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

groovyx.gpars.GParsExecutorsPool.groovy Maven / Gradle / Ivy

Go to download

The Groovy and Java high-level concurrency library offering actors, dataflow, CSP, agents, parallel collections, fork/join and more

There is a newer version: 1.2.1
Show newest version
// GPars - Groovy Parallel Systems
//
// Copyright © 2008-11  The original author or 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 groovyx.gpars

import groovyx.gpars.dataflow.DataflowVariable
import groovyx.gpars.util.PoolUtils
import java.util.concurrent.Callable
import java.util.concurrent.ExecutorService
import java.util.concurrent.Executors
import java.util.concurrent.Future
import java.util.concurrent.ThreadFactory
import java.util.concurrent.TimeUnit
import java.util.concurrent.atomic.AtomicInteger

/**
 * Enables a ExecutorService-based DSL on closures, objects and collections.
 * E.g.
 * GParsExecutorsPool.withPool(5) {ExecutorService service ->
 *     Collection result = [1, 2, 3, 4, 5].collectParallel({it * 10}.async())
 *     assertEquals(new HashSet([10, 20, 30, 40, 50]), new HashSet((Collection)result*.get()))
 *}*
 * GParsExecutorsPool.withPool(5) {ExecutorService service ->
 *     def result = [1, 2, 3, 4, 5].findParallel{Number number -> number > 2}*     assert result in [3, 4, 5]
 *}*
 * @author Vaclav Pech
 * Date: Oct 23, 2008
 */
class GParsExecutorsPool {

    /**
     * Maps threads to their appropriate thread pools
     */
    private static final ThreadLocalPools currentPoolStack = new ThreadLocalPools()

    /**
     * Caches the default pool size.
     */
    private static final int defaultPoolSize = PoolUtils.retrieveDefaultPoolSize()

    /**
     * Retrieves the pool assigned to the current thread.
     */
    protected static ExecutorService retrieveCurrentPool() {
        currentPoolStack.current
    }

    /**
     * Creates a new pool with the default size()
     */
    private static createPool() {
        return createPool(PoolUtils.retrieveDefaultPoolSize())
    }

    /**
     * Creates a new pool with the given size()
     */
    private static createPool(int poolSize) {
        return createPool(poolSize, createDefaultThreadFactory())
    }

    private static createPool(int poolSize, ThreadFactory threadFactory) {
        if (!(poolSize in 1..Integer.MAX_VALUE)) throw new IllegalArgumentException("Invalid value $poolSize for the pool size has been specified. Please supply a positive int number.")
        if (!threadFactory) throw new IllegalArgumentException("No value specified for threadFactory.")
        return Executors.newFixedThreadPool(poolSize, threadFactory)
    }

    private static ThreadFactory createDefaultThreadFactory() {
        return {Runnable runnable ->
            final Thread thread = new Thread(runnable)
            thread.daemon = false
            thread
        } as ThreadFactory
    }

    /**
     * Creates a new instance of ExecutorService, binds it to the current thread, enables the ExecutorService DSL
     * and runs the supplied closure.
     * It is an identical alternative for withPool() with a shorter name.
     * Within the supplied code block the ExecutorService is available as the only parameter, objects have been
     * enhanced with the eachParallel(), collectParallel() and other methods from the GParsExecutorsPoolUtil
     * category class as well as closures can be turned into asynchronous ones by calling the async() method on them.
     * E.g. closure,async returns a new closure, which, when run will schedule the original closure
     * for processing in the pool.
     * Calling images.eachParallel{processImage(it}} will call the potentially long-lasting processImage()
     * operation on each image in the images collection in parallel.
     * 
     * def result = new ConcurrentSkipListSet()
     * GParsExecutorsPool.withPool {ExecutorService service ->
     *     [1, 2, 3, 4, 5].eachParallel{Number number -> result.add(number * 10)}*     assertEquals(new HashSet([10, 20, 30, 40, 50]), result)
     *}* 
* @param cl The block of code to invoke with the DSL enabled */ public static withPool(Closure cl) { return withPool(defaultPoolSize, cl) } /** * Creates a new instance of ExecutorService, binds it to the current thread, enables the ExecutorService DSL * and runs the supplied closure. * It is an identical alternative for withPool() with a shorter name. * Within the supplied code block the ExecutorService is available as the only parameter, objects have been * enhanced with the eachParallel(), collectParallel() and other methods from the GParsExecutorsPoolUtil * category class as well as closures can be turned into asynchronous ones by calling the async() method on them. * E.g. closure,async returns a new closure, which, when run will schedule the original closure * for processing in the pool. * Calling images.eachParallel{processImage(it}} will call the potentially long-lasting processImage() * operation on each image in the images collection in parallel. *
     * def result = new ConcurrentSkipListSet()
     * GParsExecutorsPool.withPool(5) {ExecutorService service ->
     *     [1, 2, 3, 4, 5].eachParallel{Number number -> result.add(number * 10)}*     assertEquals(new HashSet([10, 20, 30, 40, 50]), result)
     *}* 
* @param numberOfThreads Number of threads in the newly created thread pool * @param cl The block of code to invoke with the DSL enabled */ public static withPool(int numberOfThreads, Closure cl) { return withPool(numberOfThreads, createDefaultThreadFactory(), cl) } /** * Creates a new instance of ExecutorService, binds it to the current thread, enables the ExecutorService DSL * and runs the supplied closure. * It is an identical alternative for withPool() with a shorter name. * Within the supplied code block the ExecutorService is available as the only parameter, objects have been * enhanced with the eachParallel(), collectParallel() and other methods from the GParsExecutorsPoolUtil * category class as well as closures can be turned into asynchronous ones by calling the async() method on them. * E.g. closure,async returns a new closure, which, when run will schedule the original closure * for processing in the pool. * Calling images.eachParallel{processImage(it}} will call the potentially long-lasting processImage() * operation on each image in the images collection in parallel. *
     * def result = new ConcurrentSkipListSet()
     * GParsExecutorsPool.withPool(5) {ExecutorService service ->
     *     [1, 2, 3, 4, 5].eachParallel{Number number -> result.add(number * 10)}*     assertEquals(new HashSet([10, 20, 30, 40, 50]), result)
     *}* 
* @param numberOfThreads Number of threads in the newly created thread pool * @param threadFactory Factory for threads in the pool * @param cl The block of code to invoke with the DSL enabled */ public static withPool(int numberOfThreads, ThreadFactory threadFactory, Closure cl) { final ExecutorService pool = createPool(numberOfThreads, threadFactory) try { return withExistingPool(pool, cl) } finally { pool.shutdown() pool.awaitTermination(Long.MAX_VALUE, TimeUnit.MILLISECONDS) } } /** * Creates a new instance of ExecutorService, binds it to the current thread, enables the ExecutorService DSL * and runs the supplied closure. * Within the supplied code block the ExecutorService is available as the only parameter, objects have been * enhanced with the eachParallel(), collectParallel() and other methods from the GParsExecutorsPoolUtil * category class as well as closures can be turned into asynchronous ones by calling the async() method on them. * E.g. closure,async returns a new closure, which, when run will schedule the original closure * for processing in the pool. * Calling images.eachParallel{processImage(it}} will call the potentially long-lasting processImage() * operation on each image in the images collection in parallel. *
     * def result = new ConcurrentSkipListSet()
     * GParsExecutorsPool.withPool(5) {ExecutorService service ->
     *     [1, 2, 3, 4, 5].eachParallel{Number number -> result.add(number * 10)}*     assertEquals(new HashSet([10, 20, 30, 40, 50]), result)
     *}* 
* @param pool The ExecutorService to use, the service will not be shutdown after this method returns */ public static withExistingPool(ExecutorService pool, Closure cl) { currentPoolStack << pool def result = null try { use(GParsExecutorsPoolUtil) { result = cl(pool) } } finally { currentPoolStack.pop() } return result } /** * Starts multiple closures in separate threads, collecting their return values * If an exception is thrown from the closure when called on any of the collection's elements, * it will be re-thrown in the calling thread when it calls the Future.get() method. * @return The result values of all closures * @throws AsyncException If any of the collection's elements causes the closure to throw an exception. The original exceptions will be stored in the AsyncException's concurrentExceptions field. */ public static List executeAsyncAndWait(Closure... closures) { return GParsExecutorsPoolUtil.processResult(executeAsync(closures)) } /** * Starts multiple closures in separate threads, collecting their return values * If an exception is thrown from the closure when called on any of the collection's elements, * it will be re-thrown in the calling thread when it calls the Future.get() method. * @return The result values of all closures * @throws AsyncException If any of the collection's elements causes the closure to throw an exception. The original exceptions will be stored in the AsyncException's concurrentExceptions field. */ public static List executeAsyncAndWait(List closures) { return executeAsyncAndWait(* closures) } /** * Starts multiple closures in separate threads, collecting Futures for their return values * Reuses the pool defined by the surrounding withPool() call. * If an exception is thrown from the closure when called on any of the collection's elements, * it will be re-thrown in the calling thread when it calls the Future.get() method. * @return Futures for the result values or exceptions of all closures */ public static List> executeAsync(Closure... closures) { ExecutorService executorService = retrieveCurrentPool() if (executorService == null) throw new IllegalStateException("No active thread pool available to execute closures asynchronously. Consider wrapping the function call with GParsExecutorsPool.withPool().") List> result = closures.collect {cl -> executorService.submit({ cl.call() } as Callable) } result } /** * Starts multiple closures in separate threads, collecting Futures for their return values * If an exception is thrown from the closure when called on any of the collection's elements, * it will be re-thrown in the calling thread when it calls the Future.get() method. * @return Futures for the result values or exceptions of all closures */ public static List> executeAsync(List closures) { return executeAsync(* closures) } /** * Runs the supplied closures asynchronously and in parallel, returning the first result obtained and cancelling the other (slower) calculations. * Typically used to run several different calculations in parallel, all of which are supposed to give the same result, * but may last different amount of time each. If the system has enough threads available, the calculations can be test-run * in parallel and the fastest result is then used, while the other results are cancelled or discarded. * @param alternatives All the functions to invoke in parallel * @return The fastest result obtained */ public static def speculate(List alternatives) { speculate(* alternatives) } /** * Runs the supplied closures asynchronously and in parallel, returning the first result obtained and cancelling the other (slower) calculations. * Typically used to run several different calculations in parallel, all of which are supposed to give the same result, * but may last different amount of time each. If the system has enough threads available, the calculations can be test-run * in parallel and the fastest result is then used, while the other results are cancelled or discarded. * @param alternatives All the functions to invoke in parallel * @return The fastest result obtained */ public static def speculate(Closure... alternatives) { def result = new DataflowVariable() def futures = new DataflowVariable() final AtomicInteger failureCounter = new AtomicInteger(0) futures << GParsExecutorsPool.executeAsync(alternatives.collect { original -> {-> //noinspection GroovyEmptyCatchBlock try { def localResult = original() futures.val*.cancel(true) result << localResult } catch (Exception e) { int counter = failureCounter.incrementAndGet() if (counter == alternatives.size()) { result << new IllegalStateException('All speculations failed', e) } } } }) def r = result.val if (r instanceof Exception) throw r return r } }