scheduler} * and optionally emit the very last upstream item when the upstream completes. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* *

History: 2.0.5 - experimental * @param period * the sampling rate * @param unit * the {@link TimeUnit} in which {@code period} is defined * @param scheduler * the {@code Scheduler} to use when sampling * @param emitLast * if {@code true} and the upstream completes while there is still an unsampled item available, * that item is emitted to downstream before completion * if {@code false}, an unsampled last item is ignored. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @see #throttleLast(long, TimeUnit, Scheduler) * @since 2.1 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable sample(long period, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean emitLast) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableSampleTimed<>(this, period, unit, scheduler, emitLast, null)); } /** * Returns a {@code Flowable} that emits the most recently emitted item (if any) emitted by the current {@code Flowable} * within periodic time intervals, where the intervals are defined on a particular {@link Scheduler} * and optionally emit the very last upstream item when the upstream completes. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param period * the sampling rate * @param unit * the {@link TimeUnit} in which {@code period} is defined * @param scheduler * the {@code Scheduler} to use when sampling * @param emitLast * if {@code true} and the upstream completes while there is still an unsampled item available, * that item is emitted to downstream before completion * if {@code false}, an unsampled last item is ignored. * @param onDropped * called with the current entry when it has been replaced by a new one * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} or {@code onDropped} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @see #throttleLast(long, TimeUnit, Scheduler) * @since 3.1.6 - Experimental */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @Experimental public final Flowable sample(long period, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean emitLast, @NonNull Consumer onDropped) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); Objects.requireNonNull(onDropped, "onDropped is null"); return RxJavaPlugins.onAssembly(new FlowableSampleTimed<>(this, period, unit, scheduler, emitLast, onDropped)); } /** * Returns a {@code Flowable} that, when the specified {@code sampler} {@link Publisher} emits an item or completes, * emits the most recently emitted item (if any) emitted by the current {@code Flowable} since the previous * emission from the {@code sampler} {@code Publisher}. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses the emissions of the {@code sampler} * {@code Publisher} to control data flow.
*
Scheduler:
*
This version of {@code sample} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the sampler {@code Publisher} * @param sampler * the {@code Publisher} to use for sampling the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sampler} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable sample(@NonNull Publisher sampler) { Objects.requireNonNull(sampler, "sampler is null"); return RxJavaPlugins.onAssembly(new FlowableSamplePublisher<>(this, sampler, false)); } /** * Returns a {@code Flowable} that, when the specified {@code sampler} {@link Publisher} emits an item or completes, * emits the most recently emitted item (if any) emitted by the current {@code Flowable} since the previous * emission from the {@code sampler} {@code Publisher} * and optionally emit the very last upstream item when the upstream or other {@code Publisher} complete. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses the emissions of the {@code sampler} * {@code Publisher} to control data flow.
*
Scheduler:
*
This version of {@code sample} does not operate by default on a particular {@link Scheduler}.
*
* *

History: 2.0.5 - experimental * @param the element type of the sampler {@code Publisher} * @param sampler * the {@code Publisher} to use for sampling the current {@code Flowable} * @param emitLast * if {@code true} and the upstream completes while there is still an unsampled item available, * that item is emitted to downstream before completion * if {@code false}, an unsampled last item is ignored. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code sampler} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @since 2.1 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable sample(@NonNull Publisher sampler, boolean emitLast) { Objects.requireNonNull(sampler, "sampler is null"); return RxJavaPlugins.onAssembly(new FlowableSamplePublisher<>(this, sampler, emitLast)); } /** * Returns a {@code Flowable} that emits the first value emitted by the current {@code Flowable}, then emits one value * for each subsequent value emitted by the current {@code Flowable}. Each emission after the first is the result of * applying the specified accumulator function to the previous emission and the corresponding value from the current {@code Flowable}. *

* *

* This sort of function is sometimes called an accumulator. *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * Violating this expectation, a {@link MissingBackpressureException} may get signaled somewhere downstream.
*
Scheduler:
*
{@code scan} does not operate by default on a particular {@link Scheduler}.
*
* * @param accumulator * an accumulator function to be invoked on each item emitted by the current {@code Flowable}, whose * result will be emitted to {@link Subscriber}s via {@link Subscriber#onNext onNext} and used in the * next accumulator call * @return the new {@code Flowable} instance * @throws NullPointerException if {@code accumulator} is {@code null} * @see ReactiveX operators documentation: Scan */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable scan(@NonNull BiFunction accumulator) { Objects.requireNonNull(accumulator, "accumulator is null"); return RxJavaPlugins.onAssembly(new FlowableScan<>(this, accumulator)); } /** * Returns a {@code Flowable} that emits the provided initial (seed) value, then emits one value for each value emitted * by the current {@code Flowable}. Each emission after the first is the result of applying the specified accumulator * function to the previous emission and the corresponding value from the current {@code Flowable}. *

* *

* This sort of function is sometimes called an accumulator. *

* Note that the {@code Flowable} that results from this method will emit {@code initialValue} as its first * emitted item. *

* Note that the {@code initialValue} is shared among all subscribers to the resulting {@code Flowable} * and may cause problems if it is mutable. To make sure each subscriber gets its own value, defer * the application of this operator via {@link #defer(Supplier)}: *


     * Publisher<T> source = ...
     * Flowable.defer(() -> source.scan(new ArrayList<>(), (list, item) -> list.add(item)));
     *
     * // alternatively, by using compose to stay fluent
     *
     * source.compose(o ->
     *     Flowable.defer(() -> o.scan(new ArrayList<>(), (list, item) -> list.add(item)))
     * );
     * 
*
*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * Violating this expectation, a {@link MissingBackpressureException} may get signaled somewhere downstream. * The downstream request pattern is not preserved across this operator. * The upstream is requested {@link #bufferSize()} - 1 upfront and 75% of {@link #bufferSize()} thereafter.
*
Scheduler:
*
{@code scan} does not operate by default on a particular {@link Scheduler}.
*
* * @param the initial, accumulator and result type * @param initialValue * the initial (seed) accumulator item * @param accumulator * an accumulator function to be invoked on each item emitted by the current {@code Flowable}, whose * result will be emitted to {@link Subscriber}s via {@link Subscriber#onNext onNext} and used in the * next accumulator call * @return the new {@code Flowable} instance * @throws NullPointerException if {@code initialValue} or {@code accumulator} is {@code null} * @see ReactiveX operators documentation: Scan */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable scan(R initialValue, @NonNull BiFunction accumulator) { Objects.requireNonNull(initialValue, "initialValue is null"); return scanWith(Functions.justSupplier(initialValue), accumulator); } /** * Returns a {@code Flowable} that emits the provided initial (seed) value, then emits one value for each value emitted * by the current {@code Flowable}. Each emission after the first is the result of applying the specified accumulator * function to the previous emission and the corresponding value from the current {@code Flowable}. *

* *

* This sort of function is sometimes called an accumulator. *

* Note that the {@code Flowable} that results from this method will emit the value returned by * the {@code seedSupplier} as its first item. *

*
Backpressure:
*
The operator honors downstream backpressure and expects the current {@code Flowable} to honor backpressure as well. * Violating this expectation, a {@link MissingBackpressureException} may get signaled somewhere downstream. * The downstream request pattern is not preserved across this operator. * The upstream is requested {@link #bufferSize()} - 1 upfront and 75% of {@link #bufferSize()} thereafter.
*
Scheduler:
*
{@code scanWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param the initial, accumulator and result type * @param seedSupplier * a {@link Supplier} that returns the initial (seed) accumulator item for each individual {@link Subscriber} * @param accumulator * an accumulator function to be invoked on each item emitted by the current {@code Flowable}, whose * result will be emitted to {@code Subscriber}s via {@link Subscriber#onNext onNext} and used in the * next accumulator call * @return the new {@code Flowable} instance * @throws NullPointerException if {@code seedSupplier} or {@code accumulator} is {@code null} * @see ReactiveX operators documentation: Scan */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable scanWith(@NonNull Supplier seedSupplier, @NonNull BiFunction accumulator) { Objects.requireNonNull(seedSupplier, "seedSupplier is null"); Objects.requireNonNull(accumulator, "accumulator is null"); return RxJavaPlugins.onAssembly(new FlowableScanSeed<>(this, seedSupplier, accumulator)); } /** * Forces the current {@code Flowable}'s emissions and notifications to be serialized and for it to obey * the {@code Publisher} contract in other ways. *

* It is possible for a {@link Publisher} to invoke its {@link Subscriber}s' methods asynchronously, perhaps from * different threads. This could make such a {@code Publisher} poorly-behaved, in that it might try to invoke * {@code onComplete} or {@code onError} before one of its {@code onNext} invocations, or it might call * {@code onNext} from two different threads concurrently. You can force such a {@code Publisher} to be * well-behaved and sequential by applying the {@code serialize} method to it. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code serialize} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: Serialize */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable serialize() { return RxJavaPlugins.onAssembly(new FlowableSerialized<>(this)); } /** * Returns a new {@code Flowable} that multicasts (and shares a single subscription to) the current {@code Flowable}. As long as * there is at least one {@link Subscriber}, the current {@code Flowable} will be subscribed and emitting data. * When all subscribers have canceled it will cancel the current {@code Flowable}. *

* This is an alias for {@link #publish()}.{@link ConnectableFlowable#refCount() refCount()}. *

* *

*
Backpressure:
*
The operator honors backpressure and expects the current {@code Flowable} to honor backpressure as well. * If this expectation is violated, the operator will signal a {@link MissingBackpressureException} to * its {@code Subscriber}s.
*
Scheduler:
*
{@code share} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: RefCount */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable share() { return publish().refCount(); } /** * Returns a {@link Maybe} that completes if this {@code Flowable} is empty, signals one item if this {@code Flowable} * signals exactly one item or signals an {@link IllegalArgumentException} if this {@code Flowable} signals * more than one item. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code singleElement} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Maybe} instance * @see ReactiveX operators documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Maybe singleElement() { return RxJavaPlugins.onAssembly(new FlowableSingleMaybe<>(this)); } /** * Returns a {@link Single} that emits the single item emitted by the current {@code Flowable} if it * emits only a single item, or a default item if the current {@code Flowable} emits no items. If the current * {@code Flowable} emits more than one item, an {@link IllegalArgumentException} is signaled instead. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code single} does not operate by default on a particular {@link Scheduler}.
*
* * @param defaultItem * a default value to emit if the current {@code Flowable} emits no item * @return the new {@code Single} instance * @throws NullPointerException if {@code defaultItem} is {@code null} * @see ReactiveX operators documentation: First */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Single single(@NonNull T defaultItem) { Objects.requireNonNull(defaultItem, "defaultItem is null"); return RxJavaPlugins.onAssembly(new FlowableSingleSingle<>(this, defaultItem)); } /** * Returns a {@link Single} that emits the single item emitted by this {@code Flowable}, if this {@code Flowable} * emits only a single item, otherwise * if this {@code Flowable} completes without emitting any items a {@link NoSuchElementException} will be signaled and * if this {@code Flowable} emits more than one item, an {@link IllegalArgumentException} will be signaled. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure).
*
Scheduler:
*
{@code singleOrError} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Single} instance * @see ReactiveX operators documentation: First */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single singleOrError() { return RxJavaPlugins.onAssembly(new FlowableSingleSingle<>(this, null)); } /** * Returns a {@code Flowable} that skips the first {@code count} items emitted by the current {@code Flowable} and emits * the remainder. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
This version of {@code skip} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * the number of items to skip * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code count} is negative * @see ReactiveX operators documentation: Skip */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable skip(long count) { if (count < 0) { throw new IllegalArgumentException("count >= 0 expected but it was " + count); } if (count == 0) { return RxJavaPlugins.onAssembly(this); } return RxJavaPlugins.onAssembly(new FlowableSkip<>(this, count)); } /** * Returns a {@code Flowable} that skips values emitted by the current {@code Flowable} before a specified time window * elapses. *

* *

*
Backpressure:
*
The operator doesn't support backpressure as it uses time to skip an arbitrary number of elements and * thus has to consume the current {@code Flowable} in an unbounded manner (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code skip} does not operate on any particular scheduler but uses the current time * from the {@code computation} {@link Scheduler}.
*
* * @param time * the length of the time window to skip * @param unit * the time unit of {@code time} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Skip */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable skip(long time, @NonNull TimeUnit unit) { return skipUntil(timer(time, unit)); } /** * Returns a {@code Flowable} that skips values emitted by the current {@code Flowable} before a specified time window * on a specified {@link Scheduler} elapses. *

* *

*
Backpressure:
*
The operator doesn't support backpressure as it uses time to skip an arbitrary number of elements and * thus has to consume the current {@code Flowable} in an unbounded manner (i.e., no backpressure applied to it).
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use for the timed skipping
*
* * @param time * the length of the time window to skip * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} on which the timed wait happens * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Skip */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable skip(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return skipUntil(timer(time, unit, scheduler)); } /** * Returns a {@code Flowable} that drops a specified number of items from the end of the sequence emitted by the * current {@code Flowable}. *

* *

* This {@link Subscriber} accumulates a queue long enough to store the first {@code count} items. As more items are * received, items are taken from the front of the queue and emitted by the resulting {@code Flowable}. This causes * such items to be delayed. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
This version of {@code skipLast} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * number of items to drop from the end of the source sequence * @return the new {@code Flowable} instance * @throws IllegalArgumentException * if {@code count} is less than zero * @see ReactiveX operators documentation: SkipLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable skipLast(int count) { if (count < 0) { throw new IllegalArgumentException("count >= 0 required but it was " + count); } if (count == 0) { return RxJavaPlugins.onAssembly(this); } return RxJavaPlugins.onAssembly(new FlowableSkipLast<>(this, count)); } /** * Returns a {@code Flowable} that drops items emitted by the current {@code Flowable} during a specified time window * before the source completes. *

* *

* Note: this action will cache the latest items arriving in the specified time window. *

*
Backpressure:
*
The operator doesn't support backpressure as it uses time to skip an arbitrary number of elements and * thus has to consume the current {@code Flowable} in an unbounded manner (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code skipLast} does not operate on any particular scheduler but uses the current time * from the {@code computation} {@link Scheduler}.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: SkipLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable skipLast(long time, @NonNull TimeUnit unit) { return skipLast(time, unit, Schedulers.computation(), false, bufferSize()); } /** * Returns a {@code Flowable} that drops items emitted by the current {@code Flowable} during a specified time window * before the source completes. *

* *

* Note: this action will cache the latest items arriving in the specified time window. *

*
Backpressure:
*
The operator doesn't support backpressure as it uses time to skip an arbitrary number of elements and * thus has to consume the current {@code Flowable} in an unbounded manner (i.e., no backpressure applied to it).
*
Scheduler:
*
{@code skipLast} does not operate on any particular scheduler but uses the current time * from the {@code computation} {@link Scheduler}.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param delayError * if {@code true}, an exception signaled by the current {@code Flowable} is delayed until the regular elements are consumed * by the downstream; if {@code false}, an exception is immediately signaled and all regular elements dropped * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: SkipLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable skipLast(long time, @NonNull TimeUnit unit, boolean delayError) { return skipLast(time, unit, Schedulers.computation(), delayError, bufferSize()); } /** * Returns a {@code Flowable} that drops items emitted by the current {@code Flowable} during a specified time window * (defined on a specified scheduler) before the source completes. *

* *

* Note: this action will cache the latest items arriving in the specified time window. *

*
Backpressure:
*
The operator doesn't support backpressure as it uses time to skip an arbitrary number of elements and * thus has to consume the current {@code Flowable} in an unbounded manner (i.e., no backpressure applied to it).
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use for tracking the current time
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the scheduler used as the time source * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: SkipLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable skipLast(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return skipLast(time, unit, scheduler, false, bufferSize()); } /** * Returns a {@code Flowable} that drops items emitted by the current {@code Flowable} during a specified time window * (defined on a specified scheduler) before the source completes. *

* *

* Note: this action will cache the latest items arriving in the specified time window. *

*
Backpressure:
*
The operator doesn't support backpressure as it uses time to skip an arbitrary number of elements and * thus has to consume the current {@code Flowable} in an unbounded manner (i.e., no backpressure applied to it).
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use to track the current time
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the scheduler used as the time source * @param delayError * if {@code true}, an exception signaled by the current {@code Flowable} is delayed until the regular elements are consumed * by the downstream; if {@code false}, an exception is immediately signaled and all regular elements dropped * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: SkipLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable skipLast(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean delayError) { return skipLast(time, unit, scheduler, delayError, bufferSize()); } /** * Returns a {@code Flowable} that drops items emitted by the current {@code Flowable} during a specified time window * (defined on a specified scheduler) before the source completes. *

* *

* Note: this action will cache the latest items arriving in the specified time window. *

*
Backpressure:
*
The operator doesn't support backpressure as it uses time to skip an arbitrary number of elements and * thus has to consume the current {@code Flowable} in an unbounded manner (i.e., no backpressure applied to it).
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the scheduler used as the time source * @param delayError * if {@code true}, an exception signaled by the current {@code Flowable} is delayed until the regular elements are consumed * by the downstream; if {@code false}, an exception is immediately signaled and all regular elements dropped * @param bufferSize * the hint about how many elements to expect to be skipped * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: SkipLast */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable skipLast(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean delayError, int bufferSize) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); // the internal buffer holds pairs of (timestamp, value) so double the default buffer size int s = bufferSize << 1; return RxJavaPlugins.onAssembly(new FlowableSkipLastTimed<>(this, time, unit, scheduler, s, delayError)); } /** * Returns a {@code Flowable} that skips items emitted by the current {@code Flowable} until a second {@link Publisher} emits * an item. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code skipUntil} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the other {@code Publisher} * @param other * the second {@code Publisher} that has to emit an item before the current {@code Flowable}'s elements begin * to be mirrored by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @see ReactiveX operators documentation: SkipUntil */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable skipUntil(@NonNull Publisher other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableSkipUntil<>(this, other)); } /** * Returns a {@code Flowable} that skips all items emitted by the current {@code Flowable} as long as a specified * condition holds {@code true}, but emits all further source items as soon as the condition becomes {@code false}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code skipWhile} does not operate by default on a particular {@link Scheduler}.
*
* * @param predicate * a function to test each item emitted from the current {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code predicate} is {@code null} * @see ReactiveX operators documentation: SkipWhile */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable skipWhile(@NonNull Predicate predicate) { Objects.requireNonNull(predicate, "predicate is null"); return RxJavaPlugins.onAssembly(new FlowableSkipWhile<>(this, predicate)); } /** * Returns a {@code Flowable} that emits the events emitted by source {@link Publisher}, in a * sorted order. Each item emitted by the {@code Publisher} must implement {@link Comparable} with respect to all * other items in the sequence. * *

If any item emitted by this {@code Flowable} does not implement {@code Comparable} with respect to * all other items emitted by this {@code Flowable}, no items will be emitted and the * sequence is terminated with a {@link ClassCastException}. *

Note that calling {@code sorted} with long, non-terminating or infinite sources * might cause {@link OutOfMemoryError} * *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code sorted} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Flowable} instance */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable sorted() { return toList().toFlowable().map(Functions.listSorter(Functions.naturalComparator())).flatMapIterable(Functions.identity()); } /** * Returns a {@code Flowable} that emits the events emitted by source {@link Publisher}, in a * sorted order based on a specified comparison function. * *

Note that calling {@code sorted} with long, non-terminating or infinite sources * might cause {@link OutOfMemoryError} * *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code sorted} does not operate by default on a particular {@link Scheduler}.
*
* * @param comparator * a function that compares two items emitted by the current {@code Flowable} and returns an {@link Integer} * that indicates their sort order * @throws NullPointerException if {@code comparator} is {@code null} * @return the new {@code Flowable} instance */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable sorted(@NonNull Comparator<@NonNull ? super T> comparator) { Objects.requireNonNull(comparator, "comparator is null"); return toList().toFlowable().map(Functions.listSorter(comparator)).flatMapIterable(Functions.identity()); } /** * Returns a {@code Flowable} that emits the items in a specified {@link Iterable} before it begins to emit items * emitted by the current {@code Flowable}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The Current {@code Flowable} * is expected to honor backpressure as well. If it violates this rule, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
{@code startWithIterable} does not operate by default on a particular {@link Scheduler}.
*
* * @param items * an {@code Iterable} that contains the items you want the resulting {@code Flowable} to emit first * @return the new {@code Flowable} instance * @throws NullPointerException if {@code items} is {@code null} * @see ReactiveX operators documentation: StartWith * @see #startWithArray(Object...) * @see #startWithItem(Object) * @since 3.0.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable startWithIterable(@NonNull Iterable items) { return concatArray(fromIterable(items), this); } /** * Returns a {@code Flowable} which first runs the other {@link CompletableSource} * then the current {@code Flowable} if the other completed normally. *

* *

*
Backpressure:
*
The returned {@code Flowable} honors the backpressure of the downstream consumer.
*
Scheduler:
*
{@code startWith} does not operate by default on a particular {@link Scheduler}.
*
* @param other the other {@code CompletableSource} to run first * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 3.0.0 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) public final Flowable startWith(@NonNull CompletableSource other) { Objects.requireNonNull(other, "other is null"); return Flowable.concat(Completable.wrap(other).toFlowable(), this); } /** * Returns a {@code Flowable} which first runs the other {@link SingleSource} * then the current {@code Flowable} if the other succeeded normally. *

* *

*
Backpressure:
*
The returned {@code Flowable} honors the backpressure of the downstream consumer.
*
Scheduler:
*
{@code startWith} does not operate by default on a particular {@link Scheduler}.
*
* @param other the other {@code SingleSource} to run first * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 3.0.0 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) public final Flowable startWith(@NonNull SingleSource other) { Objects.requireNonNull(other, "other is null"); return Flowable.concat(Single.wrap(other).toFlowable(), this); } /** * Returns a {@code Flowable} which first runs the other {@link MaybeSource} * then the current {@code Flowable} if the other succeeded or completed normally. *

* *

*
Backpressure:
*
The returned {@code Flowable} honors the backpressure of the downstream consumer.
*
Scheduler:
*
{@code startWith} does not operate by default on a particular {@link Scheduler}.
*
* @param other the other {@code MaybeSource} to run first * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 3.0.0 */ @CheckReturnValue @NonNull @SchedulerSupport(SchedulerSupport.NONE) @BackpressureSupport(BackpressureKind.FULL) public final Flowable startWith(@NonNull MaybeSource other) { Objects.requireNonNull(other, "other is null"); return Flowable.concat(Maybe.wrap(other).toFlowable(), this); } /** * Returns a {@code Flowable} that emits the items in a specified {@link Publisher} before it begins to emit * items emitted by the current {@code Flowable}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the {@code other} {@code Publisher}s * are expected to honor backpressure as well. If any of then violates this rule, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
{@code startWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param other * a {@code Publisher} that contains the items you want the modified {@code Publisher} to emit first * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @see ReactiveX operators documentation: StartWith */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable startWith(@NonNull Publisher other) { Objects.requireNonNull(other, "other is null"); return concatArray(other, this); } /** * Returns a {@code Flowable} that emits a specified item before it begins to emit items emitted by the current * {@code Flowable}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable} * is expected to honor backpressure as well. If it violates this rule, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
{@code startWithItem} does not operate by default on a particular {@link Scheduler}.
*
* * @param item * the item to emit first * @return the new {@code Flowable} instance * @throws NullPointerException if {@code item} is {@code null} * @see ReactiveX operators documentation: StartWith * @see #startWithArray(Object...) * @see #startWithIterable(Iterable) * @since 3.0.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable startWithItem(@NonNull T item) { Objects.requireNonNull(item, "item is null"); return concatArray(just(item), this); } /** * Returns a {@code Flowable} that emits the specified items before it begins to emit items emitted by the current * {@code Flowable}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The current {@code Flowable} * is expected to honor backpressure as well. If it violates this rule, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
{@code startWithArray} does not operate by default on a particular {@link Scheduler}.
*
* * @param items * the array of values to emit first * @return the new {@code Flowable} instance * @throws NullPointerException if {@code items} is {@code null} * @see ReactiveX operators documentation: StartWith * @see #startWithItem(Object) * @see #startWithIterable(Iterable) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @SafeVarargs @NonNull public final Flowable startWithArray(@NonNull T... items) { Flowable fromArray = fromArray(items); if (fromArray == empty()) { return RxJavaPlugins.onAssembly(this); } return concatArray(fromArray, this); } /** * Subscribes to the current {@code Flowable} and ignores {@code onNext} and {@code onComplete} emissions. *

* If the {@code Flowable} emits an error, it is wrapped into an * {@link io.reactivex.rxjava3.exceptions.OnErrorNotImplementedException OnErrorNotImplementedException} * and routed to the {@link RxJavaPlugins#onError(Throwable)} handler. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code subscribe} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@link Disposable} instance that allows cancelling the flow * @see ReactiveX operators documentation: Subscribe * @see #subscribe(Consumer, Consumer, Action, DisposableContainer) */ @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Disposable subscribe() { return subscribe(Functions.emptyConsumer(), Functions.ON_ERROR_MISSING, Functions.EMPTY_ACTION); } /** * Subscribes to the current {@code Flowable} and provides a callback to handle the items it emits. *

* If the {@code Flowable} emits an error, it is wrapped into an * {@link io.reactivex.rxjava3.exceptions.OnErrorNotImplementedException OnErrorNotImplementedException} * and routed to the {@link RxJavaPlugins#onError(Throwable)} handler. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code subscribe} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext * the {@code Consumer} you have designed to accept emissions from the current {@code Flowable} * @return the new {@link Disposable} instance that allows cancelling the flow * @throws NullPointerException * if {@code onNext} is {@code null} * @see ReactiveX operators documentation: Subscribe * @see #subscribe(Consumer, Consumer, Action, DisposableContainer) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Disposable subscribe(@NonNull Consumer onNext) { return subscribe(onNext, Functions.ON_ERROR_MISSING, Functions.EMPTY_ACTION); } /** * Subscribes to the current {@code Flowable} and provides callbacks to handle the items it emits and any error * notification it issues. *
*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code subscribe} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext * the {@code Consumer} you have designed to accept emissions from the current {@code Flowable} * @param onError * the {@code Consumer} you have designed to accept any error notification from the * current {@code Flowable} * @return the new {@link Disposable} instance that allows cancelling the flow * @throws NullPointerException * if {@code onNext} or {@code onError} is {@code null} * @see ReactiveX operators documentation: Subscribe * @see #subscribe(Consumer, Consumer, Action, DisposableContainer) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Disposable subscribe(@NonNull Consumer onNext, @NonNull Consumer onError) { return subscribe(onNext, onError, Functions.EMPTY_ACTION); } /** * Subscribes to the current {@code Flowable} and provides callbacks to handle the items it emits and any error or * completion notification it issues. *
*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code subscribe} does not operate by default on a particular {@link Scheduler}.
*
* * @param onNext * the {@code Consumer} you have designed to accept emissions from the current {@code Flowable} * @param onError * the {@code Consumer} you have designed to accept any error notification from the * current {@code Flowable} * @param onComplete * the {@link Action} you have designed to accept a completion notification from the * the current {@code Flowable} * @return the new {@link Disposable} instance that allows cancelling the flow * @throws NullPointerException * if {@code onNext}, {@code onError} or {@code onComplete} is {@code null} * @see ReactiveX operators documentation: Subscribe * @see #subscribe(Consumer, Consumer, Action, DisposableContainer) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Disposable subscribe(@NonNull Consumer onNext, @NonNull Consumer onError, @NonNull Action onComplete) { Objects.requireNonNull(onNext, "onNext is null"); Objects.requireNonNull(onError, "onError is null"); Objects.requireNonNull(onComplete, "onComplete is null"); LambdaSubscriber ls = new LambdaSubscriber<>(onNext, onError, onComplete, FlowableInternalHelper.RequestMax.INSTANCE); subscribe(ls); return ls; } /** * Wraps the given onXXX callbacks into a {@link Disposable} {@link Subscriber}, * adds it to the given {@link DisposableContainer} and ensures, that if the upstream * terminates or this particular {@code Disposable} is disposed, the {@code Subscriber} is removed * from the given container. *

* The {@code Subscriber} will be removed after the callback for the terminal event has been invoked. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner (i.e., no * backpressure is applied to it).
*
Scheduler:
*
{@code subscribe} does not operate by default on a particular {@link Scheduler}.
*
* @param onNext the callback for upstream items * @param onError the callback for an upstream error if any * @param onComplete the callback for the upstream completion if any * @param container the {@code DisposableContainer} (such as {@link CompositeDisposable}) to add and remove the * created {@code Disposable} {@code Subscriber} * @return the {@code Disposable} that allows disposing the particular subscription. * @throws NullPointerException * if {@code onNext}, {@code onError}, * {@code onComplete} or {@code container} is {@code null} * @since 3.1.0 */ @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Disposable subscribe( @NonNull Consumer onNext, @NonNull Consumer onError, @NonNull Action onComplete, @NonNull DisposableContainer container) { Objects.requireNonNull(onNext, "onNext is null"); Objects.requireNonNull(onError, "onError is null"); Objects.requireNonNull(onComplete, "onComplete is null"); Objects.requireNonNull(container, "container is null"); DisposableAutoReleaseSubscriber subscriber = new DisposableAutoReleaseSubscriber<>( container, onNext, onError, onComplete); container.add(subscriber); subscribe(subscriber); return subscriber; } @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) @Override public final void subscribe(@NonNull Subscriber subscriber) { if (subscriber instanceof FlowableSubscriber) { subscribe((FlowableSubscriber)subscriber); } else { Objects.requireNonNull(subscriber, "subscriber is null"); subscribe(new StrictSubscriber<>(subscriber)); } } /** * Establish a connection between this {@code Flowable} and the given {@link FlowableSubscriber} and * start streaming events based on the demand of the {@code FlowableSubscriber}. *

* This is a "factory method" and can be called multiple times, each time starting a new {@link Subscription}. *

* Each {@code Subscription} will work for only a single {@code FlowableSubscriber}. *

* If the same {@code FlowableSubscriber} instance is subscribed to multiple {@code Flowable}s and/or the * same {@code Flowable} multiple times, it must ensure the serialization over its {@code onXXX} * methods manually. *

* If the {@code Flowable} rejects the subscription attempt or otherwise fails it will signal * the error via {@link FlowableSubscriber#onError(Throwable)}. *

* This subscribe method relaxes the following Reactive Streams rules: *

    *
  • §1.3: {@code onNext} should not be called concurrently until {@code onSubscribe} returns. * {@link FlowableSubscriber#onSubscribe(Subscription)} should make sure a sync or async call triggered by request() is safe.
  • *
  • §2.3: {@code onError} or {@code onComplete} must not call cancel. * Calling request() or cancel() is NOP at this point.
  • *
  • §2.12: {@code onSubscribe} must be called at most once on the same instance. * {@code FlowableSubscriber} reuse is not checked and if happens, it is the responsibility of * the {@code FlowableSubscriber} to ensure proper serialization of its onXXX methods.
  • *
  • §3.9: negative requests should emit an {@code onError(IllegalArgumentException)}. * Non-positive requests signal via {@link RxJavaPlugins#onError(Throwable)} and the stream is not affected.
  • *
*
*
Backpressure:
*
The backpressure behavior/expectation is determined by the supplied {@code FlowableSubscriber}.
*
Scheduler:
*
{@code subscribe} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.0.7 - experimental; 2.1 - beta * @param subscriber the {@code FlowableSubscriber} that will consume signals from this {@code Flowable} * @throws NullPointerException if {@code subscriber} is {@code null} * @since 2.2 */ @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public final void subscribe(@NonNull FlowableSubscriber subscriber) { Objects.requireNonNull(subscriber, "subscriber is null"); try { Subscriber flowableSubscriber = RxJavaPlugins.onSubscribe(this, subscriber); Objects.requireNonNull(flowableSubscriber, "The RxJavaPlugins.onSubscribe hook returned a null FlowableSubscriber. Please check the handler provided to RxJavaPlugins.setOnFlowableSubscribe for invalid null returns. Further reading: https://github.com/ReactiveX/RxJava/wiki/Plugins"); subscribeActual(flowableSubscriber); } catch (NullPointerException e) { // NOPMD throw e; } catch (Throwable e) { Exceptions.throwIfFatal(e); // can't call onError because no way to know if a Subscription has been set or not // can't call onSubscribe because the call might have set a Subscription already RxJavaPlugins.onError(e); NullPointerException npe = new NullPointerException("Actually not, but can't throw other exceptions due to RS"); npe.initCause(e); throw npe; } } /** * Operator implementations (both source and intermediate) should implement this method that * performs the necessary business logic and handles the incoming {@link Subscriber}s. *

There is no need to call any of the plugin hooks on the current {@code Flowable} instance or * the {@code Subscriber}; all hooks and basic safeguards have been * applied by {@link #subscribe(Subscriber)} before this method gets called. * @param subscriber the incoming {@code Subscriber}, never {@code null} */ protected abstract void subscribeActual(@NonNull Subscriber subscriber); /** * Subscribes a given {@link Subscriber} (subclass) to this {@code Flowable} and returns the given * {@code Subscriber} as is. *

Usage example: *


     * Flowable<Integer> source = Flowable.range(1, 10);
     * CompositeDisposable composite = new CompositeDisposable();
     *
     * ResourceSubscriber<Integer> rs = new ResourceSubscriber<>() {
     *     // ...
     * };
     *
     * composite.add(source.subscribeWith(rs));
     * 
* *
*
Backpressure:
*
The backpressure behavior/expectation is determined by the supplied {@code Subscriber}.
*
Scheduler:
*
{@code subscribeWith} does not operate by default on a particular {@link Scheduler}.
*
* @param the type of the {@code Subscriber} to use and return * @param subscriber the {@code Subscriber} (subclass) to use and return, not {@code null} * @return the input {@code subscriber} * @throws NullPointerException if {@code subscriber} is {@code null} * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull E extends Subscriber> E subscribeWith(E subscriber) { subscribe(subscriber); return subscriber; } /** * Asynchronously subscribes {@link Subscriber}s to the current {@code Flowable} on the specified {@link Scheduler}. *

* If there is a {@link #create(FlowableOnSubscribe, BackpressureStrategy)} type source up in the * chain, it is recommended to use {@code subscribeOn(scheduler, false)} instead * to avoid same-pool deadlock because requests may pile up behind an eager/blocking emitter. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param scheduler * the {@code Scheduler} to perform subscription actions on * @return the new {@code Flowable} instance * @throws NullPointerException if {@code scheduler} is {@code null} * @see ReactiveX operators documentation: SubscribeOn * @see RxJava Threading Examples * @see #observeOn * @see #subscribeOn(Scheduler, boolean) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable subscribeOn(@NonNull Scheduler scheduler) { Objects.requireNonNull(scheduler, "scheduler is null"); return subscribeOn(scheduler, !(this instanceof FlowableCreate)); } /** * Asynchronously subscribes {@link Subscriber}s to the current {@code Flowable} on the specified {@link Scheduler} * optionally reroutes requests from other threads to the same {@code Scheduler} thread. *

* If there is a {@link #create(FlowableOnSubscribe, BackpressureStrategy)} type source up in the * chain, it is recommended to have {@code requestOn} {@code false} to avoid same-pool deadlock * because requests may pile up behind an eager/blocking emitter. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
*

History: 2.1.1 - experimental * @param scheduler * the {@code Scheduler} to perform subscription actions on * @param requestOn if {@code true}, requests are rerouted to the given {@code Scheduler} as well (strong pipelining) * if {@code false}, requests coming from any thread are simply forwarded to * the upstream on the same thread (weak pipelining) * @return the new {@code Flowable} instance * @throws NullPointerException if {@code scheduler} is {@code null} * @see ReactiveX operators documentation: SubscribeOn * @see RxJava Threading Examples * @see #observeOn * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable subscribeOn(@NonNull Scheduler scheduler, boolean requestOn) { Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableSubscribeOn<>(this, scheduler, requestOn)); } /** * Returns a {@code Flowable} that emits the items emitted by the current {@code Flowable} or the items of an alternate * {@link Publisher} if the current {@code Flowable} is empty. *

* *

*
Backpressure:
*
If the current {@code Flowable} is empty, the alternate {@code Publisher} is expected to honor backpressure. * If the current {@code Flowable} is non-empty, it is expected to honor backpressure as instead. * In either case, if violated, a {@link MissingBackpressureException} may get * signaled somewhere downstream. *
*
Scheduler:
*
{@code switchIfEmpty} does not operate by default on a particular {@link Scheduler}.
*
* * @param other * the alternate {@code Publisher} to subscribe to if the source does not emit any items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @since 1.1.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable switchIfEmpty(@NonNull Publisher other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableSwitchIfEmpty<>(this, other)); } /** * Returns a new {@code Flowable} by applying a function that you supply to each item emitted by the current * {@code Flowable} that returns a {@link Publisher}, and then emitting the items emitted by the most recently emitted * of these {@code Publisher}s. *

* The resulting {@code Flowable} completes if both the current {@code Flowable} and the last inner {@code Publisher}, if any, complete. * If the current {@code Flowable} signals an {@code onError}, the inner {@code Publisher} is canceled and the error delivered in-sequence. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed in an * unbounded manner (i.e., without backpressure) and the inner {@code Publisher}s are expected to honor * backpressure but it is not enforced; the operator won't signal a {@link MissingBackpressureException} * but the violation may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code switchMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the inner {@code Publisher}s and the output * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see ReactiveX operators documentation: FlatMap * @see #switchMapDelayError(Function) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable switchMap(@NonNull Function> mapper) { return switchMap(mapper, bufferSize()); } /** * Returns a new {@code Flowable} by applying a function that you supply to each item emitted by the current * {@code Flowable} that returns a {@link Publisher}, and then emitting the items emitted by the most recently emitted * of these {@code Publisher}s. *

* The resulting {@code Flowable} completes if both the current {@code Flowable} and the last inner {@code Publisher}, if any, complete. * If the current {@code Flowable} signals an {@code onError}, the inner {@code Publisher} is canceled and the error delivered in-sequence. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed in an * unbounded manner (i.e., without backpressure) and the inner {@code Publisher}s are expected to honor * backpressure but it is not enforced; the operator won't signal a {@link MissingBackpressureException} * but the violation may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code switchMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the inner {@code Publisher}s and the output * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @param bufferSize * the number of elements to prefetch from the current active inner {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: FlatMap * @see #switchMapDelayError(Function, int) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable switchMap(@NonNull Function> mapper, int bufferSize) { return switchMap0(mapper, bufferSize, false); } /** * Maps the upstream values into {@link CompletableSource}s, subscribes to the newer one while * disposing the subscription to the previous {@code CompletableSource}, thus keeping at most one * active {@code CompletableSource} running. *

* *

* Since a {@code CompletableSource} doesn't produce any items, the resulting reactive type of * this operator is a {@link Completable} that can only indicate successful completion or * a failure in any of the inner {@code CompletableSource}s or the failure of the current * {@code Flowable}. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner and otherwise * does not have backpressure in its return type because no items are ever produced.
*
Scheduler:
*
{@code switchMapCompletable} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
If either this {@code Flowable} or the active {@code CompletableSource} signals an {@code onError}, * the resulting {@code Completable} is terminated immediately with that {@link Throwable}. * Use the {@link #switchMapCompletableDelayError(Function)} to delay such inner failures until * every inner {@code CompletableSource}s and the main {@code Flowable} terminates in some fashion. * If they fail concurrently, the operator may combine the {@code Throwable}s into a * {@link io.reactivex.rxjava3.exceptions.CompositeException CompositeException} * and signal it to the downstream instead. If any inactivated (switched out) {@code CompletableSource} * signals an {@code onError} late, the {@code Throwable}s will be signaled to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. *
*
*

History: 2.1.11 - experimental * @param mapper the function called with each upstream item and should return a * {@code CompletableSource} to be subscribed to and awaited for * (non blockingly) for its terminal event * @return the new {@code Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #switchMapCompletableDelayError(Function) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Completable switchMapCompletable(@NonNull Function mapper) { Objects.requireNonNull(mapper, "mapper is null"); return RxJavaPlugins.onAssembly(new FlowableSwitchMapCompletable<>(this, mapper, false)); } /** * Maps the upstream values into {@link CompletableSource}s, subscribes to the newer one while * disposing the subscription to the previous {@code CompletableSource}, thus keeping at most one * active {@code CompletableSource} running and delaying any main or inner errors until all * of them terminate. *

* *

* Since a {@code CompletableSource} doesn't produce any items, the resulting reactive type of * this operator is a {@link Completable} that can only indicate successful completion or * a failure in any of the inner {@code CompletableSource}s or the failure of the current * {@code Flowable}. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner and otherwise * does not have backpressure in its return type because no items are ever produced.
*
Scheduler:
*
{@code switchMapCompletableDelayError} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
The errors of this {@code Flowable} and all the {@code CompletableSource}s, who had the chance * to run to their completion, are delayed until * all of them terminate in some fashion. At this point, if there was only one failure, the respective * {@link Throwable} is emitted to the downstream. If there was more than one failure, the * operator combines all {@code Throwable}s into a {@link io.reactivex.rxjava3.exceptions.CompositeException CompositeException} * and signals that to the downstream. * If any inactivated (switched out) {@code CompletableSource} * signals an {@code onError} late, the {@code Throwable}s will be signaled to the global error handler via * {@link RxJavaPlugins#onError(Throwable)} method as {@link UndeliverableException} errors. *
*
*

History: 2.1.11 - experimental * @param mapper the function called with each upstream item and should return a * {@code CompletableSource} to be subscribed to and awaited for * (non blockingly) for its terminal event * @return the new {@code Completable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #switchMapCompletable(Function) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Completable switchMapCompletableDelayError(@NonNull Function mapper) { Objects.requireNonNull(mapper, "mapper is null"); return RxJavaPlugins.onAssembly(new FlowableSwitchMapCompletable<>(this, mapper, true)); } /** * Returns a new {@code Flowable} by applying a function that you supply to each item emitted by the current * {@code Flowable} that returns a {@link Publisher}, and then emitting the items emitted by the most recently emitted * of these {@code Publisher}s and delays any error until all {@code Publisher}s terminate. *

* The resulting {@code Flowable} completes if both the current {@code Flowable} and the last inner {@code Publisher}, if any, complete. * If the current {@code Flowable} signals an {@code onError}, the termination of the last inner {@code Publisher} will emit that error as is * or wrapped into a {@link CompositeException} along with the other possible errors the former inner {@code Publisher}s signaled. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed in an * unbounded manner (i.e., without backpressure) and the inner {@code Publisher}s are expected to honor * backpressure but it is not enforced; the operator won't signal a {@link MissingBackpressureException} * but the violation may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code switchMapDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the inner {@code Publisher}s and the output * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see ReactiveX operators documentation: FlatMap * @see #switchMap(Function) * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable switchMapDelayError(@NonNull Function> mapper) { return switchMapDelayError(mapper, bufferSize()); } /** * Returns a new {@code Flowable} by applying a function that you supply to each item emitted by the current * {@code Flowable} that returns a {@link Publisher}, and then emitting the items emitted by the most recently emitted * of these {@code Publisher}s and delays any error until all {@code Publisher}s terminate. *

* The resulting {@code Flowable} completes if both the current {@code Flowable} and the last inner {@code Publisher}, if any, complete. * If the current {@code Flowable} signals an {@code onError}, the termination of the last inner {@code Publisher} will emit that error as is * or wrapped into a {@link CompositeException} along with the other possible errors the former inner {@code Publisher}s signaled. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The outer {@code Publisher} is consumed in an * unbounded manner (i.e., without backpressure) and the inner {@code Publisher}s are expected to honor * backpressure but it is not enforced; the operator won't signal a {@link MissingBackpressureException} * but the violation may lead to {@link OutOfMemoryError} due to internal buffer bloat.
*
Scheduler:
*
{@code switchMapDelayError} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the inner {@code Publisher}s and the output * @param mapper * a function that, when applied to an item emitted by the current {@code Flowable}, returns a * {@code Publisher} * @param bufferSize * the number of elements to prefetch from the current active inner {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: FlatMap * @see #switchMap(Function, int) * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable switchMapDelayError(@NonNull Function> mapper, int bufferSize) { return switchMap0(mapper, bufferSize, true); } Flowable switchMap0(Function> mapper, int bufferSize, boolean delayError) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); if (this instanceof ScalarSupplier) { @SuppressWarnings("unchecked") T v = ((ScalarSupplier)this).get(); if (v == null) { return empty(); } return FlowableScalarXMap.scalarXMap(v, mapper); } return RxJavaPlugins.onAssembly(new FlowableSwitchMap<>(this, mapper, bufferSize, delayError)); } /** * Maps the upstream items into {@link MaybeSource}s and switches (subscribes) to the newer ones * while disposing the older ones (and ignoring their signals) and emits the latest success value of the current one if * available while failing immediately if this {@code Flowable} or any of the * active inner {@code MaybeSource}s fail. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The main {@code Flowable} is consumed in an * unbounded manner (i.e., without backpressure).
*
Scheduler:
*
{@code switchMapMaybe} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
This operator terminates with an {@code onError} if this {@code Flowable} or any of * the inner {@code MaybeSource}s fail while they are active. When this happens concurrently, their * individual {@link Throwable} errors may get combined and emitted as a single * {@link CompositeException}. Otherwise, a late * (i.e., inactive or switched out) {@code onError} from this {@code Flowable} or from any of * the inner {@code MaybeSource}s will be forwarded to the global error handler via * {@link io.reactivex.rxjava3.plugins.RxJavaPlugins#onError(Throwable)} as * {@link io.reactivex.rxjava3.exceptions.UndeliverableException UndeliverableException}
*
*

History: 2.1.11 - experimental * @param the output value type * @param mapper the function called with the current upstream event and should * return a {@code MaybeSource} to replace the current active inner source * and get subscribed to. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #switchMapMaybeDelayError(Function) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable switchMapMaybe(@NonNull Function> mapper) { Objects.requireNonNull(mapper, "mapper is null"); return RxJavaPlugins.onAssembly(new FlowableSwitchMapMaybe<>(this, mapper, false)); } /** * Maps the upstream items into {@link MaybeSource}s and switches (subscribes) to the newer ones * while disposing the older ones (and ignoring their signals) and emits the latest success value of the current one if * available, delaying errors from this {@code Flowable} or the inner {@code MaybeSource}s until all terminate. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The main {@code Flowable} is consumed in an * unbounded manner (i.e., without backpressure).
*
Scheduler:
*
{@code switchMapMaybeDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the output value type * @param mapper the function called with the current upstream event and should * return a {@code MaybeSource} to replace the current active inner source * and get subscribed to. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #switchMapMaybe(Function) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable switchMapMaybeDelayError(@NonNull Function> mapper) { Objects.requireNonNull(mapper, "mapper is null"); return RxJavaPlugins.onAssembly(new FlowableSwitchMapMaybe<>(this, mapper, true)); } /** * Maps the upstream items into {@link SingleSource}s and switches (subscribes) to the newer ones * while disposing the older ones (and ignoring their signals) and emits the latest success value of the current one * while failing immediately if this {@code Flowable} or any of the * active inner {@code SingleSource}s fail. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The main {@code Flowable} is consumed in an * unbounded manner (i.e., without backpressure).
*
Scheduler:
*
{@code switchMapSingle} does not operate by default on a particular {@link Scheduler}.
*
Error handling:
*
This operator terminates with an {@code onError} if this {@code Flowable} or any of * the inner {@code SingleSource}s fail while they are active. When this happens concurrently, their * individual {@link Throwable} errors may get combined and emitted as a single * {@link CompositeException}. Otherwise, a late * (i.e., inactive or switched out) {@code onError} from this {@code Flowable} or from any of * the inner {@code SingleSource}s will be forwarded to the global error handler via * {@link io.reactivex.rxjava3.plugins.RxJavaPlugins#onError(Throwable)} as * {@link io.reactivex.rxjava3.exceptions.UndeliverableException UndeliverableException}
*
*

History: 2.1.11 - experimental * @param the output value type * @param mapper the function called with the current upstream event and should * return a {@code SingleSource} to replace the current active inner source * and get subscribed to. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #switchMapSingleDelayError(Function) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable switchMapSingle(@NonNull Function> mapper) { Objects.requireNonNull(mapper, "mapper is null"); return RxJavaPlugins.onAssembly(new FlowableSwitchMapSingle<>(this, mapper, false)); } /** * Maps the upstream items into {@link SingleSource}s and switches (subscribes) to the newer ones * while disposing the older ones (and ignoring their signals) and emits the latest success value of the current one, * delaying errors from this {@code Flowable} or the inner {@code SingleSource}s until all terminate. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The main {@code Flowable} is consumed in an * unbounded manner (i.e., without backpressure).
*
Scheduler:
*
{@code switchMapSingleDelayError} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.11 - experimental * @param the output value type * @param mapper the function called with the current upstream event and should * return a {@code SingleSource} to replace the current active inner source * and get subscribed to. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @see #switchMapSingle(Function) * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable switchMapSingleDelayError(@NonNull Function> mapper) { Objects.requireNonNull(mapper, "mapper is null"); return RxJavaPlugins.onAssembly(new FlowableSwitchMapSingle<>(this, mapper, true)); } /** * Returns a {@code Flowable} that emits only the first {@code count} items emitted by the current {@code Flowable}. * If the source emits fewer than {@code count} items then all of its items are emitted. *

* *

* This method returns a {@code Flowable} that will invoke a subscribing {@link Subscriber}'s * {@link Subscriber#onNext onNext} function a maximum of {@code count} times before invoking * {@link Subscriber#onComplete onComplete}. *

* Limits both the number of upstream items (after which the sequence completes) * and the total downstream request amount requested from the upstream to * possibly prevent the creation of excess items by the upstream. *

* The operator requests at most the given {@code count} of items from upstream even * if the downstream requests more than that. For example, given a {@code take(5)}, * if the downstream requests 1, a request of 1 is submitted to the upstream * and the operator remembers that only 4 items can be requested now on. A request * of 5 at this point will request 4 from the upstream and any subsequent requests will * be ignored. *

* Note that requests are negotiated on an operator boundary and {@code take}'s amount * may not be preserved further upstream. For example, * {@code source.observeOn(Schedulers.computation()).take(5)} will still request the * default (128) elements from the given {@code source}. *

*
Backpressure:
*
The current {@code Flowable} is consumed in a bounded manner.
*
Scheduler:
*
This version of {@code take} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * the maximum number of items and the total request amount, non-negative. * Zero will immediately cancel the upstream on subscription and complete * the downstream. * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code count} is negative * @see ReactiveX operators documentation: Take */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable take(long count) { if (count < 0) { throw new IllegalArgumentException("count >= 0 required but it was " + count); } return RxJavaPlugins.onAssembly(new FlowableTake<>(this, count)); } /** * Returns a {@code Flowable} that emits those items emitted by source {@link Publisher} before a specified time runs * out. *

* If time runs out before the {@code Flowable} completes normally, the {@code onComplete} event will be * signaled on the default {@code computation} {@link Scheduler}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
This version of {@code take} operates by default on the {@code computation} {@code Scheduler}.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Take */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable take(long time, @NonNull TimeUnit unit) { return takeUntil(timer(time, unit)); } /** * Returns a {@code Flowable} that emits those items emitted by source {@link Publisher} before a specified time (on a * specified {@link Scheduler}) runs out. *

* If time runs out before the {@code Flowable} completes normally, the {@code onComplete} event will be * signaled on the provided {@code Scheduler}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} used for time source * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Take */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable take(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return takeUntil(timer(time, unit, scheduler)); } /** * Returns a {@code Flowable} that emits at most the last {@code count} items emitted by the current {@code Flowable}. If the source emits fewer than * {@code count} items then all of its items are emitted. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream if the {@code count} is non-zero; ignores * backpressure if the {@code count} is zero as it doesn't signal any values.
*
Scheduler:
*
This version of {@code takeLast} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * the maximum number of items to emit from the end of the sequence of items emitted by the current * {@code Flowable} * @return the new {@code Flowable} instance * @throws IllegalArgumentException * if {@code count} is negative * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable takeLast(int count) { if (count < 0) { throw new IllegalArgumentException("count >= 0 required but it was " + count); } else if (count == 0) { return RxJavaPlugins.onAssembly(new FlowableIgnoreElements<>(this)); } else if (count == 1) { return RxJavaPlugins.onAssembly(new FlowableTakeLastOne<>(this)); } return RxJavaPlugins.onAssembly(new FlowableTakeLast<>(this, count)); } /** * Returns a {@code Flowable} that emits at most a specified number of items from the current {@code Flowable} that were * emitted in a specified window of time before the current {@code Flowable} completed. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., no backpressure is applied to it).
*
Scheduler:
*
{@code takeLast} does not operate on any particular scheduler but uses the current time * from the {@code computation} {@link Scheduler}.
*
* * @param count * the maximum number of items to emit * @param time * the length of the time window * @param unit * the time unit of {@code time} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @throws IllegalArgumentException if {@code count} is negative * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable takeLast(long count, long time, @NonNull TimeUnit unit) { return takeLast(count, time, unit, Schedulers.computation(), false, bufferSize()); } /** * Returns a {@code Flowable} that emits at most a specified number of items from the current {@code Flowable} that were * emitted in a specified window of time before the current {@code Flowable} completed, where the timing information is * provided by a given {@link Scheduler}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., no backpressure is applied to it).
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use for tracking the current time
*
* * @param count * the maximum number of items to emit * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that provides the timestamps for the observed items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException * if {@code count} is less than zero * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable takeLast(long count, long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return takeLast(count, time, unit, scheduler, false, bufferSize()); } /** * Returns a {@code Flowable} that emits at most a specified number of items from the current {@code Flowable} that were * emitted in a specified window of time before the current {@code Flowable} completed, where the timing information is * provided by a given {@link Scheduler}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., no backpressure is applied to it).
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use for tracking the current time
*
* * @param count * the maximum number of items to emit * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that provides the timestamps for the observed items * @param delayError * if {@code true}, an exception signaled by the current {@code Flowable} is delayed until the regular elements are consumed * by the downstream; if {@code false}, an exception is immediately signaled and all regular elements dropped * @param bufferSize * the hint about how many elements to expect to be last * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException * if {@code count} is negative or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable takeLast(long count, long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean delayError, int bufferSize) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); if (count < 0) { throw new IllegalArgumentException("count >= 0 required but it was " + count); } return RxJavaPlugins.onAssembly(new FlowableTakeLastTimed<>(this, count, time, unit, scheduler, bufferSize, delayError)); } /** * Returns a {@code Flowable} that emits the items from the current {@code Flowable} that were emitted in a specified * window of time before the current {@code Flowable} completed. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., no backpressure is applied to it) but note that this may * lead to {@link OutOfMemoryError} due to internal buffer bloat. * Consider using {@link #takeLast(long, long, TimeUnit)} in this case.
*
Scheduler:
*
This version of {@code takeLast} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable takeLast(long time, @NonNull TimeUnit unit) { return takeLast(time, unit, Schedulers.computation(), false, bufferSize()); } /** * Returns a {@code Flowable} that emits the items from the current {@code Flowable} that were emitted in a specified * window of time before the current {@code Flowable} completed. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., no backpressure is applied to it) but note that this may * lead to {@link OutOfMemoryError} due to internal buffer bloat. * Consider using {@link #takeLast(long, long, TimeUnit)} in this case.
*
Scheduler:
*
This version of {@code takeLast} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param delayError * if {@code true}, an exception signaled by the current {@code Flowable} is delayed until the regular elements are consumed * by the downstream; if {@code false}, an exception is immediately signaled and all regular elements dropped * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable takeLast(long time, @NonNull TimeUnit unit, boolean delayError) { return takeLast(time, unit, Schedulers.computation(), delayError, bufferSize()); } /** * Returns a {@code Flowable} that emits the items from the current {@code Flowable} that were emitted in a specified * window of time before the current {@code Flowable} completed, where the timing information is provided by a specified * {@link Scheduler}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., no backpressure is applied to it) but note that this may * lead to {@link OutOfMemoryError} due to internal buffer bloat. * Consider using {@link #takeLast(long, long, TimeUnit, Scheduler)} in this case.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that provides the timestamps for the observed items * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable takeLast(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return takeLast(time, unit, scheduler, false, bufferSize()); } /** * Returns a {@code Flowable} that emits the items from the current {@code Flowable} that were emitted in a specified * window of time before the current {@code Flowable} completed, where the timing information is provided by a specified * {@link Scheduler}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., no backpressure is applied to it) but note that this may * lead to {@link OutOfMemoryError} due to internal buffer bloat. * Consider using {@link #takeLast(long, long, TimeUnit, Scheduler)} in this case.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that provides the timestamps for the observed items * @param delayError * if {@code true}, an exception signaled by the current {@code Flowable} is delayed until the regular elements are consumed * by the downstream; if {@code false}, an exception is immediately signaled and all regular elements dropped * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable takeLast(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean delayError) { return takeLast(time, unit, scheduler, delayError, bufferSize()); } /** * Returns a {@code Flowable} that emits the items from the current {@code Flowable} that were emitted in a specified * window of time before the current {@code Flowable} completed, where the timing information is provided by a specified * {@link Scheduler}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., no backpressure is applied to it) but note that this may * lead to {@link OutOfMemoryError} due to internal buffer bloat. * Consider using {@link #takeLast(long, long, TimeUnit, Scheduler)} in this case.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param time * the length of the time window * @param unit * the time unit of {@code time} * @param scheduler * the {@code Scheduler} that provides the timestamps for the observed items * @param delayError * if {@code true}, an exception signaled by the current {@code Flowable} is delayed until the regular elements are consumed * by the downstream; if {@code false}, an exception is immediately signaled and all regular elements dropped * @param bufferSize * the hint about how many elements to expect to be last * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: TakeLast */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable takeLast(long time, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean delayError, int bufferSize) { return takeLast(Long.MAX_VALUE, time, unit, scheduler, delayError, bufferSize); } /** * Returns a {@code Flowable} that emits items emitted by the current {@code Flowable}, checks the specified predicate * for each item, and then completes when the condition is satisfied. *

* *

* The difference between this operator and {@link #takeWhile(Predicate)} is that here, the condition is * evaluated after the item is emitted. * *

*
Backpressure:
*
The operator is a pass-through for backpressure; the backpressure behavior is determined by the upstream * source and the downstream consumer.
*
Scheduler:
*
{@code takeUntil} does not operate by default on a particular {@link Scheduler}.
*
* * @param stopPredicate * a function that evaluates an item emitted by the current {@code Flowable} and returns a {@link Boolean} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code stopPredicate} is {@code null} * @see ReactiveX operators documentation: TakeUntil * @see Flowable#takeWhile(Predicate) * @since 1.1.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable takeUntil(@NonNull Predicate stopPredicate) { Objects.requireNonNull(stopPredicate, "stopPredicate is null"); return RxJavaPlugins.onAssembly(new FlowableTakeUntilPredicate<>(this, stopPredicate)); } /** * Returns a {@code Flowable} that emits the items emitted by the current {@code Flowable} until a second {@link Publisher} * emits an item or completes. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code takeUntil} does not operate by default on a particular {@link Scheduler}.
*
* * @param other * the {@code Publisher} whose first emitted item or completion will cause {@code takeUntil} to stop emitting items * from the current {@code Flowable} * @param * the type of items emitted by {@code other} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} is {@code null} * @see ReactiveX operators documentation: TakeUntil */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U> Flowable takeUntil(@NonNull Publisher other) { Objects.requireNonNull(other, "other is null"); return RxJavaPlugins.onAssembly(new FlowableTakeUntil<>(this, other)); } /** * Returns a {@code Flowable} that emits items emitted by the current {@code Flowable} so long as each item satisfied a * specified condition, and then completes as soon as this condition is not satisfied. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code takeWhile} does not operate by default on a particular {@link Scheduler}.
*
* * @param predicate * a function that evaluates an item emitted by the current {@code Flowable} and returns a {@link Boolean} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code predicate} is {@code null} * @see ReactiveX operators documentation: TakeWhile * @see Flowable#takeUntil(Predicate) */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final Flowable takeWhile(@NonNull Predicate predicate) { Objects.requireNonNull(predicate, "predicate is null"); return RxJavaPlugins.onAssembly(new FlowableTakeWhile<>(this, predicate)); } /** * Returns a {@code Flowable} that emits only the first item emitted by the current {@code Flowable} during sequential * time windows of a specified duration. *

* This differs from {@link #throttleLast} in that this only tracks the passage of time whereas * {@link #throttleLast} ticks at scheduled intervals. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
{@code throttleFirst} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param windowDuration * time to wait before emitting another item after emitting the last item * @param unit * the unit of time of {@code windowDuration} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable throttleFirst(long windowDuration, @NonNull TimeUnit unit) { return throttleFirst(windowDuration, unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits only the first item emitted by the current {@code Flowable} during sequential * time windows of a specified duration, where the windows are managed by a specified {@link Scheduler}. *

* This differs from {@link #throttleLast} in that this only tracks the passage of time whereas * {@link #throttleLast} ticks at scheduled intervals. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param skipDuration * time to wait before emitting another item after emitting the last item * @param unit * the unit of time of {@code skipDuration} * @param scheduler * the {@code Scheduler} to use internally to manage the timers that handle timeout for each * event * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable throttleFirst(long skipDuration, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableThrottleFirstTimed<>(this, skipDuration, unit, scheduler, null)); } /** * Returns a {@code Flowable} that emits only the first item emitted by the current {@code Flowable} during sequential * time windows of a specified duration, where the windows are managed by a specified {@link Scheduler}. *

* This differs from {@link #throttleLast} in that this only tracks the passage of time whereas * {@link #throttleLast} ticks at scheduled intervals. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param skipDuration * time to wait before emitting another item after emitting the last item * @param unit * the unit of time of {@code skipDuration} * @param scheduler * the {@code Scheduler} to use internally to manage the timers that handle timeout for each * event * @param onDropped * called when an item doesn't get delivered to the downstream * * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} or {@code onDropped} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @since 3.1.6 - Experimental */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @Experimental public final Flowable throttleFirst(long skipDuration, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, @NonNull Consumer onDropped) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); Objects.requireNonNull(onDropped, "onDropped is null"); return RxJavaPlugins.onAssembly(new FlowableThrottleFirstTimed<>(this, skipDuration, unit, scheduler, onDropped)); } /** * Returns a {@code Flowable} that emits only the last item emitted by the current {@code Flowable} during sequential * time windows of a specified duration. *

* This differs from {@link #throttleFirst} in that this ticks along at a scheduled interval whereas * {@link #throttleFirst} does not tick, it just tracks the passage of time. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
{@code throttleLast} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param intervalDuration * duration of windows within which the last item emitted by the current {@code Flowable} will be * emitted * @param unit * the unit of time of {@code intervalDuration} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @see #sample(long, TimeUnit) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable throttleLast(long intervalDuration, @NonNull TimeUnit unit) { return sample(intervalDuration, unit); } /** * Returns a {@code Flowable} that emits only the last item emitted by the current {@code Flowable} during sequential * time windows of a specified duration, where the duration is governed by a specified {@link Scheduler}. *

* This differs from {@link #throttleFirst(long, TimeUnit, Scheduler)} in that this ticks along at a scheduled interval whereas * {@code throttleFirst} does not tick, it just tracks the passage of time. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param intervalDuration * duration of windows within which the last item emitted by the current {@code Flowable} will be * emitted * @param unit * the unit of time of {@code intervalDuration} * @param scheduler * the {@code Scheduler} to use internally to manage the timers that handle timeout for each * event * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @see #sample(long, TimeUnit, Scheduler) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable throttleLast(long intervalDuration, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return sample(intervalDuration, unit, scheduler); } /** * Returns a {@code Flowable} that emits only the last item emitted by the current {@code Flowable} during sequential * time windows of a specified duration, where the duration is governed by a specified {@link Scheduler}. *

* This differs from {@link #throttleFirst(long, TimeUnit, Scheduler)} in that this ticks along at a scheduled interval whereas * {@code throttleFirst} does not tick, it just tracks the passage of time. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param intervalDuration * duration of windows within which the last item emitted by the current {@code Flowable} will be * emitted * @param unit * the unit of time of {@code intervalDuration} * @param scheduler * the {@code Scheduler} to use internally to manage the timers that handle timeout for each * event * @param onDropped * called with the current entry when it has been replaced by a new one * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} or {@code onDropped} is {@code null} * @see ReactiveX operators documentation: Sample * @see RxJava wiki: Backpressure * @see #sample(long, TimeUnit, Scheduler) * @since 3.1.6 - Experimental */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull @Experimental public final Flowable throttleLast(long intervalDuration, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, @NonNull Consumer onDropped) { return sample(intervalDuration, unit, scheduler, false, onDropped); } /** * Throttles items from the upstream {@code Flowable} by first emitting the next * item from upstream, then periodically emitting the latest item (if any) when * the specified timeout elapses between them. *

* *

* Unlike the option with {@link #throttleLatest(long, TimeUnit, boolean)}, the very last item being held back * (if any) is not emitted when the upstream completes. *

* If no items were emitted from the upstream during this timeout phase, the next * upstream item is emitted immediately and the timeout window starts from then. *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow. * If the downstream is not ready to receive items, a * {@link io.reactivex.rxjava3.exceptions.MissingBackpressureException MissingBackpressureException} * will be signaled.
*
Scheduler:
*
{@code throttleLatest} operates by default on the {@code computation} {@link Scheduler}.
*
*

History: 2.1.14 - experimental * @param timeout the time to wait after an item emission towards the downstream * before trying to emit the latest item from upstream again * @param unit the time unit * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @since 2.2 * @see #throttleLatest(long, TimeUnit, boolean) * @see #throttleLatest(long, TimeUnit, Scheduler) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable throttleLatest(long timeout, @NonNull TimeUnit unit) { return throttleLatest(timeout, unit, Schedulers.computation(), false); } /** * Throttles items from the upstream {@code Flowable} by first emitting the next * item from upstream, then periodically emitting the latest item (if any) when * the specified timeout elapses between them. *

* *

* If no items were emitted from the upstream during this timeout phase, the next * upstream item is emitted immediately and the timeout window starts from then. *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow. * If the downstream is not ready to receive items, a * {@link io.reactivex.rxjava3.exceptions.MissingBackpressureException MissingBackpressureException} * will be signaled.
*
Scheduler:
*
{@code throttleLatest} operates by default on the {@code computation} {@link Scheduler}.
*
*

History: 2.1.14 - experimental * @param timeout the time to wait after an item emission towards the downstream * before trying to emit the latest item from upstream again * @param unit the time unit * @param emitLast If {@code true}, the very last item from the upstream will be emitted * immediately when the upstream completes, regardless if there is * a timeout window active or not. If {@code false}, the very last * upstream item is ignored and the flow terminates. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see #throttleLatest(long, TimeUnit, Scheduler, boolean) * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable throttleLatest(long timeout, @NonNull TimeUnit unit, boolean emitLast) { return throttleLatest(timeout, unit, Schedulers.computation(), emitLast); } /** * Throttles items from the upstream {@code Flowable} by first emitting the next * item from upstream, then periodically emitting the latest item (if any) when * the specified timeout elapses between them. *

* *

* Unlike the option with {@link #throttleLatest(long, TimeUnit, Scheduler, boolean)}, the very last item being held back * (if any) is not emitted when the upstream completes. *

* If no items were emitted from the upstream during this timeout phase, the next * upstream item is emitted immediately and the timeout window starts from then. *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow. * If the downstream is not ready to receive items, a * {@link io.reactivex.rxjava3.exceptions.MissingBackpressureException MissingBackpressureException} * will be signaled.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
*

History: 2.1.14 - experimental * @param timeout the time to wait after an item emission towards the downstream * before trying to emit the latest item from upstream again * @param unit the time unit * @param scheduler the {@code Scheduler} where the timed wait and latest item * emission will be performed * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see #throttleLatest(long, TimeUnit, Scheduler, boolean) * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable throttleLatest(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return throttleLatest(timeout, unit, scheduler, false); } /** * Throttles items from the upstream {@code Flowable} by first emitting the next * item from upstream, then periodically emitting the latest item (if any) when * the specified timeout elapses between them. *

* *

* If no items were emitted from the upstream during this timeout phase, the next * upstream item is emitted immediately and the timeout window starts from then. *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow. * If the downstream is not ready to receive items, a * {@link io.reactivex.rxjava3.exceptions.MissingBackpressureException MissingBackpressureException} * will be signaled.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
*

History: 2.1.14 - experimental * @param timeout the time to wait after an item emission towards the downstream * before trying to emit the latest item from upstream again * @param unit the time unit * @param scheduler the {@code Scheduler} where the timed wait and latest item * emission will be performed * @param emitLast If {@code true}, the very last item from the upstream will be emitted * immediately when the upstream completes, regardless if there is * a timeout window active or not. If {@code false}, the very last * upstream item is ignored and the flow terminates. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @since 2.2 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable throttleLatest(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean emitLast) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableThrottleLatest<>(this, timeout, unit, scheduler, emitLast, null)); } /** * Throttles items from the upstream {@code Flowable} by first emitting the next * item from upstream, then periodically emitting the latest item (if any) when * the specified timeout elapses between them, invoking the consumer for any dropped item. *

* *

* If no items were emitted from the upstream during this timeout phase, the next * upstream item is emitted immediately and the timeout window starts from then. *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow. * If the downstream is not ready to receive items, a * {@link io.reactivex.rxjava3.exceptions.MissingBackpressureException MissingBackpressureException} * will be signaled.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
Error handling:
*
* If the upstream signals an {@code onError} or {@code onDropped} callback crashes, * the error is delivered immediately to the downstream. If both happen, a {@link CompositeException} * is created, containing both the upstream and the callback error. * If the {@code onDropped} callback crashes during cancellation, the exception is forwarded * to the global error handler via {@link RxJavaPlugins#onError(Throwable)}. *
*
* @param timeout the time to wait after an item emission towards the downstream * before trying to emit the latest item from upstream again * @param unit the time unit * @param scheduler the {@code Scheduler} where the timed wait and latest item * emission will be performed * @param emitLast If {@code true}, the very last item from the upstream will be emitted * immediately when the upstream completes, regardless if there is * a timeout window active or not. If {@code false}, the very last * upstream item is ignored and the flow terminates. * @param onDropped called when an item is replaced by a newer item that doesn't get delivered * to the downstream, including the very last item if {@code emitLast} is {@code false} * and the current undelivered item when the sequence gets canceled. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit}, {@code scheduler} or {@code onDropped} is {@code null} * @since 3.1.6 - Experimental */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @Experimental public final Flowable throttleLatest(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, boolean emitLast, @NonNull Consumer onDropped) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); Objects.requireNonNull(onDropped, "onDropped is null"); return RxJavaPlugins.onAssembly(new FlowableThrottleLatest<>(this, timeout, unit, scheduler, emitLast, onDropped)); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, except that it drops items emitted by the * current {@code Flowable} that are followed by newer items before a timeout value expires. The timer resets on * each emission (alias to {@link #debounce(long, TimeUnit)}). *

* Note: If items keep being emitted by the current {@code Flowable} faster than the timeout then no items * will be emitted by the resulting {@code Flowable}. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
{@code throttleWithTimeout} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timeout * the length of the window of time that must pass after the emission of an item from the current * {@code Flowable} in which it emits no items in order for the item to be emitted by the * resulting {@code Flowable} * @param unit * the unit of time for the specified {@code timeout} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Debounce * @see RxJava wiki: Backpressure * @see #debounce(long, TimeUnit) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable throttleWithTimeout(long timeout, @NonNull TimeUnit unit) { return debounce(timeout, unit); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, except that it drops items emitted by the * current {@code Flowable} that are followed by newer items before a timeout value expires on a specified * {@link Scheduler}. The timer resets on each emission (alias to {@link #debounce(long, TimeUnit, Scheduler)}). *

* Note: If items keep being emitted by the current {@code Flowable} faster than the timeout then no items * will be emitted by the resulting {@code Flowable}. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param timeout * the length of the window of time that must pass after the emission of an item from the current * {@code Flowable} in which it emits no items in order for the item to be emitted by the * resulting {@code Flowable} * @param unit * the unit of time for the specified {@code timeout} * @param scheduler * the {@code Scheduler} to use internally to manage the timers that handle the timeout for each * item * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Debounce * @see RxJava wiki: Backpressure * @see #debounce(long, TimeUnit, Scheduler) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable throttleWithTimeout(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return debounce(timeout, unit, scheduler); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, except that it drops items emitted by the * current {@code Flowable} that are followed by newer items before a timeout value expires on a specified * {@link Scheduler}. The timer resets on each emission (alias to {@link #debounce(long, TimeUnit, Scheduler, Consumer)}). *

* Note: If items keep being emitted by the current {@code Flowable} faster than the timeout then no items * will be emitted by the resulting {@code Flowable}. *

* *

*
Backpressure:
*
This operator does not support backpressure as it uses time to control data flow.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param timeout * the length of the window of time that must pass after the emission of an item from the current * {@code Flowable} in which it emits no items in order for the item to be emitted by the * resulting {@code Flowable} * @param unit * the unit of time for the specified {@code timeout} * @param scheduler * the {@code Scheduler} to use internally to manage the timers that handle the timeout for each * item * @param onDropped * called with the current entry when it has been replaced by a new one * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} or {@code onDropped} is {@code null} * @see ReactiveX operators documentation: Debounce * @see RxJava wiki: Backpressure * @see #debounce(long, TimeUnit, Scheduler, Consumer) * @since 3.1.6 - Experimental */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull @Experimental public final Flowable throttleWithTimeout(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, @NonNull Consumer onDropped) { return debounce(timeout, unit, scheduler, onDropped); } /** * Returns a {@code Flowable} that emits records of the time interval between consecutive items emitted by the * current {@code Flowable}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code timeInterval} does not operate on any particular scheduler but uses the current time * from the {@code computation} {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: TimeInterval */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> timeInterval() { return timeInterval(TimeUnit.MILLISECONDS, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits records of the time interval between consecutive items emitted by the * current {@code Flowable}, where this interval is computed on a specified {@link Scheduler}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code timeInterval} does not operate on any particular scheduler but uses the current time * from the specified {@code Scheduler}.
*
* * @param scheduler * the {@code Scheduler} used to compute time intervals * @return the new {@code Flowable} instance * @throws NullPointerException if {@code scheduler} is {@code null} * @see ReactiveX operators documentation: TimeInterval */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) // Supplied scheduler is only used for creating timestamps. @NonNull public final Flowable> timeInterval(@NonNull Scheduler scheduler) { return timeInterval(TimeUnit.MILLISECONDS, scheduler); } /** * Returns a {@code Flowable} that emits records of the time interval between consecutive items emitted by the * current {@code Flowable}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code timeInterval} does not operate on any particular scheduler but uses the current time * from the {@code computation} {@link Scheduler}.
*
* * @param unit the time unit for the current time * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: TimeInterval */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> timeInterval(@NonNull TimeUnit unit) { return timeInterval(unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits records of the time interval between consecutive items emitted by the * current {@code Flowable}, where this interval is computed on a specified {@link Scheduler}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code timeInterval} does not operate on any particular scheduler but uses the current time * from the specified {@code Scheduler}.
*
* * @param unit the time unit for the current time * @param scheduler * the {@code Scheduler} used to compute time intervals * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: TimeInterval */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) // Supplied scheduler is only used for creating timestamps. @NonNull public final Flowable> timeInterval(@NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableTimeInterval<>(this, unit, scheduler)); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, but notifies {@link Subscriber}s of a * {@link TimeoutException} if an item emitted by the current {@code Flowable} doesn't arrive within a window of * time after the emission of the previous item, where that period of time is measured by a {@link Publisher} that * is a function of the previous item. *

* *

* Note: The arrival of the first source item is never timed out. *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the current {@code Flowable}s violate this, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
This version of {@code timeout} operates by default on the {@code immediate} {@link Scheduler}.
*
* * @param * the timeout value type (ignored) * @param itemTimeoutIndicator * a function that returns a {@code Publisher} for each item emitted by the current * {@code Flowable} and that determines the timeout window for the subsequent item * @return the new {@code Flowable} instance * @throws NullPointerException if {@code itemTimeoutIndicator} is {@code null} * @see ReactiveX operators documentation: Timeout */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull V> Flowable timeout(@NonNull Function> itemTimeoutIndicator) { return timeout0(null, itemTimeoutIndicator, null); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, but that switches to a fallback {@link Publisher} if * an item emitted by the current {@code Flowable} doesn't arrive within a window of time after the emission of the * previous item, where that period of time is measured by a {@code Publisher} that is a function of the previous * item. *

* *

* Note: The arrival of the first source item is never timed out. *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the current {@code Flowable}s violate this, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
This version of {@code timeout} operates by default on the {@code immediate} {@link Scheduler}.
*
* * @param * the timeout value type (ignored) * @param itemTimeoutIndicator * a function that returns a {@code Publisher}, for each item emitted by the current {@code Flowable}, that * determines the timeout window for the subsequent item * @param fallback * the fallback {@code Publisher} to switch to if the current {@code Flowable} times out * @return the new {@code Flowable} instance * @throws NullPointerException if {@code itemTimeoutIndicator} or {@code fallback} is {@code null} * @see ReactiveX operators documentation: Timeout */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull V> Flowable timeout(@NonNull Function> itemTimeoutIndicator, @NonNull Publisher fallback) { Objects.requireNonNull(fallback, "fallback is null"); return timeout0(null, itemTimeoutIndicator, fallback); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable} but applies a timeout policy for each emitted * item. If the next item isn't emitted within the specified timeout duration starting from its predecessor, * the resulting {@code Flowable} terminates and notifies {@link Subscriber}s of a {@link TimeoutException}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
This version of {@code timeout} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timeout * maximum duration between emitted items before a timeout occurs * @param unit * the unit of time that applies to the {@code timeout} argument. * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Timeout */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable timeout(long timeout, @NonNull TimeUnit unit) { return timeout0(timeout, unit, null, Schedulers.computation()); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable} but applies a timeout policy for each emitted * item. If the next item isn't emitted within the specified timeout duration starting from its predecessor, * the current {@code Flowable} is disposed and the resulting {@code Flowable} begins instead to mirror a fallback {@link Publisher}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the current {@code Flowable}s violate this, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
This version of {@code timeout} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timeout * maximum duration between items before a timeout occurs * @param unit * the unit of time that applies to the {@code timeout} argument * @param fallback * the fallback {@code Publisher} to use in case of a timeout * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code fallback} is {@code null} * @see ReactiveX operators documentation: Timeout */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.COMPUTATION) public final Flowable timeout(long timeout, @NonNull TimeUnit unit, @NonNull Publisher fallback) { Objects.requireNonNull(fallback, "fallback is null"); return timeout0(timeout, unit, fallback, Schedulers.computation()); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable} but applies a timeout policy for each emitted * item using a specified {@link Scheduler}. If the next item isn't emitted within the specified timeout duration * starting from its predecessor, the current {@code Flowable} is disposed and the resulting {@code Flowable} begins * instead to mirror a fallback {@link Publisher}. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the current {@code Flowable}s violate this, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param timeout * maximum duration between items before a timeout occurs * @param unit * the unit of time that applies to the {@code timeout} argument * @param scheduler * the {@code Scheduler} to run the timeout timers on * @param fallback * the {@code Publisher} to use as the fallback in case of a timeout * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit}, {@code scheduler} or {@code fallback} is {@code null} * @see ReactiveX operators documentation: Timeout */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable timeout(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, @NonNull Publisher fallback) { Objects.requireNonNull(fallback, "fallback is null"); return timeout0(timeout, unit, fallback, scheduler); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable} but applies a timeout policy for each emitted * item, where this policy is governed by a specified {@link Scheduler}. If the next item isn't emitted within the * specified timeout duration starting from its predecessor, the resulting {@code Flowable} terminates and * notifies {@link Subscriber}s of a {@link TimeoutException}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param timeout * maximum duration between items before a timeout occurs * @param unit * the unit of time that applies to the {@code timeout} argument * @param scheduler * the {@code Scheduler} to run the timeout timers on * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Timeout */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable timeout(long timeout, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return timeout0(timeout, unit, null, scheduler); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, but notifies {@link Subscriber}s of a * {@link TimeoutException} if either the first item emitted by the current {@code Flowable} or any subsequent item * doesn't arrive within time windows defined by other {@link Publisher}s. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. Both this and the returned {@code Publisher}s * are expected to honor backpressure as well. If any of then violates this rule, it may throw an * {@link IllegalStateException} when the {@code Publisher} completes.
*
Scheduler:
*
{@code timeout} does not operate by default on any {@link Scheduler}.
*
* * @param * the first timeout value type (ignored) * @param * the subsequent timeout value type (ignored) * @param firstTimeoutIndicator * a function that returns a {@code Publisher} that determines the timeout window for the first source * item * @param itemTimeoutIndicator * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable} and that * determines the timeout window in which the subsequent source item must arrive in order to * continue the sequence * @return the new {@code Flowable} instance * @throws NullPointerException if {@code firstTimeoutIndicator} or {@code itemTimeoutIndicator} is {@code null} * @see ReactiveX operators documentation: Timeout */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull V> Flowable timeout(@NonNull Publisher firstTimeoutIndicator, @NonNull Function> itemTimeoutIndicator) { Objects.requireNonNull(firstTimeoutIndicator, "firstTimeoutIndicator is null"); return timeout0(firstTimeoutIndicator, itemTimeoutIndicator, null); } /** * Returns a {@code Flowable} that mirrors the current {@code Flowable}, but switches to a fallback {@link Publisher} if either * the first item emitted by the current {@code Flowable} or any subsequent item doesn't arrive within time windows * defined by other {@code Publisher}s. *

* *

*
Backpressure:
*
The operator honors backpressure from downstream. The {@code Publisher} * sources are expected to honor backpressure as well. * If any of the current {@code Flowable}s violate this, it may throw an * {@link IllegalStateException} when the current {@code Flowable} completes.
*
Scheduler:
*
{@code timeout} does not operate by default on any {@link Scheduler}.
*
* * @param * the first timeout value type (ignored) * @param * the subsequent timeout value type (ignored) * @param firstTimeoutIndicator * a function that returns a {@code Publisher} which determines the timeout window for the first source * item * @param itemTimeoutIndicator * a function that returns a {@code Publisher} for each item emitted by the current {@code Flowable} and that * determines the timeout window in which the subsequent source item must arrive in order to * continue the sequence * @param fallback * the fallback {@code Publisher} to switch to if the current {@code Flowable} times out * @return the new {@code Flowable} instance * @throws NullPointerException if {@code firstTimeoutIndicator}, {@code itemTimeoutIndicator} or {@code fallback} is {@code null} * @see ReactiveX operators documentation: Timeout */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull V> Flowable timeout( @NonNull Publisher firstTimeoutIndicator, @NonNull Function> itemTimeoutIndicator, @NonNull Publisher fallback) { Objects.requireNonNull(firstTimeoutIndicator, "firstTimeoutIndicator is null"); Objects.requireNonNull(fallback, "fallback is null"); return timeout0(firstTimeoutIndicator, itemTimeoutIndicator, fallback); } private Flowable timeout0(long timeout, TimeUnit unit, Publisher fallback, Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableTimeoutTimed<>(this, timeout, unit, scheduler, fallback)); } private <@NonNull U, @NonNull V> Flowable timeout0( Publisher firstTimeoutIndicator, Function> itemTimeoutIndicator, Publisher fallback) { Objects.requireNonNull(itemTimeoutIndicator, "itemTimeoutIndicator is null"); return RxJavaPlugins.onAssembly(new FlowableTimeout<>(this, firstTimeoutIndicator, itemTimeoutIndicator, fallback)); } /** * Returns a {@code Flowable} that emits each item emitted by the current {@code Flowable}, wrapped in a * {@link Timed} object. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code timestamp} does not operate on any particular scheduler but uses the current time * from the {@code computation} {@link Scheduler}.
*
* * @return the new {@code Flowable} instance * @see ReactiveX operators documentation: Timestamp */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> timestamp() { return timestamp(TimeUnit.MILLISECONDS, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits each item emitted by the current {@code Flowable}, wrapped in a * {@link Timed} object whose timestamps are provided by a specified {@link Scheduler}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
This operator does not operate on any particular scheduler but uses the current time * from the specified {@code Scheduler}.
*
* * @param scheduler * the {@code Scheduler} to use as a time source * @return the new {@code Flowable} instance * @throws NullPointerException if {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Timestamp */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) // Supplied scheduler is only used for creating timestamps. @NonNull public final Flowable> timestamp(@NonNull Scheduler scheduler) { return timestamp(TimeUnit.MILLISECONDS, scheduler); } /** * Returns a {@code Flowable} that emits each item emitted by the current {@code Flowable}, wrapped in a * {@link Timed} object. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
{@code timestamp} does not operate on any particular scheduler but uses the current time * from the {@code computation} {@link Scheduler}.
*
* * @param unit the time unit for the current time * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Timestamp */ @CheckReturnValue @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> timestamp(@NonNull TimeUnit unit) { return timestamp(unit, Schedulers.computation()); } /** * Returns a {@code Flowable} that emits each item emitted by the current {@code Flowable}, wrapped in a * {@link Timed} object whose timestamps are provided by a specified {@link Scheduler}. *

* *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
This operator does not operate on any particular scheduler but uses the current time * from the specified {@code Scheduler}.
*
* * @param unit the time unit for the current time * @param scheduler * the {@code Scheduler} to use as a time source * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Timestamp */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) // Supplied scheduler is only used for creating timestamps. public final Flowable> timestamp(@NonNull TimeUnit unit, @NonNull Scheduler scheduler) { Objects.requireNonNull(unit, "unit is null"); Objects.requireNonNull(scheduler, "scheduler is null"); return map(Functions.timestampWith(unit, scheduler)); } /** * Calls the specified converter function during assembly time and returns its resulting value. *

* This allows fluent conversion to any other type. *

*
Backpressure:
*
The backpressure behavior depends on what happens in the {@code converter} function.
*
Scheduler:
*
{@code to} does not operate by default on a particular {@link Scheduler}.
*
*

History: 2.1.7 - experimental * @param the resulting object type * @param converter the function that receives the current {@code Flowable} instance and returns a value * @return the converted value * @throws NullPointerException if {@code converter} is {@code null} * @since 2.2 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.SPECIAL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> R to(@NonNull FlowableConverter converter) { return Objects.requireNonNull(converter, "converter is null").apply(this); } /** * Returns a {@link Single} that emits a single item, a list composed of all the items emitted by the * finite upstream source {@link Publisher}. *

* *

* Normally, a {@code Publisher} that returns multiple items will do so by invoking its {@link Subscriber}'s * {@link Subscriber#onNext onNext} method for each such item. You can change this behavior by having the * operator compose a list of all of these items and then to invoke the {@link SingleObserver}'s {@code onSuccess} * method once, passing it the entire list, by calling the {@code Flowable}'s {@code toList} method prior to * calling its {@link #subscribe} method. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated list to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toList} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Single} instance * @see ReactiveX operators documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single> toList() { return RxJavaPlugins.onAssembly(new FlowableToListSingle<>(this)); } /** * Returns a {@link Single} that emits a single item, a list composed of all the items emitted by the * finite source {@link Publisher}. *

* *

* Normally, a {@code Publisher} that returns multiple items will do so by invoking its {@link Subscriber}'s * {@link Subscriber#onNext onNext} method for each such item. You can change this behavior by having the * operator compose a list of all of these items and then to invoke the {@link SingleObserver}'s {@code onSuccess} * method once, passing it the entire list, by calling the {@code Flowable}'s {@code toList} method prior to * calling its {@link #subscribe} method. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated list to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toList} does not operate by default on a particular {@link Scheduler}.
*
* * @param capacityHint * the number of elements expected from the current {@code Flowable} * @return the new {@code Single} instance * @throws IllegalArgumentException if {@code capacityHint} is non-positive * @see ReactiveX operators documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single> toList(int capacityHint) { ObjectHelper.verifyPositive(capacityHint, "capacityHint"); return RxJavaPlugins.onAssembly(new FlowableToListSingle<>(this, Functions.createArrayList(capacityHint))); } /** * Returns a {@link Single} that emits a single item, a list composed of all the items emitted by the * finite source {@link Publisher}. *

* *

* Normally, a {@code Publisher} that returns multiple items will do so by invoking its {@link Subscriber}'s * {@link Subscriber#onNext onNext} method for each such item. You can change this behavior by having the * operator compose a collection of all of these items and then to invoke the {@link SingleObserver}'s {@code onSuccess} * method once, passing it the entire collection, by calling the {@code Flowable}'s {@code toList} method prior to * calling its {@link #subscribe} method. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated collection to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toList} does not operate by default on a particular {@link Scheduler}.
*
* * @param the subclass of a collection of Ts * @param collectionSupplier * the {@link Supplier} returning the collection (for each individual {@code Subscriber}) to be filled in * @return the new {@code Single} instance * @throws NullPointerException if {@code collectionSupplier} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U extends Collection> Single toList(@NonNull Supplier collectionSupplier) { Objects.requireNonNull(collectionSupplier, "collectionSupplier is null"); return RxJavaPlugins.onAssembly(new FlowableToListSingle<>(this, collectionSupplier)); } /** * Returns a {@link Single} that emits a single {@link HashMap} containing all items emitted by the finite source {@link Publisher}, * mapped by the keys returned by a specified {@code keySelector} function. *

* *

* If more than one source item maps to the same key, the {@code HashMap} will contain the latest of those items. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated map to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type of the Map * @param keySelector * the function that extracts the key from a source item to be used in the {@code HashMap} * @return the new {@code Single} instance * @throws NullPointerException if {@code keySelector} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull K> Single> toMap(@NonNull Function keySelector) { Objects.requireNonNull(keySelector, "keySelector is null"); return collect(HashMapSupplier.asSupplier(), Functions.toMapKeySelector(keySelector)); } /** * Returns a {@link Single} that emits a single {@link HashMap} containing values corresponding to items emitted by the * finite source {@link Publisher}, mapped by the keys returned by a specified {@code keySelector} function. *

* *

* If more than one source item maps to the same key, the {@code HashMap} will contain a single entry that * corresponds to the latest of those items. *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated map to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type of the Map * @param the value type of the Map * @param keySelector * the function that extracts the key from a source item to be used in the {@code HashMap} * @param valueSelector * the function that extracts the value from a source item to be used in the {@code HashMap} * @return the new {@code Single} instance * @throws NullPointerException if {@code keySelector} or {@code valueSelector} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull K, @NonNull V> Single> toMap(@NonNull Function keySelector, @NonNull Function valueSelector) { Objects.requireNonNull(keySelector, "keySelector is null"); Objects.requireNonNull(valueSelector, "valueSelector is null"); return collect(HashMapSupplier.asSupplier(), Functions.toMapKeyValueSelector(keySelector, valueSelector)); } /** * Returns a {@link Single} that emits a single {@link Map}, returned by a specified {@code mapFactory} function, that * contains keys and values extracted from the items emitted by the finite source {@link Publisher}. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated map to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toMap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type of the Map * @param the value type of the Map * @param keySelector * the function that extracts the key from a source item to be used in the Map * @param valueSelector * the function that extracts the value from the source items to be used as value in the Map * @param mapSupplier * the function that returns a {@code Map} instance to be used * @return the new {@code Single} instance * @throws NullPointerException if {@code keySelector}, {@code valueSelector} or {@code mapSupplier} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull K, @NonNull V> Single> toMap(@NonNull Function keySelector, @NonNull Function valueSelector, @NonNull Supplier> mapSupplier) { Objects.requireNonNull(keySelector, "keySelector is null"); Objects.requireNonNull(valueSelector, "valueSelector is null"); return collect(mapSupplier, Functions.toMapKeyValueSelector(keySelector, valueSelector)); } /** * Returns a {@link Single} that emits a single {@link HashMap} that contains an {@link ArrayList} of items emitted by the * finite source {@link Publisher} keyed by a specified {@code keySelector} function. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated map to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
This operator does not support backpressure as by intent it is requesting and buffering everything.
*
Scheduler:
*
{@code toMultimap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type of the Map * @param keySelector * the function that extracts the key from the source items to be used as key in the {@code HashMap} * @return the new {@code Single} instance * @throws NullPointerException if {@code keySelector} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K> Single>> toMultimap(@NonNull Function keySelector) { Function valueSelector = Functions.identity(); Supplier>> mapSupplier = HashMapSupplier.asSupplier(); Function> collectionFactory = ArrayListSupplier.asFunction(); return toMultimap(keySelector, valueSelector, mapSupplier, collectionFactory); } /** * Returns a {@link Single} that emits a single {@link HashMap} that contains an {@link ArrayList} of values extracted by a * specified {@code valueSelector} function from items emitted by the finite source {@link Publisher}, keyed by a * specified {@code keySelector} function. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated map to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toMultimap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type of the Map * @param the value type of the Map * @param keySelector * the function that extracts a key from the source items to be used as key in the {@code HashMap} * @param valueSelector * the function that extracts a value from the source items to be used as value in the {@code HashMap} * @return the new {@code Single} instance * @throws NullPointerException if {@code keySelector} or {@code valueSelector} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K, @NonNull V> Single>> toMultimap(@NonNull Function keySelector, @NonNull Function valueSelector) { Supplier>> mapSupplier = HashMapSupplier.asSupplier(); Function> collectionFactory = ArrayListSupplier.asFunction(); return toMultimap(keySelector, valueSelector, mapSupplier, collectionFactory); } /** * Returns a {@link Single} that emits a single {@link Map}, returned by a specified {@code mapFactory} function, that * contains a custom collection of values, extracted by a specified {@code valueSelector} function from * items emitted by the finite source {@link Publisher}, and keyed by the {@code keySelector} function. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated map to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toMultimap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type of the Map * @param the value type of the Map * @param keySelector * the function that extracts a key from the source items to be used as the key in the Map * @param valueSelector * the function that extracts a value from the source items to be used as the value in the {@code Map} * @param mapSupplier * the function that returns a Map instance to be used * @param collectionFactory * the function that returns a {@link Collection} instance for a particular key to be used in the {@code Map} * @return the new {@code Single} instance * @throws NullPointerException if {@code keySelector}, {@code valueSelector}, {@code mapSupplier} or {@code collectionFactory} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull K, @NonNull V> Single>> toMultimap( @NonNull Function keySelector, @NonNull Function valueSelector, @NonNull Supplier>> mapSupplier, @NonNull Function> collectionFactory) { Objects.requireNonNull(keySelector, "keySelector is null"); Objects.requireNonNull(valueSelector, "valueSelector is null"); Objects.requireNonNull(mapSupplier, "mapSupplier is null"); Objects.requireNonNull(collectionFactory, "collectionFactory is null"); return collect(mapSupplier, Functions.toMultimapKeyValueSelector(keySelector, valueSelector, collectionFactory)); } /** * Returns a {@link Single} that emits a single {@link Map}, returned by a specified {@code mapFactory} function, that * contains an {@link ArrayList} of values, extracted by a specified {@code valueSelector} function from items * emitted by the finite source {@link Publisher} and keyed by the {@code keySelector} function. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated map to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toMultimap} does not operate by default on a particular {@link Scheduler}.
*
* * @param the key type of the {@code Map} * @param the value type of the {@code Map} * @param keySelector * the function that extracts a key from the source items to be used as the key in the {@code Map} * @param valueSelector * the function that extracts a value from the source items to be used as the value in the {@code Map} * @param mapSupplier * the function that returns a {@code Map} instance to be used * @return the new {@code Single} instance * @throws NullPointerException if {@code keySelector}, {@code valueSelector} or {@code mapSupplier} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull K, @NonNull V> Single>> toMultimap( @NonNull Function keySelector, @NonNull Function valueSelector, @NonNull Supplier>> mapSupplier ) { return toMultimap(keySelector, valueSelector, mapSupplier, ArrayListSupplier.asFunction()); } /** * Converts the current {@code Flowable} into a non-backpressured {@link Observable}. *
*
Backpressure:
*
{@code Observable}s don't support backpressure thus the current {@code Flowable} is consumed in an unbounded * manner (by requesting {@link Long#MAX_VALUE}).
*
Scheduler:
*
{@code toObservable} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code Observable} instance * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Observable toObservable() { return RxJavaPlugins.onAssembly(new ObservableFromPublisher<>(this)); } /** * Returns a {@link Single} that emits a {@link List} that contains the items emitted by the finite source {@link Publisher}, in a * sorted order. Each item emitted by the {@code Publisher} must implement {@link Comparable} with respect to all * other items in the sequence. * *

If any item emitted by this {@code Flowable} does not implement {@code Comparable} with respect to * all other items emitted by this {@code Flowable}, no items will be emitted and the * sequence is terminated with a {@link ClassCastException}. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated list to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toSortedList} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code Single} instance * @see ReactiveX operators documentation: To */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single> toSortedList() { return toSortedList(Functions.naturalComparator()); } /** * Returns a {@link Single} that emits a {@link List} that contains the items emitted by the finite source {@link Publisher}, in a * sorted order based on a specified comparison function. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated list to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toSortedList} does not operate by default on a particular {@link Scheduler}.
*
* * @param comparator * a function that compares two items emitted by the current {@code Flowable} and returns an {@code int} * that indicates their sort order * @return the new {@code Single} instance * @throws NullPointerException if {@code comparator} is {@code null} * @see ReactiveX operators documentation: To */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Single> toSortedList(@NonNull Comparator comparator) { Objects.requireNonNull(comparator, "comparator is null"); return toList().map(Functions.listSorter(comparator)); } /** * Returns a {@link Single} that emits a {@link List} that contains the items emitted by the finite source {@link Publisher}, in a * sorted order based on a specified comparison function. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated list to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toSortedList} does not operate by default on a particular {@link Scheduler}.
*
* * @param comparator * a function that compares two items emitted by the current {@code Flowable} and returns an {@code int} * that indicates their sort order * @param capacityHint * the initial capacity of the {@link ArrayList} used to accumulate items before sorting * @return the new {@code Single} instance * @throws NullPointerException if {@code comparator} is {@code null} * @throws IllegalArgumentException if {@code capacityHint} is non-positive * @see ReactiveX operators documentation: To * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) public final Single> toSortedList(@NonNull Comparator comparator, int capacityHint) { Objects.requireNonNull(comparator, "comparator is null"); return toList(capacityHint).map(Functions.listSorter(comparator)); } /** * Returns a {@link Single} that emits a {@link List} that contains the items emitted by the finite source {@link Publisher}, in a * sorted order. Each item emitted by the {@code Publisher} must implement {@link Comparable} with respect to all * other items in the sequence. * *

If any item emitted by this {@code Flowable} does not implement {@code Comparable} with respect to * all other items emitted by this {@code Flowable}, no items will be emitted and the * sequence is terminated with a {@link ClassCastException}. *

* *

* Note that this operator requires the upstream to signal {@code onComplete} for the accumulated list to * be emitted. Sources that are infinite and never complete will never emit anything through this * operator and an infinite source may lead to a fatal {@link OutOfMemoryError}. *

*
Backpressure:
*
The operator honors backpressure from downstream and consumes the current {@code Flowable} in an * unbounded manner (i.e., without applying backpressure to it).
*
Scheduler:
*
{@code toSortedList} does not operate by default on a particular {@link Scheduler}.
*
* * @param capacityHint * the initial capacity of the {@link ArrayList} used to accumulate items before sorting * @return the new {@code Single} instance * @throws IllegalArgumentException if {@code capacityHint} is non-positive * @see ReactiveX operators documentation: To * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Single> toSortedList(int capacityHint) { return toSortedList(Functions.naturalComparator(), capacityHint); } /** * Cancels the current {@code Flowable} asynchronously by invoking {@link Subscription#cancel()} * on the specified {@link Scheduler}. *

* The operator suppresses signals from the current {@code Flowable} immediately when the * downstream cancels the flow because the actual cancellation itself could take an arbitrary amount of time * to take effect and make the flow stop producing items. *

*
Backpressure:
*
The operator doesn't interfere with backpressure which is determined by the current {@code Flowable}'s backpressure * behavior.
*
Scheduler:
*
You specify which {@code Scheduler} this operator will use.
*
* * @param scheduler * the {@code Scheduler} to perform cancellation actions on * @return the new {@code Flowable} instance * @throws NullPointerException if {@code scheduler} is {@code null} * @see ReactiveX operators documentation: SubscribeOn */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable unsubscribeOn(@NonNull Scheduler scheduler) { Objects.requireNonNull(scheduler, "scheduler is null"); return RxJavaPlugins.onAssembly(new FlowableUnsubscribeOn<>(this, scheduler)); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping windows, each containing {@code count} items. When the current * {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the current window and * propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window will only contain one element. The behavior is * a trade-off between no-dataloss and ensuring upstream cancellation can happen. *

*
Backpressure:
*
The operator honors backpressure of its inner and outer subscribers, however, the inner {@code Flowable} uses an * unbounded buffer that may hold at most {@code count} elements.
*
Scheduler:
*
This version of {@code window} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * the maximum size of each window before it should be emitted * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> window(long count) { return window(count, count, bufferSize()); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits windows every {@code skip} items, each containing no more than {@code count} items. When * the current {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the current window * and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off between no-dataloss and ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator honors backpressure of its inner and outer subscribers, however, the inner {@code Flowable} uses an * unbounded buffer that may hold at most {@code count} elements.
*
Scheduler:
*
This version of {@code window} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * the maximum size of each window before it should be emitted * @param skip * how many items need to be skipped before starting a new window. Note that if {@code skip} and * {@code count} are equal this is the same operation as {@link #window(long)}. * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code count} or {@code skip} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> window(long count, long skip) { return window(count, skip, bufferSize()); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits windows every {@code skip} items, each containing no more than {@code count} items. When * the current {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the current window * and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off between no-dataloss and ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator honors backpressure of its inner and outer subscribers, however, the inner {@code Flowable} uses an * unbounded buffer that may hold at most {@code count} elements.
*
Scheduler:
*
This version of {@code window} does not operate by default on a particular {@link Scheduler}.
*
* * @param count * the maximum size of each window before it should be emitted * @param skip * how many items need to be skipped before starting a new window. Note that if {@code skip} and * {@code count} are equal this is the same operation as {@link #window(long)}. * @param bufferSize * the capacity hint for the buffer in the inner windows * @return the new {@code Flowable} instance * @throws IllegalArgumentException if {@code count}, {@code skip} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Flowable> window(long count, long skip, int bufferSize) { ObjectHelper.verifyPositive(skip, "skip"); ObjectHelper.verifyPositive(count, "count"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableWindow<>(this, count, skip, bufferSize)); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} starts a new window periodically, as determined by the {@code timeskip} argument. It emits * each window after a fixed timespan, specified by the {@code timespan} argument. When the current * {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the * current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The emitted inner {@code Flowable}s honor * backpressure but have an unbounded inner buffer that may lead to {@link OutOfMemoryError} * if left unconsumed.
*
Scheduler:
*
This version of {@code window} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timespan * the period of time each window collects items before it should be emitted * @param timeskip * the period of time after which a new window will be created * @param unit * the unit of time that applies to the {@code timespan} and {@code timeskip} arguments * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable> window(long timespan, long timeskip, @NonNull TimeUnit unit) { return window(timespan, timeskip, unit, Schedulers.computation(), bufferSize()); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} starts a new window periodically, as determined by the {@code timeskip} argument. It emits * each window after a fixed timespan, specified by the {@code timespan} argument. When the current * {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the * current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The returned inner {@code Flowable}s honor * backpressure but have an unbounded inner buffer that may lead to {@link OutOfMemoryError} * if left unconsumed.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each window collects items before it should be emitted * @param timeskip * the period of time after which a new window will be created * @param unit * the unit of time that applies to the {@code timespan} and {@code timeskip} arguments * @param scheduler * the {@code Scheduler} to use when determining the end and start of a window * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable> window(long timespan, long timeskip, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return window(timespan, timeskip, unit, scheduler, bufferSize()); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} starts a new window periodically, as determined by the {@code timeskip} argument. It emits * each window after a fixed timespan, specified by the {@code timespan} argument. When the current * {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the * current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The returned inner {@code Flowable}s honor * backpressure but have an unbounded inner buffer that may lead to {@link OutOfMemoryError} * if left unconsumed.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each window collects items before it should be emitted * @param timeskip * the period of time after which a new window will be created * @param unit * the unit of time that applies to the {@code timespan} and {@code timeskip} arguments * @param scheduler * the {@code Scheduler} to use when determining the end and start of a window * @param bufferSize * the capacity hint for the buffer in the inner windows * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code timespan}, {@code timeskip} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable> window(long timespan, long timeskip, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, int bufferSize) { ObjectHelper.verifyPositive(bufferSize, "bufferSize"); ObjectHelper.verifyPositive(timespan, "timespan"); ObjectHelper.verifyPositive(timeskip, "timeskip"); Objects.requireNonNull(scheduler, "scheduler is null"); Objects.requireNonNull(unit, "unit is null"); return RxJavaPlugins.onAssembly(new FlowableWindowTimed<>(this, timespan, timeskip, unit, scheduler, Long.MAX_VALUE, bufferSize, false)); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping windows, each of a fixed duration specified by the * {@code timespan} argument. When the current {@code Flowable} completes or encounters an error, the resulting * {@code Flowable} emits the current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The emitted inner {@code Flowable}s honor * backpressure and may hold up to {@code count} elements at most.
*
Scheduler:
*
This version of {@code window} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timespan * the period of time each window collects items before it should be emitted and replaced with a * new window * @param unit * the unit of time that applies to the {@code timespan} argument * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable> window(long timespan, @NonNull TimeUnit unit) { return window(timespan, unit, Schedulers.computation(), Long.MAX_VALUE, false); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping windows, each of a fixed duration as specified by the * {@code timespan} argument or a maximum size as specified by the {@code count} argument (whichever is * reached first). When the current {@code Flowable} completes or encounters an error, the resulting {@code Flowable} * emits the current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The emitted inner {@code Flowable}s honor * backpressure and may hold up to {@code count} elements at most.
*
Scheduler:
*
This version of {@code window} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timespan * the period of time each window collects items before it should be emitted and replaced with a * new window * @param unit * the unit of time that applies to the {@code timespan} argument * @param count * the maximum size of each window before it should be emitted * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable> window(long timespan, @NonNull TimeUnit unit, long count) { return window(timespan, unit, Schedulers.computation(), count, false); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping windows, each of a fixed duration as specified by the * {@code timespan} argument or a maximum size as specified by the {@code count} argument (whichever is * reached first). When the current {@code Flowable} completes or encounters an error, the resulting {@code Flowable} * emits the current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The emitted inner {@code Flowable}s honor * backpressure and may hold up to {@code count} elements at most.
*
Scheduler:
*
This version of {@code window} operates by default on the {@code computation} {@link Scheduler}.
*
* * @param timespan * the period of time each window collects items before it should be emitted and replaced with a * new window * @param unit * the unit of time that applies to the {@code timespan} argument * @param count * the maximum size of each window before it should be emitted * @param restart * if {@code true}, when a window reaches the capacity limit, the timer is restarted as well * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} is {@code null} * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.COMPUTATION) @NonNull public final Flowable> window(long timespan, @NonNull TimeUnit unit, long count, boolean restart) { return window(timespan, unit, Schedulers.computation(), count, restart); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping windows, each of a fixed duration as specified by the * {@code timespan} argument. When the current {@code Flowable} completes or encounters an error, the resulting * {@code Flowable} emits the current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The emitted inner {@code Flowable}s honor * backpressure but have an unbounded inner buffer that may lead to {@link OutOfMemoryError} * if left unconsumed.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each window collects items before it should be emitted and replaced with a * new window * @param unit * the unit of time which applies to the {@code timespan} argument * @param scheduler * the {@code Scheduler} to use when determining the end and start of a window * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable> window(long timespan, @NonNull TimeUnit unit, @NonNull Scheduler scheduler) { return window(timespan, unit, scheduler, Long.MAX_VALUE, false); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping windows, each of a fixed duration specified by the * {@code timespan} argument or a maximum size specified by the {@code count} argument (whichever is reached * first). When the current {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the * current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The emitted inner {@code Flowable}s honor * backpressure and may hold up to {@code count} elements at most.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each window collects items before it should be emitted and replaced with a * new window * @param unit * the unit of time which applies to the {@code timespan} argument * @param count * the maximum size of each window before it should be emitted * @param scheduler * the {@code Scheduler} to use when determining the end and start of a window * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable> window(long timespan, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, long count) { return window(timespan, unit, scheduler, count, false); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping windows, each of a fixed duration specified by the * {@code timespan} argument or a maximum size specified by the {@code count} argument (whichever is reached * first). When the current {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the * current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The emitted inner {@code Flowable}s honor * backpressure and may hold up to {@code count} elements at most.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each window collects items before it should be emitted and replaced with a * new window * @param unit * the unit of time which applies to the {@code timespan} argument * @param count * the maximum size of each window before it should be emitted * @param scheduler * the {@code Scheduler} to use when determining the end and start of a window * @param restart * if {@code true}, when a window reaches the capacity limit, the timer is restarted as well * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code count} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) @NonNull public final Flowable> window(long timespan, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, long count, boolean restart) { return window(timespan, unit, scheduler, count, restart, bufferSize()); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits connected, non-overlapping windows, each of a fixed duration specified by the * {@code timespan} argument or a maximum size specified by the {@code count} argument (whichever is reached * first). When the current {@code Flowable} completes or encounters an error, the resulting {@code Flowable} emits the * current window and propagates the notification from the current {@code Flowable}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The operator consumes the current {@code Flowable} in an unbounded manner. * The resulting {@code Flowable} doesn't support backpressure as it uses * time to control the creation of windows. The emitted inner {@code Flowable}s honor * backpressure and may hold up to {@code count} elements at most.
*
Scheduler:
*
You specify which {@link Scheduler} this operator will use.
*
* * @param timespan * the period of time each window collects items before it should be emitted and replaced with a * new window * @param unit * the unit of time which applies to the {@code timespan} argument * @param count * the maximum size of each window before it should be emitted * @param scheduler * the {@code Scheduler} to use when determining the end and start of a window * @param restart * if {@code true}, when a window reaches the capacity limit, the timer is restarted as well * @param bufferSize * the capacity hint for the buffer in the inner windows * @return the new {@code Flowable} instance * @throws NullPointerException if {@code unit} or {@code scheduler} is {@code null} * @throws IllegalArgumentException if {@code count}, {@code timespan} or {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.CUSTOM) public final Flowable> window( long timespan, @NonNull TimeUnit unit, @NonNull Scheduler scheduler, long count, boolean restart, int bufferSize) { ObjectHelper.verifyPositive(bufferSize, "bufferSize"); Objects.requireNonNull(scheduler, "scheduler is null"); Objects.requireNonNull(unit, "unit is null"); ObjectHelper.verifyPositive(count, "count"); return RxJavaPlugins.onAssembly(new FlowableWindowTimed<>(this, timespan, timespan, unit, scheduler, count, bufferSize, restart)); } /** * Returns a {@code Flowable} that emits non-overlapping windows of items it collects from the current {@code Flowable} * where the boundary of each window is determined by the items emitted from a specified boundary-governing * {@link Publisher}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The outer {@code Publisher} of this operator does not support backpressure as it uses a {@code boundary} {@code Publisher} to control data * flow. The inner {@code Publisher}s honor backpressure and buffer everything until the boundary signals the next element.
*
Scheduler:
*
This version of {@code window} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the window element type (ignored) * @param boundaryIndicator * a {@code Publisher} whose emitted items close and open windows * @return the new {@code Flowable} instance * @throws NullPointerException if {@code boundaryIndicator} is {@code null} * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull B> Flowable> window(@NonNull Publisher boundaryIndicator) { return window(boundaryIndicator, bufferSize()); } /** * Returns a {@code Flowable} that emits non-overlapping windows of items it collects from the current {@code Flowable} * where the boundary of each window is determined by the items emitted from a specified boundary-governing * {@link Publisher}. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The outer {@code Publisher} of this operator does not support backpressure as it uses a {@code boundary} {@code Publisher} to control data * flow. The inner {@code Publisher}s honor backpressure and buffer everything until the boundary signals the next element.
*
Scheduler:
*
This version of {@code window} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the window element type (ignored) * @param boundaryIndicator * a {@code Publisher} whose emitted items close and open windows * @param bufferSize * the capacity hint for the buffer in the inner windows * @return the new {@code Flowable} instance * @throws NullPointerException if {@code boundaryIndicator} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull B> Flowable> window(@NonNull Publisher boundaryIndicator, int bufferSize) { Objects.requireNonNull(boundaryIndicator, "boundaryIndicator is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableWindowBoundary<>(this, boundaryIndicator, bufferSize)); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits windows that contain those items emitted by the current {@code Flowable} between the time when * the {@code windowOpenings} {@link Publisher} emits an item and when the {@code Publisher} returned by * {@code closingSelector} emits an item. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The outer {@code Publisher} of this operator doesn't support backpressure because the emission of new * inner {@code Publisher}s are controlled by the {@code windowOpenings} {@code Publisher}. * The inner {@code Publisher}s honor backpressure and buffer everything until the associated closing * {@code Publisher} signals or completes.
*
Scheduler:
*
This version of {@code window} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the window-opening {@code Publisher} * @param the element type of the window-closing {@code Publisher}s * @param openingIndicator * a {@code Publisher} that, when it emits an item, causes another window to be created * @param closingIndicator * a {@link Function} that produces a {@code Publisher} for every window created. When this {@code Publisher} * emits an item, the associated window is closed and emitted * @return the new {@code Flowable} instance * @throws NullPointerException if {@code openingIndicator} or {@code closingIndicator} is {@code null} * @see ReactiveX operators documentation: Window */ @CheckReturnValue @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U, @NonNull V> Flowable> window( @NonNull Publisher openingIndicator, @NonNull Function> closingIndicator) { return window(openingIndicator, closingIndicator, bufferSize()); } /** * Returns a {@code Flowable} that emits windows of items it collects from the current {@code Flowable}. The resulting * {@code Flowable} emits windows that contain those items emitted by the current {@code Flowable} between the time when * the {@code windowOpenings} {@link Publisher} emits an item and when the {@code Publisher} returned by * {@code closingSelector} emits an item. *

* *

* Note that ignoring windows or subscribing later (i.e., on another thread) will result in * so-called window abandonment where a window may not contain any elements. In this case, subsequent * elements will be dropped until the condition for the next window boundary is satisfied. The behavior is * a trade-off for ensuring upstream cancellation can happen under some race conditions. *

*
Backpressure:
*
The outer {@code Publisher} of this operator doesn't support backpressure because the emission of new * inner {@code Publisher}s are controlled by the {@code windowOpenings} {@code Publisher}. * The inner {@code Publisher}s honor backpressure and buffer everything until the associated closing * {@code Publisher} signals or completes.
*
Scheduler:
*
This version of {@code window} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the window-opening {@code Publisher} * @param the element type of the window-closing {@code Publisher}s * @param openingIndicator * a {@code Publisher} that, when it emits an item, causes another window to be created * @param closingIndicator * a {@link Function} that produces a {@code Publisher} for every window created. When this {@code Publisher} * emits an item, the associated window is closed and emitted * @param bufferSize * the capacity hint for the buffer in the inner windows * @return the new {@code Flowable} instance * @throws NullPointerException if {@code openingIndicator} or {@code closingIndicator} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Window */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.ERROR) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull V> Flowable> window( @NonNull Publisher openingIndicator, @NonNull Function> closingIndicator, int bufferSize) { Objects.requireNonNull(openingIndicator, "openingIndicator is null"); Objects.requireNonNull(closingIndicator, "closingIndicator is null"); ObjectHelper.verifyPositive(bufferSize, "bufferSize"); return RxJavaPlugins.onAssembly(new FlowableWindowBoundarySelector<>(this, openingIndicator, closingIndicator, bufferSize)); } /** * Merges the specified {@link Publisher} into the current {@code Flowable} sequence by using the {@code resultSelector} * function only when the current {@code Flowable} (this instance) emits an item. * *

Note that this operator doesn't emit anything until the other source has produced at * least one value. The resulting emission only happens when the current {@code Flowable} emits (and * not when the other source emits, unlike combineLatest). * If the other source doesn't produce any value and just completes, the sequence is completed immediately. * If the upstream completes before the other source has produced at least one value, the sequence completes * without emission. *

* * *

*
Backpressure:
*
The operator is a pass-through for backpressure: the backpressure support * depends on the upstream and downstream's backpressure behavior. The other {@code Publisher} * is consumed in an unbounded fashion.
*
Scheduler:
*
This operator, by default, doesn't run any particular {@link Scheduler}.
*
* * @param the element type of the other {@code Publisher} * @param the result type of the combination * @param other * the other {@code Publisher} * @param combiner * the function to call when the current {@code Flowable} emits an item and the other {@code Publisher} has already * emitted an item, to generate the item to be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} or {@code combiner} is {@code null} * @since 2.0 * @see ReactiveX operators documentation: CombineLatest */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull R> Flowable withLatestFrom(@NonNull Publisher other, @NonNull BiFunction combiner) { Objects.requireNonNull(other, "other is null"); Objects.requireNonNull(combiner, "combiner is null"); return RxJavaPlugins.onAssembly(new FlowableWithLatestFrom(this, combiner, other)); } /** * Combines the value emission from the current {@code Flowable} with the latest emissions from the * other {@link Publisher}s via a function to produce the output item. * *

Note that this operator doesn't emit anything until all other sources have produced at * least one value. The resulting emission only happens when the current {@code Flowable} emits (and * not when any of the other sources emit, unlike combineLatest). * If a source doesn't produce any value and just completes, the sequence is completed immediately. * If the upstream completes before all other sources have produced at least one value, the sequence completes * without emission. * *

*
Backpressure:
*
This operator is a pass-through for backpressure behavior between the current {@code Flowable} * and the downstream {@link Subscriber}. The other {@code Publisher}s are consumed in an unbounded manner.
*
Scheduler:
*
This operator does not operate by default on a particular {@link Scheduler}.
*
* * @param the first other source's value type * @param the second other source's value type * @param the result value type * @param source1 the first other {@code Publisher} * @param source2 the second other {@code Publisher} * @param combiner the function called with an array of values from each participating {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2} or {@code combiner} is {@code null} * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull T1, @NonNull T2, @NonNull R> Flowable withLatestFrom(@NonNull Publisher source1, @NonNull Publisher source2, @NonNull Function3 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(combiner, "combiner is null"); Function f = Functions.toFunction(combiner); return withLatestFrom(new Publisher[] { source1, source2 }, f); } /** * Combines the value emission from the current {@code Flowable} with the latest emissions from the * other {@link Publisher}s via a function to produce the output item. * *

Note that this operator doesn't emit anything until all other sources have produced at * least one value. The resulting emission only happens when the current {@code Flowable} emits (and * not when any of the other sources emit, unlike combineLatest). * If a source doesn't produce any value and just completes, the sequence is completed immediately. * If the upstream completes before all other sources have produced at least one value, the sequence completes * without emission. * *

*
Backpressure:
*
This operator is a pass-through for backpressure behavior between the current {@code Flowable} * and the downstream {@link Subscriber}. The other {@code Publisher}s are consumed in an unbounded manner.
*
Scheduler:
*
This operator does not operate by default on a particular {@link Scheduler}.
*
* * @param the first other source's value type * @param the second other source's value type * @param the third other source's value type * @param the result value type * @param source1 the first other {@code Publisher} * @param source2 the second other {@code Publisher} * @param source3 the third other {@code Publisher} * @param combiner the function called with an array of values from each participating {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3} or {@code combiner} is {@code null} * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull R> Flowable withLatestFrom( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Function4 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(combiner, "combiner is null"); Function f = Functions.toFunction(combiner); return withLatestFrom(new Publisher[] { source1, source2, source3 }, f); } /** * Combines the value emission from the current {@code Flowable} with the latest emissions from the * other {@link Publisher}s via a function to produce the output item. * *

Note that this operator doesn't emit anything until all other sources have produced at * least one value. The resulting emission only happens when the current {@code Flowable} emits (and * not when any of the other sources emit, unlike combineLatest). * If a source doesn't produce any value and just completes, the sequence is completed immediately. * If the upstream completes before all other sources have produced at least one value, the sequence completes * without emission. * *

*
Backpressure:
*
This operator is a pass-through for backpressure behavior between the current {@code Flowable} * and the downstream {@link Subscriber}. The other {@code Publisher}s are consumed in an unbounded manner.
*
Scheduler:
*
This operator does not operate by default on a particular {@link Scheduler}.
*
* * @param the first other source's value type * @param the second other source's value type * @param the third other source's value type * @param the fourth other source's value type * @param the result value type * @param source1 the first other {@code Publisher} * @param source2 the second other {@code Publisher} * @param source3 the third other {@code Publisher} * @param source4 the fourth other {@code Publisher} * @param combiner the function called with an array of values from each participating {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code source1}, {@code source2}, {@code source3}, * {@code source4} or {@code combiner} is {@code null} * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull T1, @NonNull T2, @NonNull T3, @NonNull T4, @NonNull R> Flowable withLatestFrom( @NonNull Publisher source1, @NonNull Publisher source2, @NonNull Publisher source3, @NonNull Publisher source4, @NonNull Function5 combiner) { Objects.requireNonNull(source1, "source1 is null"); Objects.requireNonNull(source2, "source2 is null"); Objects.requireNonNull(source3, "source3 is null"); Objects.requireNonNull(source4, "source4 is null"); Objects.requireNonNull(combiner, "combiner is null"); Function f = Functions.toFunction(combiner); return withLatestFrom(new Publisher[] { source1, source2, source3, source4 }, f); } /** * Combines the value emission from the current {@code Flowable} with the latest emissions from the * other {@link Publisher}s via a function to produce the output item. * *

Note that this operator doesn't emit anything until all other sources have produced at * least one value. The resulting emission only happens when the current {@code Flowable} emits (and * not when any of the other sources emit, unlike combineLatest). * If a source doesn't produce any value and just completes, the sequence is completed immediately. * If the upstream completes before all other sources have produced at least one value, the sequence completes * without emission. * *

*
Backpressure:
*
This operator is a pass-through for backpressure behavior between the current {@code Flowable} * and the downstream {@link Subscriber}. The other {@code Publisher}s are consumed in an unbounded manner.
*
Scheduler:
*
This operator does not operate by default on a particular {@link Scheduler}.
*
* * @param the result value type * @param others the array of other sources * @param combiner the function called with an array of values from each participating {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code others} or {@code combiner} is {@code null} * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable withLatestFrom(@NonNull Publisher<@NonNull ?>[] others, @NonNull Function combiner) { Objects.requireNonNull(others, "others is null"); Objects.requireNonNull(combiner, "combiner is null"); return RxJavaPlugins.onAssembly(new FlowableWithLatestFromMany<>(this, others, combiner)); } /** * Combines the value emission from the current {@code Flowable} with the latest emissions from the * other {@link Publisher}s via a function to produce the output item. * *

Note that this operator doesn't emit anything until all other sources have produced at * least one value. The resulting emission only happens when the current {@code Flowable} emits (and * not when any of the other sources emit, unlike combineLatest). * If a source doesn't produce any value and just completes, the sequence is completed immediately. * If the upstream completes before all other sources have produced at least one value, the sequence completes * without emission. * *

*
Backpressure:
*
This operator is a pass-through for backpressure behavior between the current {@code Flowable} * and the downstream {@link Subscriber}. The other {@code Publisher}s are consumed in an unbounded manner.
*
Scheduler:
*
This operator does not operate by default on a particular {@link Scheduler}.
*
* * @param the result value type * @param others the iterable of other sources * @param combiner the function called with an array of values from each participating {@code Publisher} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code others} or {@code combiner} is {@code null} * @since 2.0 */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.PASS_THROUGH) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull R> Flowable withLatestFrom(@NonNull Iterable<@NonNull ? extends Publisher<@NonNull ?>> others, @NonNull Function combiner) { Objects.requireNonNull(others, "others is null"); Objects.requireNonNull(combiner, "combiner is null"); return RxJavaPlugins.onAssembly(new FlowableWithLatestFromMany<>(this, others, combiner)); } /** * Returns a {@code Flowable} that emits items that are the result of applying a specified function to pairs of * values, one each from the current {@code Flowable} and a specified {@link Iterable} sequence. *

* *

* Note that the {@code other} {@code Iterable} is evaluated as items are observed from the current {@code Flowable}; it is * not pre-consumed. This allows you to zip infinite streams on either side. *

*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zipWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items in the {@code other} {@code Iterable} * @param * the type of items emitted by the resulting {@code Flowable} * @param other * the {@code Iterable} sequence * @param zipper * a function that combines the pairs of items from the current {@code Flowable} and the {@code Iterable} to generate * the items to be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull R> Flowable zipWith(@NonNull Iterable other, @NonNull BiFunction zipper) { Objects.requireNonNull(other, "other is null"); Objects.requireNonNull(zipper, "zipper is null"); return RxJavaPlugins.onAssembly(new FlowableZipIterable<>(this, other, zipper)); } /** * Returns a {@code Flowable} that emits items that are the result of applying a specified function to pairs of * values, one each from the current {@code Flowable} and another specified {@link Publisher}. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

range(1, 5).doOnComplete(action1).zipWith(range(6, 5).doOnComplete(action2), (a, b) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *

* *

*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zipWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the {@code other} {@code Publisher} * @param * the type of items emitted by the resulting {@code Flowable} * @param other * the other {@code Publisher} * @param zipper * a function that combines the pairs of items from the two {@code Publisher}s to generate the items to * be emitted by the resulting {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip */ @CheckReturnValue @NonNull @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) public final <@NonNull U, @NonNull R> Flowable zipWith(@NonNull Publisher other, @NonNull BiFunction zipper) { Objects.requireNonNull(other, "other is null"); return zip(this, other, zipper); } /** * Returns a {@code Flowable} that emits items that are the result of applying a specified function to pairs of * values, one each from the current {@code Flowable} and another specified {@link Publisher}. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

range(1, 5).doOnComplete(action1).zipWith(range(6, 5).doOnComplete(action2), (a, b) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *

* *

*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zipWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the {@code other} {@code Publisher} * @param * the type of items emitted by the resulting {@code Flowable} * @param other * the other {@code Publisher} * @param zipper * a function that combines the pairs of items from the two {@code Publisher}s to generate the items to * be emitted by the resulting {@code Flowable} * @param delayError * if {@code true}, errors from the current {@code Flowable} or the other {@code Publisher} is delayed until both terminate * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} or {@code zipper} is {@code null} * @see ReactiveX operators documentation: Zip * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U, @NonNull R> Flowable zipWith(@NonNull Publisher other, @NonNull BiFunction zipper, boolean delayError) { return zip(this, other, zipper, delayError); } /** * Returns a {@code Flowable} that emits items that are the result of applying a specified function to pairs of * values, one each from the current {@code Flowable} and another specified {@link Publisher}. *

* The operator subscribes to its sources in the order they are specified and completes eagerly if * one of the sources is shorter than the rest while canceling the other sources. Therefore, it * is possible those other sources will never be able to run to completion (and thus not calling * {@code doOnComplete()}). This can also happen if the sources are exactly the same length; if * source A completes and B has been consumed and is about to complete, the operator detects A won't * be sending further values and it will cancel B immediately. For example: *

range(1, 5).doOnComplete(action1).zipWith(range(6, 5).doOnComplete(action2), (a, b) -> a + b)
* {@code action1} will be called but {@code action2} won't. *
To work around this termination property, * use {@link #doOnCancel(Action)} as well or use {@code using()} to do cleanup in case of completion * or cancellation. *

* *

*
Backpressure:
*
The operator expects backpressure from the sources and honors backpressure from the downstream. * (I.e., zipping with {@link #interval(long, TimeUnit)} may result in {@link MissingBackpressureException}, use * one of the {@code onBackpressureX} to handle similar, backpressure-ignoring sources.
*
Scheduler:
*
{@code zipWith} does not operate by default on a particular {@link Scheduler}.
*
* * @param * the type of items emitted by the {@code other} {@code Publisher} * @param * the type of items emitted by the resulting {@code Flowable} * @param other * the other {@code Publisher} * @param zipper * a function that combines the pairs of items from the two {@code Publisher}s to generate the items to * be emitted by the resulting {@code Flowable} * @param bufferSize * the capacity hint for the buffer in the inner windows * @param delayError * if {@code true}, errors from the current {@code Flowable} or the other {@code Publisher} is delayed until both terminate * @return the new {@code Flowable} instance * @throws NullPointerException if {@code other} or {@code zipper} is {@code null} * @throws IllegalArgumentException if {@code bufferSize} is non-positive * @see ReactiveX operators documentation: Zip * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull U, @NonNull R> Flowable zipWith(@NonNull Publisher other, @NonNull BiFunction zipper, boolean delayError, int bufferSize) { return zip(this, other, zipper, delayError, bufferSize); } // ------------------------------------------------------------------------- // Fluent test support, super handy and reduces test preparation boilerplate // ------------------------------------------------------------------------- /** * Creates a {@link TestSubscriber} that requests {@link Long#MAX_VALUE} and subscribes * it to this {@code Flowable}. *
*
Backpressure:
*
The returned {@code TestSubscriber} consumes this {@code Flowable} in an unbounded fashion.
*
Scheduler:
*
{@code test} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code TestSubscriber} instance * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final TestSubscriber test() { // NoPMD TestSubscriber ts = new TestSubscriber<>(); subscribe(ts); return ts; } /** * Creates a {@link TestSubscriber} with the given initial request amount and subscribes * it to this {@code Flowable}. *
*
Backpressure:
*
The returned {@code TestSubscriber} requests the given {@code initialRequest} amount upfront.
*
Scheduler:
*
{@code test} does not operate by default on a particular {@link Scheduler}.
*
* @param initialRequest the initial request amount, positive * @return the new {@code TestSubscriber} instance * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final TestSubscriber test(long initialRequest) { // NoPMD TestSubscriber ts = new TestSubscriber<>(initialRequest); subscribe(ts); return ts; } /** * Creates a {@link TestSubscriber} with the given initial request amount, * optionally cancels it before the subscription and subscribes * it to this {@code Flowable}. *
*
Backpressure:
*
The returned {@code TestSubscriber} requests the given {@code initialRequest} amount upfront.
*
Scheduler:
*
{@code test} does not operate by default on a particular {@link Scheduler}.
*
* @param initialRequest the initial request amount, positive * @param cancel should the {@code TestSubscriber} be canceled before the subscription? * @return the new {@code TestSubscriber} instance * @since 2.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final TestSubscriber test(long initialRequest, boolean cancel) { // NoPMD TestSubscriber ts = new TestSubscriber<>(initialRequest); if (cancel) { ts.cancel(); } subscribe(ts); return ts; } // ------------------------------------------------------------------------- // JDK 8 Support // ------------------------------------------------------------------------- /** * Converts the existing value of the provided optional into a {@link #just(Object)} * or an empty optional into an {@link #empty()} {@code Flowable} instance. *

* *

* Note that the operator takes an already instantiated optional reference and does not * by any means create this original optional. If the optional is to be created per * consumer upon subscription, use {@link #defer(Supplier)} around {@code fromOptional}: *


     * Flowable.defer(() -> Flowable.fromOptional(createOptional()));
     * 
*
*
Backpressure:
*
The returned {@code Flowable} supports backpressure.
*
Scheduler:
*
{@code fromOptional} does not operate by default on a particular {@link Scheduler}.
*
* @param the element type of the optional value * @param optional the optional value to convert into a {@code Flowable} * @return the new {@code Flowable} instance * @throws NullPointerException if {@code optional} is {@code null} * @since 3.0.0 * @see #just(Object) * @see #empty() */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable<@NonNull T> fromOptional(@NonNull Optional optional) { Objects.requireNonNull(optional, "optional is null"); return optional.map(Flowable::just).orElseGet(Flowable::empty); } /** * Signals the completion value or error of the given (hot) {@link CompletionStage}-based asynchronous calculation. *

* *

* Note that the operator takes an already instantiated, running or terminated {@code CompletionStage}. * If the {@code CompletionStage} is to be created per consumer upon subscription, use {@link #defer(Supplier)} * around {@code fromCompletionStage}: *


     * Flowable.defer(() -> Flowable.fromCompletionStage(createCompletionStage()));
     * 
*

* If the {@code CompletionStage} completes with {@code null}, a {@link NullPointerException} is signaled. *

* Canceling the flow can't cancel the execution of the {@code CompletionStage} because {@code CompletionStage} * itself doesn't support cancellation. Instead, the operator detaches from the {@code CompletionStage}. *

*
Backpressure:
*
The returned {@code Flowable} supports backpressure and caches the completion value until the * downstream is ready to receive it.
*
Scheduler:
*
{@code fromCompletionStage} does not operate by default on a particular {@link Scheduler}.
*
* @param the element type of the {@code CompletionStage} * @param stage the {@code CompletionStage} to convert to {@code Flowable} and signal its terminal value or error * @return the new {@code Flowable} instance * @throws NullPointerException if {@code stage} is {@code null} * @since 3.0.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable<@NonNull T> fromCompletionStage(@NonNull CompletionStage stage) { Objects.requireNonNull(stage, "stage is null"); return RxJavaPlugins.onAssembly(new FlowableFromCompletionStage<>(stage)); } /** * Converts a {@link Stream} into a finite {@code Flowable} and emits its items in the sequence. *

* *

* The operator closes the {@code Stream} upon cancellation and when it terminates. Any exceptions raised when * closing a {@code Stream} are routed to the global error handler ({@link RxJavaPlugins#onError(Throwable)}. * If a {@code Stream} should not be closed, turn it into an {@link Iterable} and use {@link #fromIterable(Iterable)}: *


     * Stream<T> stream = ...
     * Flowable.fromIterable(stream::iterator);
     * 
*

* Note that {@code Stream}s can be consumed only once; any subsequent attempt to consume a {@code Stream} * will result in an {@link IllegalStateException}. *

* Primitive streams are not supported and items have to be boxed manually (e.g., via {@link IntStream#boxed()}): *


     * IntStream intStream = IntStream.rangeClosed(1, 10);
     * Flowable.fromStream(intStream.boxed());
     * 
*

* {@code Stream} does not support concurrent usage so creating and/or consuming the same instance multiple times * from multiple threads can lead to undefined behavior. *

*
Backpressure:
*
The operator honors backpressure from downstream and iterates the given {@code Stream} * on demand (i.e., when requested).
*
Scheduler:
*
{@code fromStream} does not operate by default on a particular {@link Scheduler}.
*
* @param the element type of the source {@code Stream} * @param stream the {@code Stream} of values to emit * @return the new {@code Flowable} instance * @throws NullPointerException if {@code stream} is {@code null} * @since 3.0.0 * @see #fromIterable(Iterable) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public static <@NonNull T> Flowable<@NonNull T> fromStream(@NonNull Stream stream) { Objects.requireNonNull(stream, "stream is null"); return RxJavaPlugins.onAssembly(new FlowableFromStream<>(stream)); } /** * Maps each upstream value into an {@link Optional} and emits the contained item if not empty. *

* * *

*
Backpressure:
*
The operator is a pass-through for downstream requests but issues {@code request(1)} whenever the * mapped {@code Optional} is empty.
*
Scheduler:
*
{@code mapOptional} does not operate by default on a particular {@link Scheduler}.
*
* @param the non-{@code null} output type * @param mapper the function that receives the upstream item and should return a non-empty {@code Optional} * to emit as the output or an empty {@code Optional} to skip to the next upstream value * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @since 3.0.0 * @see #map(Function) * @see #filter(Predicate) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable mapOptional(@NonNull Function> mapper) { Objects.requireNonNull(mapper, "mapper is null"); return RxJavaPlugins.onAssembly(new FlowableMapOptional<>(this, mapper)); } /** * Collects the finite upstream's values into a container via a {@link Stream} {@link Collector} callback set and emits * it as the success result. *

* * *

*
Backpressure:
*
The operator consumes the upstream in an unbounded manner.
*
Scheduler:
*
{@code collect} does not operate by default on a particular {@link Scheduler}.
*
* @param the non-{@code null} result type * @param the intermediate container type used for the accumulation * @param collector the interface defining the container supplier, accumulator and finisher functions; * see {@link Collectors} for some standard implementations * @return the new {@link Single} instance * @throws NullPointerException if {@code collector} is {@code null} * @since 3.0.0 * @see Collectors * @see #collect(Supplier, BiConsumer) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R, @NonNull A> Single collect(@NonNull Collector collector) { Objects.requireNonNull(collector, "collector is null"); return RxJavaPlugins.onAssembly(new FlowableCollectWithCollectorSingle<>(this, collector)); } /** * Signals the first upstream item (or the default item if the upstream is empty) via * a {@link CompletionStage}. *

* *

* The upstream can be canceled by converting the resulting {@code CompletionStage} into * {@link CompletableFuture} via {@link CompletionStage#toCompletableFuture()} and * calling {@link CompletableFuture#cancel(boolean)} on it. * The upstream will be also cancelled if the resulting {@code CompletionStage} is converted to and * completed manually by {@link CompletableFuture#complete(Object)} or {@link CompletableFuture#completeExceptionally(Throwable)}. *

* {@code CompletionStage}s don't have a notion of emptiness and allow {@code null}s, therefore, one can either use * a {@code defaultItem} of {@code null} or turn the flow into a sequence of {@link Optional}s and default to {@link Optional#empty()}: *


     * CompletionStage<Optional<T>> stage = source.map(Optional::of).firstStage(Optional.empty());
     * 
*
*
Backpressure:
*
The operator requests one item from upstream and then when received, cancels the upstream.
*
Scheduler:
*
{@code firstStage} does not operate by default on a particular {@link Scheduler}.
*
* @param defaultItem the item to signal if the upstream is empty * @return the new {@code CompletionStage} instance * @since 3.0.0 * @see #firstOrErrorStage() */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final CompletionStage firstStage(@Nullable T defaultItem) { return subscribeWith(new FlowableFirstStageSubscriber<>(true, defaultItem)); } /** * Signals the only expected upstream item (or the default item if the upstream is empty) * or signals {@link IllegalArgumentException} if the upstream has more than one item * via a {@link CompletionStage}. *

* *

* The upstream can be canceled by converting the resulting {@code CompletionStage} into * {@link CompletableFuture} via {@link CompletionStage#toCompletableFuture()} and * calling {@link CompletableFuture#cancel(boolean)} on it. * The upstream will be also cancelled if the resulting {@code CompletionStage} is converted to and * completed manually by {@link CompletableFuture#complete(Object)} or {@link CompletableFuture#completeExceptionally(Throwable)}. *

* {@code CompletionStage}s don't have a notion of emptiness and allow {@code null}s, therefore, one can either use * a {@code defaultItem} of {@code null} or turn the flow into a sequence of {@link Optional}s and default to {@link Optional#empty()}: *


     * CompletionStage<Optional<T>> stage = source.map(Optional::of).singleStage(Optional.empty());
     * 
*
*
Backpressure:
*
The operator requests two items from upstream and then when more than one item is received, cancels the upstream.
*
Scheduler:
*
{@code singleStage} does not operate by default on a particular {@link Scheduler}.
*
* @param defaultItem the item to signal if the upstream is empty * @return the new {@code CompletionStage} instance * @since 3.0.0 * @see #singleOrErrorStage() */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final CompletionStage singleStage(@Nullable T defaultItem) { return subscribeWith(new FlowableSingleStageSubscriber<>(true, defaultItem)); } /** * Signals the last upstream item (or the default item if the upstream is empty) via * a {@link CompletionStage}. *

* *

* The upstream can be canceled by converting the resulting {@code CompletionStage} into * {@link CompletableFuture} via {@link CompletionStage#toCompletableFuture()} and * calling {@link CompletableFuture#cancel(boolean)} on it. * The upstream will be also cancelled if the resulting {@code CompletionStage} is converted to and * completed manually by {@link CompletableFuture#complete(Object)} or {@link CompletableFuture#completeExceptionally(Throwable)}. *

* {@code CompletionStage}s don't have a notion of emptiness and allow {@code null}s, therefore, one can either use * a {@code defaultItem} of {@code null} or turn the flow into a sequence of {@link Optional}s and default to {@link Optional#empty()}: *


     * CompletionStage<Optional<T>> stage = source.map(Optional::of).lastStage(Optional.empty());
     * 
*
*
Backpressure:
*
The operator requests an unbounded number of items from the upstream.
*
Scheduler:
*
{@code lastStage} does not operate by default on a particular {@link Scheduler}.
*
* @param defaultItem the item to signal if the upstream is empty * @return the new {@code CompletionStage} instance * @since 3.0.0 * @see #lastOrErrorStage() */ @CheckReturnValue @BackpressureSupport(BackpressureKind.UNBOUNDED_IN) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final CompletionStage lastStage(@Nullable T defaultItem) { return subscribeWith(new FlowableLastStageSubscriber<>(true, defaultItem)); } /** * Signals the first upstream item or a {@link NoSuchElementException} if the upstream is empty via * a {@link CompletionStage}. *

* *

* The upstream can be canceled by converting the resulting {@code CompletionStage} into * {@link CompletableFuture} via {@link CompletionStage#toCompletableFuture()} and * calling {@link CompletableFuture#cancel(boolean)} on it. * The upstream will be also cancelled if the resulting {@code CompletionStage} is converted to and * completed manually by {@link CompletableFuture#complete(Object)} or {@link CompletableFuture#completeExceptionally(Throwable)}. *

*
Backpressure:
*
The operator requests one item from upstream and then when received, cancels the upstream.
*
Scheduler:
*
{@code firstOrErrorStage} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code CompletionStage} instance * @since 3.0.0 * @see #firstStage(Object) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final CompletionStage firstOrErrorStage() { return subscribeWith(new FlowableFirstStageSubscriber<>(false, null)); } /** * Signals the only expected upstream item, a {@link NoSuchElementException} if the upstream is empty * or signals {@link IllegalArgumentException} if the upstream has more than one item * via a {@link CompletionStage}. *

* *

* The upstream can be canceled by converting the resulting {@code CompletionStage} into * {@link CompletableFuture} via {@link CompletionStage#toCompletableFuture()} and * calling {@link CompletableFuture#cancel(boolean)} on it. * The upstream will be also cancelled if the resulting {@code CompletionStage} is converted to and * completed manually by {@link CompletableFuture#complete(Object)} or {@link CompletableFuture#completeExceptionally(Throwable)}. *

*
Backpressure:
*
The operator requests two items from upstream and then when more than one item is received, cancels the upstream.
*
Scheduler:
*
{@code singleOrErrorStage} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code CompletionStage} instance * @since 3.0.0 * @see #singleStage(Object) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final CompletionStage singleOrErrorStage() { return subscribeWith(new FlowableSingleStageSubscriber<>(false, null)); } /** * Signals the last upstream item or a {@link NoSuchElementException} if the upstream is empty via * a {@link CompletionStage}. *

* *

* The upstream can be canceled by converting the resulting {@code CompletionStage} into * {@link CompletableFuture} via {@link CompletionStage#toCompletableFuture()} and * calling {@link CompletableFuture#cancel(boolean)} on it. * The upstream will be also cancelled if the resulting {@code CompletionStage} is converted to and * completed manually by {@link CompletableFuture#complete(Object)} or {@link CompletableFuture#completeExceptionally(Throwable)}. *

*
Backpressure:
*
The operator requests an unbounded number of items from the upstream.
*
Scheduler:
*
{@code lastOrErrorStage} does not operate by default on a particular {@link Scheduler}.
*
* @return the new {@code CompletionStage} instance * @since 3.0.0 * @see #lastStage(Object) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final CompletionStage lastOrErrorStage() { return subscribeWith(new FlowableLastStageSubscriber<>(false, null)); } /** * Creates a sequential {@link Stream} to consume or process this {@code Flowable} in a blocking manner via * the Java {@code Stream} API. *

* *

* Cancellation of the upstream is done via {@link Stream#close()}, therefore, it is strongly recommended the * consumption is performed within a try-with-resources construct: *


     * Flowable<Integer> source = Flowable.range(1, 10)
     *        .subscribeOn(Schedulers.computation());
     *
     * try (Stream<Integer> stream = source.blockingStream()) {
     *     stream.limit(3).forEach(System.out::println);
     * }
     * 
*
*
Backpressure:
*
The operator requests {@link #bufferSize()} amount upfront and 75% of it after each 75% of the amount received.
*
Scheduler:
*
{@code blockingStream} does not operate by default on a particular {@link Scheduler}.
*
* * @return the new {@code Stream} instance * @since 3.0.0 * @see #blockingStream(int) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Stream blockingStream() { return blockingStream(bufferSize()); } /** * Creates a sequential {@link Stream} to consume or process this {@code Flowable} in a blocking manner via * the Java {@code Stream} API. *

* *

* Cancellation of the upstream is done via {@link Stream#close()}, therefore, it is strongly recommended the * consumption is performed within a try-with-resources construct: *


     * Flowable<Integer> source = Flowable.range(1, 10)
     *        .subscribeOn(Schedulers.computation());
     *
     * try (Stream<Integer> stream = source.blockingStream(4)) {
     *     stream.limit(3).forEach(System.out::println);
     * }
     * 
*
*
Backpressure:
*
The operator requests the given {@code prefetch} amount upfront and 75% of it after each 75% of the amount received.
*
Scheduler:
*
{@code blockingStream} does not operate by default on a particular {@link Scheduler}.
*
* * @param prefetch the number of items to request from the upstream to limit the number of * in-flight items and item generation. * @return the new {@code Stream} instance * @throws IllegalArgumentException if {@code prefetch} is non-positive * @since 3.0.0 */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final Stream blockingStream(int prefetch) { Iterator iterator = blockingIterable(prefetch).iterator(); return StreamSupport.stream(Spliterators.spliteratorUnknownSize(iterator, 0), false) .onClose(((Disposable) iterator)::dispose); } /** * Maps each upstream item into a {@link Stream} and emits the {@code Stream}'s items to the downstream in a sequential fashion. *

* *

* Due to the blocking and sequential nature of Java {@code Stream}s, the streams are mapped and consumed in a sequential fashion * without interleaving (unlike a more general {@link #flatMap(Function)}). Therefore, {@code flatMapStream} and * {@code concatMapStream} are identical operators and are provided as aliases. *

* The operator closes the {@code Stream} upon cancellation and when it terminates. Any exceptions raised when * closing a {@code Stream} are routed to the global error handler ({@link RxJavaPlugins#onError(Throwable)}. * If a {@code Stream} should not be closed, turn it into an {@link Iterable} and use {@link #concatMapIterable(Function)}: *


     * source.concatMapIterable(v -> createStream(v)::iterator);
     * 
*

* Note that {@code Stream}s can be consumed only once; any subsequent attempt to consume a {@code Stream} * will result in an {@link IllegalStateException}. *

* Primitive streams are not supported and items have to be boxed manually (e.g., via {@link IntStream#boxed()}): *


     * source.concatMapStream(v -> IntStream.rangeClosed(v + 1, v + 10).boxed());
     * 
*

* {@code Stream} does not support concurrent usage so creating and/or consuming the same instance multiple times * from multiple threads can lead to undefined behavior. *

*
Backpressure:
*
The operator honors the downstream backpressure and consumes the inner stream only on demand. The operator * prefetches {@link #bufferSize} items of the upstream (then 75% of it after the 75% received) * and caches them until they are ready to be mapped into {@code Stream}s * after the current {@code Stream} has been consumed.
*
Scheduler:
*
{@code concatMapStream} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the {@code Stream}s and the result * @param mapper the function that receives an upstream item and should return a {@code Stream} whose elements * will be emitted to the downstream * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @since 3.0.0 * @see #concatMap(Function) * @see #concatMapIterable(Function) * @see #concatMapStream(Function, int) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapStream(@NonNull Function> mapper) { return flatMapStream(mapper, bufferSize()); } /** * Maps each upstream item into a {@link Stream} and emits the {@code Stream}'s items to the downstream in a sequential fashion. *

* *

* Due to the blocking and sequential nature of Java {@code Stream}s, the streams are mapped and consumed in a sequential fashion * without interleaving (unlike a more general {@link #flatMap(Function)}). Therefore, {@code flatMapStream} and * {@code concatMapStream} are identical operators and are provided as aliases. *

* The operator closes the {@code Stream} upon cancellation and when it terminates. Any exceptions raised when * closing a {@code Stream} are routed to the global error handler ({@link RxJavaPlugins#onError(Throwable)}. * If a {@code Stream} should not be closed, turn it into an {@link Iterable} and use {@link #concatMapIterable(Function, int)}: *


     * source.concatMapIterable(v -> createStream(v)::iterator, 32);
     * 
*

* Note that {@code Stream}s can be consumed only once; any subsequent attempt to consume a {@code Stream} * will result in an {@link IllegalStateException}. *

* Primitive streams are not supported and items have to be boxed manually (e.g., via {@link IntStream#boxed()}): *


     * source.concatMapStream(v -> IntStream.rangeClosed(v + 1, v + 10).boxed(), 32);
     * 
*

* {@code Stream} does not support concurrent usage so creating and/or consuming the same instance multiple times * from multiple threads can lead to undefined behavior. *

*
Backpressure:
*
The operator honors the downstream backpressure and consumes the inner stream only on demand. The operator * prefetches the given amount of upstream items and caches them until they are ready to be mapped into {@code Stream}s * after the current {@code Stream} has been consumed.
*
Scheduler:
*
{@code concatMapStream} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the {@code Stream}s and the result * @param mapper the function that receives an upstream item and should return a {@code Stream} whose elements * will be emitted to the downstream * @param prefetch the number of upstream items to request upfront, then 75% of this amount after each 75% upstream items received * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @since 3.0.0 * @see #concatMap(Function, int) * @see #concatMapIterable(Function, int) * @see #flatMapStream(Function, int) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable concatMapStream(@NonNull Function> mapper, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableFlatMapStream<>(this, mapper, prefetch)); } /** * Maps each upstream item into a {@link Stream} and emits the {@code Stream}'s items to the downstream in a sequential fashion. *

* *

* Due to the blocking and sequential nature of Java {@code Stream}s, the streams are mapped and consumed in a sequential fashion * without interleaving (unlike a more general {@link #flatMap(Function)}). Therefore, {@code flatMapStream} and * {@code concatMapStream} are identical operators and are provided as aliases. *

* The operator closes the {@code Stream} upon cancellation and when it terminates. Any exceptions raised when * closing a {@code Stream} are routed to the global error handler ({@link RxJavaPlugins#onError(Throwable)}. * If a {@code Stream} should not be closed, turn it into an {@link Iterable} and use {@link #flatMapIterable(Function)}: *


     * source.flatMapIterable(v -> createStream(v)::iterator);
     * 
*

* Note that {@code Stream}s can be consumed only once; any subsequent attempt to consume a {@code Stream} * will result in an {@link IllegalStateException}. *

* Primitive streams are not supported and items have to be boxed manually (e.g., via {@link IntStream#boxed()}): *


     * source.flatMapStream(v -> IntStream.rangeClosed(v + 1, v + 10).boxed());
     * 
*

* {@code Stream} does not support concurrent usage so creating and/or consuming the same instance multiple times * from multiple threads can lead to undefined behavior. *

*
Backpressure:
*
The operator honors the downstream backpressure and consumes the inner stream only on demand. The operator * prefetches {@link #bufferSize} items of the upstream (then 75% of it after the 75% received) * and caches them until they are ready to be mapped into {@code Stream}s * after the current {@code Stream} has been consumed.
*
Scheduler:
*
{@code flatMapStream} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the {@code Stream}s and the result * @param mapper the function that receives an upstream item and should return a {@code Stream} whose elements * will be emitted to the downstream * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @since 3.0.0 * @see #flatMap(Function) * @see #flatMapIterable(Function) * @see #flatMapStream(Function, int) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable flatMapStream(@NonNull Function> mapper) { return flatMapStream(mapper, bufferSize()); } /** * Maps each upstream item into a {@link Stream} and emits the {@code Stream}'s items to the downstream in a sequential fashion. *

* *

* Due to the blocking and sequential nature of Java {@code Stream}s, the streams are mapped and consumed in a sequential fashion * without interleaving (unlike a more general {@link #flatMap(Function)}). Therefore, {@code flatMapStream} and * {@code concatMapStream} are identical operators and are provided as aliases. *

* The operator closes the {@code Stream} upon cancellation and when it terminates. Any exceptions raised when * closing a {@code Stream} are routed to the global error handler ({@link RxJavaPlugins#onError(Throwable)}. * If a {@code Stream} should not be closed, turn it into an {@link Iterable} and use {@link #flatMapIterable(Function, int)}: *


     * source.flatMapIterable(v -> createStream(v)::iterator, 32);
     * 
*

* Note that {@code Stream}s can be consumed only once; any subsequent attempt to consume a {@code Stream} * will result in an {@link IllegalStateException}. *

* Primitive streams are not supported and items have to be boxed manually (e.g., via {@link IntStream#boxed()}): *


     * source.flatMapStream(v -> IntStream.rangeClosed(v + 1, v + 10).boxed(), 32);
     * 
*

* {@code Stream} does not support concurrent usage so creating and/or consuming the same instance multiple times * from multiple threads can lead to undefined behavior. *

*
Backpressure:
*
The operator honors the downstream backpressure and consumes the inner stream only on demand. The operator * prefetches the given amount of upstream items and caches them until they are ready to be mapped into {@code Stream}s * after the current {@code Stream} has been consumed.
*
Scheduler:
*
{@code flatMapStream} does not operate by default on a particular {@link Scheduler}.
*
* * @param the element type of the {@code Stream}s and the result * @param mapper the function that receives an upstream item and should return a {@code Stream} whose elements * will be emitted to the downstream * @param prefetch the number of upstream items to request upfront, then 75% of this amount after each 75% upstream items received * @return the new {@code Flowable} instance * @throws NullPointerException if {@code mapper} is {@code null} * @throws IllegalArgumentException if {@code prefetch} is non-positive * @since 3.0.0 * @see #flatMap(Function, int) * @see #flatMapIterable(Function, int) * @see #concatMapStream(Function, int) */ @CheckReturnValue @BackpressureSupport(BackpressureKind.FULL) @SchedulerSupport(SchedulerSupport.NONE) @NonNull public final <@NonNull R> Flowable flatMapStream(@NonNull Function> mapper, int prefetch) { Objects.requireNonNull(mapper, "mapper is null"); ObjectHelper.verifyPositive(prefetch, "prefetch"); return RxJavaPlugins.onAssembly(new FlowableFlatMapStream<>(this, mapper, prefetch)); } }