gwtrpc.shaded.com.google.common.util.concurrent.Futures Maven / Gradle / Ivy
/*
* Copyright (C) 2006 The Guava Authors
*
* 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 com.google.common.util.concurrent;
import static com.google.common.base.Preconditions.checkNotNull;
import static com.google.common.base.Preconditions.checkState;
import static com.google.common.util.concurrent.MoreExecutors.directExecutor;
import static com.google.common.util.concurrent.Uninterruptibles.getUninterruptibly;
import com.google.common.annotations.Beta;
import com.google.common.annotations.GwtCompatible;
import com.google.common.annotations.GwtIncompatible;
import com.google.common.base.Function;
import com.google.common.base.MoreObjects;
import com.google.common.base.Preconditions;
import com.google.common.collect.ImmutableList;
import com.google.common.util.concurrent.CollectionFuture.ListFuture;
import com.google.common.util.concurrent.ImmediateFuture.ImmediateCancelledFuture;
import com.google.common.util.concurrent.ImmediateFuture.ImmediateFailedCheckedFuture;
import com.google.common.util.concurrent.ImmediateFuture.ImmediateFailedFuture;
import com.google.common.util.concurrent.ImmediateFuture.ImmediateSuccessfulCheckedFuture;
import com.google.common.util.concurrent.ImmediateFuture.ImmediateSuccessfulFuture;
import com.google.errorprone.annotations.CanIgnoreReturnValue;
import java.util.Collection;
import java.util.List;
import java.util.concurrent.Callable;
import java.util.concurrent.CancellationException;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.Executor;
import java.util.concurrent.Future;
import java.util.concurrent.ScheduledExecutorService;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;
import java.util.concurrent.atomic.AtomicInteger;
import javax.annotation.Nullable;
/**
* Static utility methods pertaining to the {@link Future} interface.
*
* Many of these methods use the {@link ListenableFuture} API; consult the Guava User Guide
* article on {@code
* ListenableFuture}.
*
*
The main purpose of {@code ListenableFuture} is to help you chain together a graph of
* asynchronous operations. You can chain them together manually with calls to methods like {@link
* Futures#transform(ListenableFuture, Function, Executor) Futures.transform}, but you will often
* find it easier to use a framework. Frameworks automate the process, often adding features like
* monitoring, debugging, and cancellation. Examples of frameworks include:
*
*
* - Dagger Producers
*
*
* If you do chain your operations manually, you may want to use {@link FluentFuture}.
*
* @author Kevin Bourrillion
* @author Nishant Thakkar
* @author Sven Mawson
* @since 1.0
*/
@Beta
@GwtCompatible(emulated = true)
public final class Futures extends GwtFuturesCatchingSpecialization {
// A note on memory visibility.
// Many of the utilities in this class (transform, withFallback, withTimeout, asList, combine)
// have two requirements that significantly complicate their design.
// 1. Cancellation should propagate from the returned future to the input future(s).
// 2. The returned futures shouldn't unnecessarily 'pin' their inputs after completion.
//
// A consequence of these requirements is that the delegate futures cannot be stored in
// final fields.
//
// For simplicity the rest of this description will discuss Futures.catching since it is the
// simplest instance, though very similar descriptions apply to many other classes in this file.
//
// In the constructor of AbstractCatchingFuture, the delegate future is assigned to a field
// 'inputFuture'. That field is non-final and non-volatile. There are 2 places where the
// 'inputFuture' field is read and where we will have to consider visibility of the write
// operation in the constructor.
//
// 1. In the listener that performs the callback. In this case it is fine since inputFuture is
// assigned prior to calling addListener, and addListener happens-before any invocation of the
// listener. Notably, this means that 'volatile' is unnecessary to make 'inputFuture' visible
// to the listener.
//
// 2. In done() where we may propagate cancellation to the input. In this case it is _not_ fine.
// There is currently nothing that enforces that the write to inputFuture in the constructor is
// visible to done(). This is because there is no happens before edge between the write and a
// (hypothetical) unsafe read by our caller. Note: adding 'volatile' does not fix this issue,
// it would just add an edge such that if done() observed non-null, then it would also
// definitely observe all earlier writes, but we still have no guarantee that done() would see
// the inital write (just stronger guarantees if it does).
//
// See: http://cs.oswego.edu/pipermail/concurrency-interest/2015-January/013800.html
// For a (long) discussion about this specific issue and the general futility of life.
//
// For the time being we are OK with the problem discussed above since it requires a caller to
// introduce a very specific kind of data-race. And given the other operations performed by these
// methods that involve volatile read/write operations, in practice there is no issue. Also, the
// way in such a visibility issue would surface is most likely as a failure of cancel() to
// propagate to the input. Cancellation propagation is fundamentally racy so this is fine.
//
// Future versions of the JMM may revise safe construction semantics in such a way that we can
// safely publish these objects and we won't need this whole discussion.
// TODO(user,lukes): consider adding volatile to all these fields since in current known JVMs
// that should resolve the issue. This comes at the cost of adding more write barriers to the
// implementations.
private Futures() {}
/**
* Creates a {@link CheckedFuture} out of a normal {@link ListenableFuture} and a {@link Function}
* that maps from {@link Exception} instances into the appropriate checked type.
*
*
Warning: We recommend against using {@code CheckedFuture} in new projects. {@code
* CheckedFuture} is difficult to build libraries atop. {@code CheckedFuture} ports of methods
* like {@link Futures#transformAsync} have historically had bugs, and some of these bugs are
* necessary, unavoidable consequences of the {@code CheckedFuture} API. Additionally, {@code
* CheckedFuture} encourages users to take exceptions from one thread and rethrow them in another,
* producing confusing stack traces.
*
*
The given mapping function will be applied to an {@link InterruptedException}, a {@link
* CancellationException}, or an {@link ExecutionException}. See {@link Future#get()} for details
* on the exceptions thrown.
*
* @since 9.0 (source-compatible since 1.0)
* @deprecated {@link CheckedFuture} cannot properly support the chained operations that are the
* primary goal of {@link ListenableFuture}. {@code CheckedFuture} also encourages users to
* rethrow exceptions from one thread in another thread, producing misleading stack traces.
* Additionally, it has a surprising policy about which exceptions to map and which to leave
* untouched. Guava users who want a {@code CheckedFuture} can fork the classes for their own
* use, possibly specializing them to the particular exception type they use. We recommend
* that most people use {@code ListenableFuture} and perform any exception wrapping
* themselves. This method is scheduled for removal from Guava in February 2018.
*/
@Deprecated
@GwtIncompatible // TODO
public static CheckedFuture makeChecked(
ListenableFuture future, Function mapper) {
return new MappingCheckedFuture<>(checkNotNull(future), mapper);
}
/**
* Creates a {@code ListenableFuture} which has its value set immediately upon construction. The
* getters just return the value. This {@code Future} can't be canceled or timed out and its
* {@code isDone()} method always returns {@code true}.
*/
public static ListenableFuture immediateFuture(@Nullable V value) {
if (value == null) {
// This cast is safe because null is assignable to V for all V (i.e. it is covariant)
@SuppressWarnings({"unchecked", "rawtypes"})
ListenableFuture typedNull = (ListenableFuture) ImmediateSuccessfulFuture.NULL;
return typedNull;
}
return new ImmediateSuccessfulFuture(value);
}
/**
* Returns a {@code CheckedFuture} which has its value set immediately upon construction.
*
* The returned {@code Future} can't be cancelled, and its {@code isDone()} method always
* returns {@code true}. Calling {@code get()} or {@code checkedGet()} will immediately return the
* provided value.
*
* @deprecated {@link CheckedFuture} cannot properly support the chained operations that are the
* primary goal of {@link ListenableFuture}. {@code CheckedFuture} also encourages users to
* rethrow exceptions from one thread in another thread, producing misleading stack traces.
* Additionally, it has a surprising policy about which exceptions to map and which to leave
* untouched. Guava users who want a {@code CheckedFuture} can fork the classes for their own
* use, possibly specializing them to the particular exception type they use. We recommend
* that most people use {@code ListenableFuture} and perform any exception wrapping
* themselves. This method is scheduled for removal from Guava in February 2018.
*/
@Deprecated
@GwtIncompatible // TODO
public static CheckedFuture immediateCheckedFuture(
@Nullable V value) {
return new ImmediateSuccessfulCheckedFuture<>(value);
}
/**
* Returns a {@code ListenableFuture} which has an exception set immediately upon construction.
*
* The returned {@code Future} can't be cancelled, and its {@code isDone()} method always
* returns {@code true}. Calling {@code get()} will immediately throw the provided {@code
* Throwable} wrapped in an {@code ExecutionException}.
*/
public static ListenableFuture immediateFailedFuture(Throwable throwable) {
checkNotNull(throwable);
return new ImmediateFailedFuture(throwable);
}
/**
* Creates a {@code ListenableFuture} which is cancelled immediately upon construction, so that
* {@code isCancelled()} always returns {@code true}.
*
* @since 14.0
*/
public static ListenableFuture immediateCancelledFuture() {
return new ImmediateCancelledFuture();
}
/**
* Returns a {@code CheckedFuture} which has an exception set immediately upon construction.
*
* The returned {@code Future} can't be cancelled, and its {@code isDone()} method always
* returns {@code true}. Calling {@code get()} will immediately throw the provided {@code
* Exception} wrapped in an {@code ExecutionException}, and calling {@code checkedGet()} will
* throw the provided exception itself.
*
* @deprecated {@link CheckedFuture} cannot properly support the chained operations that are the
* primary goal of {@link ListenableFuture}. {@code CheckedFuture} also encourages users to
* rethrow exceptions from one thread in another thread, producing misleading stack traces.
* Additionally, it has a surprising policy about which exceptions to map and which to leave
* untouched. Guava users who want a {@code CheckedFuture} can fork the classes for their own
* use, possibly specializing them to the particular exception type they use. We recommend
* that most people use {@code ListenableFuture} and perform any exception wrapping
* themselves. This method is scheduled for removal from Guava in February 2018.
*/
@Deprecated
@GwtIncompatible // TODO
public static CheckedFuture immediateFailedCheckedFuture(
X exception) {
checkNotNull(exception);
return new ImmediateFailedCheckedFuture<>(exception);
}
/**
* Executes {@code callable} on the specified {@code executor}, returning a {@code Future}.
*
* @throws RejectedExecutionException if the task cannot be scheduled for execution
* @since 23.0
*/
public static ListenableFuture submitAsync(AsyncCallable callable, Executor executor) {
TrustedListenableFutureTask task = TrustedListenableFutureTask.create(callable);
executor.execute(task);
return task;
}
/**
* Schedules {@code callable} on the specified {@code executor}, returning a {@code Future}.
*
* @throws RejectedExecutionException if the task cannot be scheduled for execution
* @since 23.0
*/
@GwtIncompatible // java.util.concurrent.ScheduledExecutorService
public static ListenableFuture scheduleAsync(
AsyncCallable callable,
long delay,
TimeUnit timeUnit,
ScheduledExecutorService executorService) {
TrustedListenableFutureTask task = TrustedListenableFutureTask.create(callable);
final Future scheduled = executorService.schedule(task, delay, timeUnit);
task.addListener(
new Runnable() {
@Override
public void run() {
// Don't want to interrupt twice
scheduled.cancel(false);
}
},
directExecutor());
return task;
}
/**
* Returns a {@code Future} whose result is taken from the given primary {@code input} or, if the
* primary input fails with the given {@code exceptionType}, from the result provided by the
* {@code fallback}. {@link Function#apply} is not invoked until the primary input has failed, so
* if the primary input succeeds, it is never invoked. If, during the invocation of {@code
* fallback}, an exception is thrown, this exception is used as the result of the output {@code
* Future}.
*
* Usage example:
*
*
{@code
* ListenableFuture fetchCounterFuture = ...;
*
* // Falling back to a zero counter in case an exception happens when
* // processing the RPC to fetch counters.
* ListenableFuture faultTolerantFuture = Futures.catching(
* fetchCounterFuture, FetchException.class, x -> 0);
* }
*
* This overload, which does not accept an executor, uses {@code directExecutor}, a dangerous
* choice in some cases. See the discussion in the {@link ListenableFuture#addListener
* ListenableFuture.addListener} documentation. All its warnings about heavyweight listeners are
* also applicable to heavyweight functions passed to this method.
*
* @param input the primary input {@code Future}
* @param exceptionType the exception type that triggers use of {@code fallback}. The exception
* type is matched against the input's exception. "The input's exception" means the cause of
* the {@link ExecutionException} thrown by {@code input.get()} or, if {@code get()} throws a
* different kind of exception, that exception itself. To avoid hiding bugs and other
* unrecoverable errors, callers should prefer more specific types, avoiding {@code
* Throwable.class} in particular.
* @param fallback the {@link Function} to be called if {@code input} fails with the expected
* exception type. The function's argument is the input's exception. "The input's exception"
* means the cause of the {@link ExecutionException} thrown by {@code input.get()} or, if
* {@code get()} throws a different kind of exception, that exception itself.
* @since 19.0
* @deprecated Use {@linkplain #catching(ListenableFuture, Class, Function, Executor) the overload
* that requires an executor}. For identical behavior, pass {@link
* MoreExecutors#directExecutor}, but consider whether another executor would be safer, as
* discussed in the {@link ListenableFuture#addListener ListenableFuture.addListener}
* documentation. This method is scheduled to be removed in April 2018.
*/
@Deprecated
@Partially.GwtIncompatible("AVAILABLE but requires exceptionType to be Throwable.class")
public static ListenableFuture catching(
ListenableFuture input,
Class exceptionType,
Function fallback) {
return AbstractCatchingFuture.create(input, exceptionType, fallback, directExecutor());
}
/**
* Returns a {@code Future} whose result is taken from the given primary {@code input} or, if the
* primary input fails with the given {@code exceptionType}, from the result provided by the
* {@code fallback}. {@link Function#apply} is not invoked until the primary input has failed, so
* if the primary input succeeds, it is never invoked. If, during the invocation of {@code
* fallback}, an exception is thrown, this exception is used as the result of the output {@code
* Future}.
*
* Usage example:
*
*
{@code
* ListenableFuture fetchCounterFuture = ...;
*
* // Falling back to a zero counter in case an exception happens when
* // processing the RPC to fetch counters.
* ListenableFuture faultTolerantFuture = Futures.catching(
* fetchCounterFuture, FetchException.class, x -> 0, directExecutor());
* }
*
* When selecting an executor, note that {@code directExecutor} is dangerous in some cases. See
* the discussion in the {@link ListenableFuture#addListener ListenableFuture.addListener}
* documentation. All its warnings about heavyweight listeners are also applicable to heavyweight
* functions passed to this method.
*
* @param input the primary input {@code Future}
* @param exceptionType the exception type that triggers use of {@code fallback}. The exception
* type is matched against the input's exception. "The input's exception" means the cause of
* the {@link ExecutionException} thrown by {@code input.get()} or, if {@code get()} throws a
* different kind of exception, that exception itself. To avoid hiding bugs and other
* unrecoverable errors, callers should prefer more specific types, avoiding {@code
* Throwable.class} in particular.
* @param fallback the {@link Function} to be called if {@code input} fails with the expected
* exception type. The function's argument is the input's exception. "The input's exception"
* means the cause of the {@link ExecutionException} thrown by {@code input.get()} or, if
* {@code get()} throws a different kind of exception, that exception itself.
* @param executor the executor that runs {@code fallback} if {@code input} fails
* @since 19.0
*/
@Partially.GwtIncompatible("AVAILABLE but requires exceptionType to be Throwable.class")
public static ListenableFuture catching(
ListenableFuture input,
Class exceptionType,
Function fallback,
Executor executor) {
return AbstractCatchingFuture.create(input, exceptionType, fallback, executor);
}
/**
* Returns a {@code Future} whose result is taken from the given primary {@code input} or, if the
* primary input fails with the given {@code exceptionType}, from the result provided by the
* {@code fallback}. {@link AsyncFunction#apply} is not invoked until the primary input has
* failed, so if the primary input succeeds, it is never invoked. If, during the invocation of
* {@code fallback}, an exception is thrown, this exception is used as the result of the output
* {@code Future}.
*
* Usage examples:
*
*
{@code
* ListenableFuture fetchCounterFuture = ...;
*
* // Falling back to a zero counter in case an exception happens when
* // processing the RPC to fetch counters.
* ListenableFuture faultTolerantFuture = Futures.catchingAsync(
* fetchCounterFuture, FetchException.class, x -> immediateFuture(0));
* }
*
* The fallback can also choose to propagate the original exception when desired:
*
*
{@code
* ListenableFuture fetchCounterFuture = ...;
*
* // Falling back to a zero counter only in case the exception was a
* // TimeoutException.
* ListenableFuture faultTolerantFuture = Futures.catchingAsync(
* fetchCounterFuture,
* FetchException.class,
* e -> {
* if (omitDataOnFetchFailure) {
* return immediateFuture(0);
* }
* throw e;
* });
* }
*
* This overload, which does not accept an executor, uses {@code directExecutor}, a dangerous
* choice in some cases. See the discussion in the {@link ListenableFuture#addListener
* ListenableFuture.addListener} documentation. All its warnings about heavyweight listeners are
* also applicable to heavyweight functions passed to this method. (Specifically, {@code
* directExecutor} functions should avoid heavyweight operations inside {@code
* AsyncFunction.apply}. Any heavyweight operations should occur in other threads responsible for
* completing the returned {@code Future}.)
*
* @param input the primary input {@code Future}
* @param exceptionType the exception type that triggers use of {@code fallback}. The exception
* type is matched against the input's exception. "The input's exception" means the cause of
* the {@link ExecutionException} thrown by {@code input.get()} or, if {@code get()} throws a
* different kind of exception, that exception itself. To avoid hiding bugs and other
* unrecoverable errors, callers should prefer more specific types, avoiding {@code
* Throwable.class} in particular.
* @param fallback the {@link AsyncFunction} to be called if {@code input} fails with the expected
* exception type. The function's argument is the input's exception. "The input's exception"
* means the cause of the {@link ExecutionException} thrown by {@code input.get()} or, if
* {@code get()} throws a different kind of exception, that exception itself.
* @since 19.0 (similar functionality in 14.0 as {@code withFallback})
* @deprecated Use {@linkplain #catchingAsync(ListenableFuture, Class, AsyncFunction, Executor)
* the overload that requires an executor}. For identical behavior, pass {@link
* MoreExecutors#directExecutor}, but consider whether another executor would be safer, as
* discussed in the {@link ListenableFuture#addListener ListenableFuture.addListener}
* documentation. This method is scheduled to be removed in April 2018.
*/
@CanIgnoreReturnValue // TODO(kak): @CheckReturnValue
@Deprecated
@Partially.GwtIncompatible("AVAILABLE but requires exceptionType to be Throwable.class")
public static ListenableFuture catchingAsync(
ListenableFuture input,
Class exceptionType,
AsyncFunction fallback) {
return AbstractCatchingFuture.create(input, exceptionType, fallback, directExecutor());
}
/**
* Returns a {@code Future} whose result is taken from the given primary {@code input} or, if the
* primary input fails with the given {@code exceptionType}, from the result provided by the
* {@code fallback}. {@link AsyncFunction#apply} is not invoked until the primary input has
* failed, so if the primary input succeeds, it is never invoked. If, during the invocation of
* {@code fallback}, an exception is thrown, this exception is used as the result of the output
* {@code Future}.
*
* Usage examples:
*
*
{@code
* ListenableFuture fetchCounterFuture = ...;
*
* // Falling back to a zero counter in case an exception happens when
* // processing the RPC to fetch counters.
* ListenableFuture faultTolerantFuture = Futures.catchingAsync(
* fetchCounterFuture, FetchException.class, x -> immediateFuture(0), directExecutor());
* }
*
* The fallback can also choose to propagate the original exception when desired:
*
*
{@code
* ListenableFuture fetchCounterFuture = ...;
*
* // Falling back to a zero counter only in case the exception was a
* // TimeoutException.
* ListenableFuture faultTolerantFuture = Futures.catchingAsync(
* fetchCounterFuture,
* FetchException.class,
* e -> {
* if (omitDataOnFetchFailure) {
* return immediateFuture(0);
* }
* throw e;
* },
* directExecutor());
* }
*
* When selecting an executor, note that {@code directExecutor} is dangerous in some cases. See
* the discussion in the {@link ListenableFuture#addListener ListenableFuture.addListener}
* documentation. All its warnings about heavyweight listeners are also applicable to heavyweight
* functions passed to this method. (Specifically, {@code directExecutor} functions should avoid
* heavyweight operations inside {@code AsyncFunction.apply}. Any heavyweight operations should
* occur in other threads responsible for completing the returned {@code Future}.)
*
* @param input the primary input {@code Future}
* @param exceptionType the exception type that triggers use of {@code fallback}. The exception
* type is matched against the input's exception. "The input's exception" means the cause of
* the {@link ExecutionException} thrown by {@code input.get()} or, if {@code get()} throws a
* different kind of exception, that exception itself. To avoid hiding bugs and other
* unrecoverable errors, callers should prefer more specific types, avoiding {@code
* Throwable.class} in particular.
* @param fallback the {@link AsyncFunction} to be called if {@code input} fails with the expected
* exception type. The function's argument is the input's exception. "The input's exception"
* means the cause of the {@link ExecutionException} thrown by {@code input.get()} or, if
* {@code get()} throws a different kind of exception, that exception itself.
* @param executor the executor that runs {@code fallback} if {@code input} fails
* @since 19.0 (similar functionality in 14.0 as {@code withFallback})
*/
@CanIgnoreReturnValue // TODO(kak): @CheckReturnValue
@Partially.GwtIncompatible("AVAILABLE but requires exceptionType to be Throwable.class")
public static ListenableFuture catchingAsync(
ListenableFuture input,
Class exceptionType,
AsyncFunction fallback,
Executor executor) {
return AbstractCatchingFuture.create(input, exceptionType, fallback, executor);
}
/**
* Returns a future that delegates to another but will finish early (via a {@link
* TimeoutException} wrapped in an {@link ExecutionException}) if the specified duration expires.
*
* The delegate future is interrupted and cancelled if it times out.
*
* @param delegate The future to delegate to.
* @param time when to timeout the future
* @param unit the time unit of the time parameter
* @param scheduledExecutor The executor service to enforce the timeout.
* @since 19.0
*/
@GwtIncompatible // java.util.concurrent.ScheduledExecutorService
public static ListenableFuture withTimeout(
ListenableFuture delegate,
long time,
TimeUnit unit,
ScheduledExecutorService scheduledExecutor) {
return TimeoutFuture.create(delegate, time, unit, scheduledExecutor);
}
/**
* Returns a new {@code Future} whose result is asynchronously derived from the result of the
* given {@code Future}. If the given {@code Future} fails, the returned {@code Future} fails with
* the same exception (and the function is not invoked).
*
* More precisely, the returned {@code Future} takes its result from a {@code Future} produced
* by applying the given {@code AsyncFunction} to the result of the original {@code Future}.
* Example usage:
*
*
{@code
* ListenableFuture rowKeyFuture = indexService.lookUp(query);
* ListenableFuture queryFuture =
* transformAsync(rowKeyFuture, dataService::readFuture);
* }
*
* This overload, which does not accept an executor, uses {@code directExecutor}, a dangerous
* choice in some cases. See the discussion in the {@link ListenableFuture#addListener
* ListenableFuture.addListener} documentation. All its warnings about heavyweight listeners are
* also applicable to heavyweight functions passed to this method. (Specifically, {@code
* directExecutor} functions should avoid heavyweight operations inside {@code
* AsyncFunction.apply}. Any heavyweight operations should occur in other threads responsible for
* completing the returned {@code Future}.)
*
*
The returned {@code Future} attempts to keep its cancellation state in sync with that of the
* input future and that of the future returned by the function. That is, if the returned {@code
* Future} is cancelled, it will attempt to cancel the other two, and if either of the other two
* is cancelled, the returned {@code Future} will receive a callback in which it will attempt to
* cancel itself.
*
* @param input The future to transform
* @param function A function to transform the result of the input future to the result of the
* output future
* @return A future that holds result of the function (if the input succeeded) or the original
* input's failure (if not)
* @since 19.0 (in 11.0 as {@code transform})
* @deprecated Use {@linkplain #transformAsync(ListenableFuture, AsyncFunction, Executor) the
* overload that requires an executor}. For identical behavior, pass {@link
* MoreExecutors#directExecutor}, but consider whether another executor would be safer, as
* discussed in the {@link ListenableFuture#addListener ListenableFuture.addListener}
* documentation. This method is scheduled to be removed in April 2018.
*/
@Deprecated
public static ListenableFuture transformAsync(
ListenableFuture input, AsyncFunction function) {
return AbstractTransformFuture.create(input, function, directExecutor());
}
/**
* Returns a new {@code Future} whose result is asynchronously derived from the result of the
* given {@code Future}. If the given {@code Future} fails, the returned {@code Future} fails with
* the same exception (and the function is not invoked).
*
* More precisely, the returned {@code Future} takes its result from a {@code Future} produced
* by applying the given {@code AsyncFunction} to the result of the original {@code Future}.
* Example usage:
*
*
{@code
* ListenableFuture rowKeyFuture = indexService.lookUp(query);
* ListenableFuture queryFuture =
* transformAsync(rowKeyFuture, dataService::readFuture, executor);
* }
*
* When selecting an executor, note that {@code directExecutor} is dangerous in some cases. See
* the discussion in the {@link ListenableFuture#addListener ListenableFuture.addListener}
* documentation. All its warnings about heavyweight listeners are also applicable to heavyweight
* functions passed to this method. (Specifically, {@code directExecutor} functions should avoid
* heavyweight operations inside {@code AsyncFunction.apply}. Any heavyweight operations should
* occur in other threads responsible for completing the returned {@code Future}.)
*
*
The returned {@code Future} attempts to keep its cancellation state in sync with that of the
* input future and that of the future returned by the chain function. That is, if the returned
* {@code Future} is cancelled, it will attempt to cancel the other two, and if either of the
* other two is cancelled, the returned {@code Future} will receive a callback in which it will
* attempt to cancel itself.
*
* @param input The future to transform
* @param function A function to transform the result of the input future to the result of the
* output future
* @param executor Executor to run the function in.
* @return A future that holds result of the function (if the input succeeded) or the original
* input's failure (if not)
* @since 19.0 (in 11.0 as {@code transform})
*/
public static ListenableFuture transformAsync(
ListenableFuture input,
AsyncFunction function,
Executor executor) {
return AbstractTransformFuture.create(input, function, executor);
}
/**
* Returns a new {@code Future} whose result is derived from the result of the given {@code
* Future}. If {@code input} fails, the returned {@code Future} fails with the same exception (and
* the function is not invoked). Example usage:
*
* {@code
* ListenableFuture queryFuture = ...;
* ListenableFuture> rowsFuture =
* transform(queryFuture, QueryResult::getRows);
* }
*
* This overload, which does not accept an executor, uses {@code directExecutor}, a dangerous
* choice in some cases. See the discussion in the {@link ListenableFuture#addListener
* ListenableFuture.addListener} documentation. All its warnings about heavyweight listeners are
* also applicable to heavyweight functions passed to this method.
*
*
The returned {@code Future} attempts to keep its cancellation state in sync with that of the
* input future. That is, if the returned {@code Future} is cancelled, it will attempt to cancel
* the input, and if the input is cancelled, the returned {@code Future} will receive a callback
* in which it will attempt to cancel itself.
*
*
An example use of this method is to convert a serializable object returned from an RPC into
* a POJO.
*
* @param input The future to transform
* @param function A Function to transform the results of the provided future to the results of
* the returned future. This will be run in the thread that notifies input it is complete.
* @return A future that holds result of the transformation.
* @since 9.0 (in 1.0 as {@code compose})
* @deprecated Use {@linkplain #transform(ListenableFuture, Function, Executor) the overload that
* requires an executor}. For identical behavior, pass {@link MoreExecutors#directExecutor},
* but consider whether another executor would be safer, as discussed in the {@link
* ListenableFuture#addListener ListenableFuture.addListener} documentation. This method is
* scheduled to be removed in April 2018.
*/
@Deprecated
public static ListenableFuture transform(
ListenableFuture input, Function function) {
return AbstractTransformFuture.create(input, function, directExecutor());
}
/**
* Returns a new {@code Future} whose result is derived from the result of the given {@code
* Future}. If {@code input} fails, the returned {@code Future} fails with the same exception (and
* the function is not invoked). Example usage:
*
* {@code
* ListenableFuture queryFuture = ...;
* ListenableFuture> rowsFuture =
* transform(queryFuture, QueryResult::getRows, executor);
* }
*
* When selecting an executor, note that {@code directExecutor} is dangerous in some cases. See
* the discussion in the {@link ListenableFuture#addListener ListenableFuture.addListener}
* documentation. All its warnings about heavyweight listeners are also applicable to heavyweight
* functions passed to this method.
*
*
The returned {@code Future} attempts to keep its cancellation state in sync with that of the
* input future. That is, if the returned {@code Future} is cancelled, it will attempt to cancel
* the input, and if the input is cancelled, the returned {@code Future} will receive a callback
* in which it will attempt to cancel itself.
*
*
An example use of this method is to convert a serializable object returned from an RPC into
* a POJO.
*
* @param input The future to transform
* @param function A Function to transform the results of the provided future to the results of
* the returned future.
* @param executor Executor to run the function in.
* @return A future that holds result of the transformation.
* @since 9.0 (in 2.0 as {@code compose})
*/
public static ListenableFuture transform(
ListenableFuture input, Function function, Executor executor) {
return AbstractTransformFuture.create(input, function, executor);
}
/**
* Like {@link #transform(ListenableFuture, Function, Executor)} except that the transformation
* {@code function} is invoked on each call to {@link Future#get() get()} on the returned future.
*
* The returned {@code Future} reflects the input's cancellation state directly, and any
* attempt to cancel the returned Future is likewise passed through to the input Future.
*
*
Note that calls to {@linkplain Future#get(long, TimeUnit) timed get} only apply the timeout
* to the execution of the underlying {@code Future}, not to the execution of the
* transformation function.
*
*
The primary audience of this method is callers of {@code transform} who don't have a {@code
* ListenableFuture} available and do not mind repeated, lazy function evaluation.
*
* @param input The future to transform
* @param function A Function to transform the results of the provided future to the results of
* the returned future.
* @return A future that returns the result of the transformation.
* @since 10.0
*/
@GwtIncompatible // TODO
public static Future lazyTransform(
final Future input, final Function function) {
checkNotNull(input);
checkNotNull(function);
return new Future() {
@Override
public boolean cancel(boolean mayInterruptIfRunning) {
return input.cancel(mayInterruptIfRunning);
}
@Override
public boolean isCancelled() {
return input.isCancelled();
}
@Override
public boolean isDone() {
return input.isDone();
}
@Override
public O get() throws InterruptedException, ExecutionException {
return applyTransformation(input.get());
}
@Override
public O get(long timeout, TimeUnit unit)
throws InterruptedException, ExecutionException, TimeoutException {
return applyTransformation(input.get(timeout, unit));
}
private O applyTransformation(I input) throws ExecutionException {
try {
return function.apply(input);
} catch (Throwable t) {
throw new ExecutionException(t);
}
}
};
}
/**
* Returns a new {@code ListenableFuture} whose result is the product of calling {@code get()} on
* the {@code Future} nested within the given {@code Future}, effectively chaining the futures one
* after the other. Example:
*
* {@code
* SettableFuture> nested = SettableFuture.create();
* ListenableFuture dereferenced = dereference(nested);
* }
*
* Most users will not need this method. To create a {@code Future} that completes with the
* result of another {@code Future}, create a {@link SettableFuture}, and call {@link
* SettableFuture#setFuture setFuture(otherFuture)} on it.
*
*
{@code dereference} has the same cancellation and execution semantics as {@link
* #transformAsync(ListenableFuture, AsyncFunction, Executor)}, in that the returned {@code
* Future} attempts to keep its cancellation state in sync with both the input {@code Future} and
* the nested {@code Future}. The transformation is very lightweight and therefore takes place in
* the same thread (either the thread that called {@code dereference}, or the thread in which the
* dereferenced future completes).
*
* @deprecated Use {@link #submitAsync(AsyncCallable, Executor)} or {@link
* SettableFuture#setFuture} instead. Or, if you're dereferencing the output of {@link
* #transform} or {@link #catching}, switch to {@link #transformAsync} or {@link
* #catchingAsync} (and likewise for similar APIs). If the cancellation of this method's
* output future races with completion of the outer input future, cancellation may not be
* propagated to the inner input future. This method is scheduled to be removed in January
* 2018.
* @param nested The nested future to transform.
* @return A future that holds result of the inner future.
* @since 13.0
*/
@Deprecated
@SuppressWarnings({"rawtypes", "unchecked"})
public static ListenableFuture dereference(
ListenableFuture> nested) {
return transformAsync(
(ListenableFuture) nested, (AsyncFunction) DEREFERENCER, directExecutor());
}
/**
* Helper {@code Function} for {@link #dereference}.
*/
private static final AsyncFunction, Object> DEREFERENCER =
new AsyncFunction, Object>() {
@Override
public ListenableFuture