Future、 CompletableFuture、ThreadPoolTaskExecutor简单实践

一 Future（jdk5引入）

简介： Future接口是Java多线程Future模式的实现，可以来进行异步计算。

可以使用isDone方法检查计算是否完成，或者使用get阻塞住调用线程，直到计算完成返回结果，也可以使用cancel方法停止任务的执行。

简单测试 -  主题 ： Future模式可以理解成：我有一个任务，提交给了Future，Future替我完成这个任务，期间程序可以处理其他任务。

                              一段时间之后，主线程可以从Future那儿取出结果。

public class ThreadPoolTest {

    private static  Logger logger= LoggerFactory.getLogger(ThreadPoolTest.class);

    public static void main(String[] args) {
        ExecutorService threadPool = Executors.newSingleThreadExecutor();//创建单一线程
        Future<String> future = threadPool.submit(new Callable<String>() {
            @Override
            public String call() throws Exception {
                Thread.sleep(2000);
                logger.info("out2-----");
                return "Hello world";
            }
        });

        try {
            try {
                BigDecimal bigDecimal=new BigDecimal(0);
                logger.info("out1:"+bigDecimal.toString());
                logger.info(future.get(3000,TimeUnit.MILLISECONDS));//Hello world
            } catch (TimeoutException e) {
                logger.error("timeout-exception",e);
            }
        } catch (InterruptedException e) {
            logger.error("interrupted-exception",e);
        } catch (ExecutionException e) {
            logger.error("execution-exception",e);
        }finally {
            threadPool.shutdown();
        }
}

console输出：

11:17:17.558 [main] INFO tk.mybatis.springboot.util.thread.ThreadPoolTest - out1:0
Disconnected from the target VM, address: '127.0.0.1:54966', transport: 'socket'
11:17:19.556 [pool-1-thread-1] INFO tk.mybatis.springboot.util.thread.ThreadPoolTest - out2-----
11:17:19.557 [main] INFO tk.mybatis.springboot.util.thread.ThreadPoolTest - Hello world

超时会报错：

SLF4J: Actual binding is of type [ch.qos.logback.classic.util.ContextSelectorStaticBinder]
10:44:07.413 [main] ERROR tk.mybatis.springboot.util.thread.ThreadPoolTest - timeout-exception
java.util.concurrent.TimeoutException: null
    at java.util.concurrent.FutureTask.get(FutureTask.java:205)
    at tk.mybatis.springboot.util.thread.ThreadPoolTest.main(ThreadPoolTest.java:30)

Process finished with exit code 0

总结：1  比起future.get()，推荐使用get (long timeout, TimeUnit unit) 方法，

设置了超时时间可以防止程序无限制的等待future的结果，可以进行异常处理逻辑。

2  虽然Future以及相关使用方法提供了异步执行任务的能力，但是对于结果的获取却是很不方便，

只能通过阻塞或者轮询的方式得到任务的结果。

阻塞的方式显然和我们的异步编程的初衷相违背，轮询的方式又会耗费无谓的CPU资源，而且也不能及时地得到计算结果，

为什么不能用观察者设计模式当计算结果完成及时通知监听者呢？

Future的接口很简单，只有五个方法。jdk8中源码如下：

package java.util.concurrent;

/**
 * A {@code Future} represents the result of an asynchronous
 * computation.  Methods are provided to check if the computation is
 * complete, to wait for its completion, and to retrieve the result of
 * the computation.  The result can only be retrieved using method
 * {@code get} when the computation has completed, blocking if
 * necessary until it is ready.  Cancellation is performed by the
 * {@code cancel} method.  Additional methods are provided to
 * determine if the task completed normally or was cancelled. Once a
 * computation has completed, the computation cannot be cancelled.
 * If you would like to use a {@code Future} for the sake
 * of cancellability but not provide a usable result, you can
 * declare types of the form {@code Future<?>} and
 * return {@code null} as a result of the underlying task.
 *
 * <p>
 * <b>Sample Usage</b> (Note that the following classes are all
 * made-up.)
 * <pre> {@code
 * interface ArchiveSearcher { String search(String target); }
 * class App {
 *   ExecutorService executor = ...
 *   ArchiveSearcher searcher = ...
 *   void showSearch(final String target)
 *       throws InterruptedException {
 *     Future<String> future
 *       = executor.submit(new Callable<String>() {
 *         public String call() {
 *             return searcher.search(target);
 *         }});
 *     displayOtherThings(); // do other things while searching
 *     try {
 *       displayText(future.get()); // use future
 *     } catch (ExecutionException ex) { cleanup(); return; }
 *   }
 * }}</pre>
 *
 * The {@link FutureTask} class is an implementation of {@code Future} that
 * implements {@code Runnable}, and so may be executed by an {@code Executor}.
 * For example, the above construction with {@code submit} could be replaced by:
 *  <pre> {@code
 * FutureTask<String> future =
 *   new FutureTask<String>(new Callable<String>() {
 *     public String call() {
 *       return searcher.search(target);
 *   }});
 * executor.execute(future);}</pre>
 *
 * <p>Memory consistency effects: Actions taken by the asynchronous computation
 * <a href="package-summary.html#MemoryVisibility"> <i>happen-before</i></a>
 * actions following the corresponding {@code Future.get()} in another thread.
 *
 * @see FutureTask
 * @see Executor
 * @since 1.5
 * @author Doug Lea
 * @param <V> The result type returned by this Future's {@code get} method
 */
public interface Future<V> {

    /**
     * Attempts to cancel execution of this task.  This attempt will
     * fail if the task has already completed, has already been cancelled,
     * or could not be cancelled for some other reason. If successful,
     * and this task has not started when {@code cancel} is called,
     * this task should never run.  If the task has already started,
     * then the {@code mayInterruptIfRunning} parameter determines
     * whether the thread executing this task should be interrupted in
     * an attempt to stop the task.
     *
     * <p>After this method returns, subsequent calls to {@link #isDone} will
     * always return {@code true}.  Subsequent calls to {@link #isCancelled}
     * will always return {@code true} if this method returned {@code true}.
     *
     * @param mayInterruptIfRunning {@code true} if the thread executing this
     * task should be interrupted; otherwise, in-progress tasks are allowed
     * to complete
     * @return {@code false} if the task could not be cancelled,
     * typically because it has already completed normally;
     * {@code true} otherwise
     */
    boolean cancel(boolean mayInterruptIfRunning);

    /**
     * Returns {@code true} if this task was cancelled before it completed
     * normally.
     *
     * @return {@code true} if this task was cancelled before it completed
     */
    boolean isCancelled();

    /**
     * Returns {@code true} if this task completed.
     *
     * Completion may be due to normal termination, an exception, or
     * cancellation -- in all of these cases, this method will return
     * {@code true}.
     *
     * @return {@code true} if this task completed
     */
    boolean isDone();

    /**
     * Waits if necessary for the computation to complete, and then
     * retrieves its result.
     *
     * @return the computed result
     * @throws CancellationException if the computation was cancelled
     * @throws ExecutionException if the computation threw an
     * exception
     * @throws InterruptedException if the current thread was interrupted
     * while waiting
     */
    V get() throws InterruptedException, ExecutionException;

    /**
     * Waits if necessary for at most the given time for the computation
     * to complete, and then retrieves its result, if available.
     *
     * @param timeout the maximum time to wait
     * @param unit the time unit of the timeout argument
     * @return the computed result
     * @throws CancellationException if the computation was cancelled
     * @throws ExecutionException if the computation threw an
     * exception
     * @throws InterruptedException if the current thread was interrupted
     * while waiting
     * @throws TimeoutException if the wait timed out
     */
    V get(long timeout, TimeUnit unit)
        throws InterruptedException, ExecutionException, TimeoutException;
}

看懂英文还是最直接的。

二. CompletableFuture介绍

2.1 Future模式的缺点

• Future虽然可以实现获取异步执行结果的需求，但是它没有提供通知的机制，我们无法得知Future什么时候完成。

• 要么使用阻塞，在future.get()的地方等待future返回的结果，这时又变成同步操作。要么使用isDone()轮询地判断Future是否完成，这样会耗费CPU的资源。

2.2 CompletableFuture介绍

 public class CompletableFuture<T> implements Future<T>, CompletionStage<T> {
 }

Netty、Guava分别扩展了Java 的 Future 接口，方便异步编程。

在Java 8中, 新增加了一个包含50个方法左右的类: CompletableFuture，提供了非常强大的Future的扩展功能，

可以帮助我们简化异步编程的复杂性，提供了函数式编程的能力，可以通过回调的方式处理计算结果，

并且提供了转换和组合CompletableFuture的方法。

1 )主动完成计算
   CompletableFuture类实现了CompletionStage和Future接口，所以你还是可以像以前一样通过阻塞或者轮询的方式获得结果，尽管这种方式不推荐使用。

public T     get()
public T     get(long timeout, TimeUnit unit)
public T     getNow(T valueIfAbsent) //如果结果已经计算完则返回结果或者抛出异常，否则返回给定的valueIfAbsent值
public T     join()

 join返回计算的结果或者抛出一个unchecked异常(CompletionException)，它和get对抛出的异常的处理有些细微的区别

@Test
    public void test(){
        CompletableFuture<Integer> future = CompletableFuture.supplyAsync(()->{
           int i=1/0;
           return 100;
        });
        logger.info(String.format("join=%d", future.join()));
        try {
            logger.info(String.format("get=%s", future.get()));
        } catch (InterruptedException e) {
           logger.error("error=%s",e);
        } catch (ExecutionException e) {
            logger.error("error=%s",e);
        }
    }

尽管Future可以代表在另外的线程中执行的一段异步代码，但是你还是可以在本身线程中执行：

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.concurrent.CompletableFuture;
import java.util.concurrent.ExecutionException;

public class BasicMain {
    private static Logger logger = LoggerFactory.getLogger(BasicMain.class);
   
    public static CompletableFuture<Integer> compute() {
        final CompletableFuture<Integer> future = new CompletableFuture<>();
        return future;
    }

    public static void main(String[] args) throws Exception {
        final CompletableFuture<Integer> f = compute();

        class Client extends Thread {
            CompletableFuture<Integer> f;
            Client(String threadName, CompletableFuture<Integer> f) {
                super(threadName);
                this.f = f;
            }

            @Override
            public void run() {
                try {
                   logger.info(this.getName() + ": " + f.get());
                } catch (InterruptedException e) {
                    logger.error(e.getMessage());
                } catch (ExecutionException e) {
                    logger.error(e.getMessage());
                }
            }
        }
        new Client("Client1", f).start();
        new Client("Client2", f).start();
        logger.info("waiting....");
        f.complete(100);
        f.obtrudeValue(200);   
      //logger Client1: 200  Client2: 200 或者 Client1: 100 Client1: 100 或者 Client1: 200 Client2: 100 ，都有可能

      //        f.completeExceptionally(new Exception()); 

    }
}

说明：

可以看到我们并没有把f.complete(100) 放在另外的线程中去执行，但是在大部分情况下我们可能会用一个线程池去执行这些异步任 务。CompletableFuture.complete()、CompletableFuture.completeExceptionally只能被调用一次。

但是我们有两个后门方法可以重设这个值:obtrudeValue、obtrudeException，但是使用的时候要小心，

因为complete已经触发了客户端，有可能导致客户端会得到不期望的结果

2）创建CompletableFuture对象。

CompletableFuture的静态工厂方法

方法名	描述
runAsync(Runnable runnable)	使用ForkJoinPool.commonPool()作为它的线程池执行异步代码。
runAsync(Runnable runnable, Executor executor)	使用指定的thread pool执行异步代码。
supplyAsync(Supplier<U> supplier)	使用ForkJoinPool.commonPool()作为它的线程池执行异步代码，异步操作有返回值
supplyAsync(Supplier<U> supplier, Executor executor)	使用指定的thread pool执行异步代码，异步操作有返回值

eg： runAsync 和 supplyAsync 方法比较

区别是：

runAsync返回的CompletableFuture是没有返回值的

    原因：

      Runnable函数式接口类型为参数，所以CompletableFuture的计算结果为空

共同点：

因为方法的参数类型都是函数式接口，所以可以使用lambda表达式实现异步任务，比如：

 CompletableFuture<String> future = CompletableFuture.supplyAsync(() -> {
            return "Hello";
        });

        try {
           logger.info("out = "+ future.get());//out = Hello
        } catch (InterruptedException e) {
            e.printStackTrace();
        } catch (ExecutionException e) {
            e.printStackTrace();
        }

      CompletableFuture<Void> future = CompletableFuture.runAsync(() -> {
            System.out.println("Hello");
        });

        try {
           logger.info("out = "+ future.get());//out = null
        } catch (InterruptedException e) {
            e.printStackTrace();
        } catch (ExecutionException e) {
            e.printStackTrace();
        }

简单试验 - 主题：在两个线程里并行执行任务A和任务B，只要有一个任务完成了，就执行任务C

package tk.mybatis.springboot.util.thread;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import java.time.LocalTime;
import java.util.Random;
import java.util.concurrent.*;

/**
 * 在两个线程里并行执行任务A和任务B，只要有一个任务完成了，就执行任务C
 *
 * 两种方法useFuture和useCompletableFuture相比：
 *
 * 首先，useCompletableFuture 比 useFuture 的代码简单。
 *  在useFuture 里，既要自己照顾线程池的创建和销毁，还要负责对任务A和任务B的监控。
 *  而useCompletableFuture，只需要用CompletableFuture的runAfterEither就完成了任务A、任务B和任务C之间的依赖关系的定义。
 *
 */
public class CompletableFutureTest {

    private static Logger logger = LoggerFactory.getLogger(CompletableFutureTest.class);
    private static Random random = new Random();

    /**
     * useFuture test
     * @throws InterruptedException
     * @throws ExecutionException
     */
    private static void useFuture() throws InterruptedException, ExecutionException {
        logger.info("useFuture");
        ExecutorService exector = Executors.newFixedThreadPool(3);
        Future<Void> futureA = exector.submit(() -> work("A1"));
        Future<Void> futureB = exector.submit(() -> work("B1"));
        while (true) {
            try {
                futureA.get(1, TimeUnit.SECONDS);
                break;
            } catch (TimeoutException e) {
            }
            try {
                futureB.get(1, TimeUnit.SECONDS);
                break;
            } catch (TimeoutException e) {
            }
        }
        exector.submit(() -> work("C1")).get();
        exector.shutdown();
    }

    private static void useCompletableFuture() throws InterruptedException, ExecutionException {
        logger.info("useCompletableFuture");
        CompletableFuture<Void> futureA = CompletableFuture.runAsync(() -> work("A2"));
        CompletableFuture<Void> futureB = CompletableFuture.runAsync(() -> work("B2"));
        logger.info("get="+futureA.runAfterEither(futureB, () -> work("C2")).get());

//     或者
//        CompletableFuture.runAsync(() -> work("A2"))
//                .runAfterEither(CompletableFuture.runAsync(() -> work("B2"))
//                        , () -> work("C2"))
//                .get();
    }

    /**
     * logger 输出
     * @param name
     * @return
     */
    public static Void work(String name) {
        logger.info(name + " starts at " + LocalTime.now());
        try {
            TimeUnit.SECONDS.sleep(random.nextInt(10));
        } catch (InterruptedException e) {
        }
        logger.info(name + " ends at " + LocalTime.now());
        return null;
    }

    /**
     * 从useFuture的输出可以看出，
     *  任务C1的开始并不是紧随着任务A1的完成，差了0.001秒，
     *  原因是在方法1里，是对任务A1和任务B1都用get(1,TimeUnit.SECONDS)来询问他们的状态，
     *  当其中一个任务先完成时，主线程可能正阻塞在另一个未完成任务的get上
     *
     * 而从useCompletableFuture完全不存在这样的问题，
     *      任务C2的开始于任务A1的结束之间没有任何的时间差
     *
     * @param args
     * @throws InterruptedException
     * @throws ExecutionException
     */
    public static void main(String[] args) throws InterruptedException, ExecutionException {
//        useFuture();
        //logger 输出
        // B1 starts at 10:28:09.562
//         A1 starts at 10:28:09.562
//         B1 ends at 10:28:17.566
//         C1 starts at 10:28:17.567
//         A1 ends at 10:28:18.563
//         C1 ends at 10:28:20.570

//        TimeUnit.SECONDS.sleep(10);
        useCompletableFuture();
        //logger 输出
//        A2 starts at 10:43:46.867
//        B2 starts at 10:43:46.867
//        B2 ends at 10:43:51.871
//        A2 ends at 10:43:51.871
//        C2 starts at 10:43:51.871
//        C2 ends at 10:44:00.874
        TimeUnit.SECONDS.sleep(10);//避免打印出的start - end 不全

    }

}

 3）计算结果完成时的处理

 当CompletableFuture的计算结果完成，或者抛出异常的时候，我们可以执行特定的Action。主要是下面的方法：

   public CompletableFuture<T> whenComplete(BiConsumer<? super T, ? super Throwable> action) {
        return uniWhenCompleteStage(null, action);
    }

    public CompletableFuture<T> whenCompleteAsync(BiConsumer<? super T, ? super Throwable> action) {
        return uniWhenCompleteStage(asyncPool, action);
    }

    public CompletableFuture<T> whenCompleteAsync(BiConsumer<? super T, ? super Throwable> action, Executor executor) {
        return uniWhenCompleteStage(screenExecutor(executor), action);
    }
   public CompletableFuture<T> exceptionally(Function<Throwable, ? extends T> fn) {
        return uniExceptionallyStage(fn);
    }

可以看到Action的类型是BiConsumer<? super T,? super Throwable>，它可以处理正常的计算结果，或者异常情况。
       方法不以Async结尾，意味着Action使用相同的线程执行，

而Async可能会使用其它的线程去执行(如果使用相同的线程池，也可能会被同一个线程选中执行)。

注意这几个方法都会返回CompletableFuture，当Action执行完毕后它的结果返回原始的CompletableFuture的计算结果或者返回异常。

package tk.mybatis.springboot.util.thread;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import java.util.HashMap;
import java.util.Map;
import java.util.Random;
import java.util.concurrent.CompletableFuture;
import java.util.concurrent.Future;

public class MainTest {
    private static Logger logger = LoggerFactory.getLogger(MainTest.class);

    private static Random random = new Random();
    private static long t = System.currentTimeMillis();

    static Map<String,String> getMoreData() {
        logger.info("start");
        try {
            Thread.sleep(3000);
        } catch (InterruptedException e) {
            throw new RuntimeException(e);
        }
        logger.info("end ");
        Map map=new HashMap();
        map.put(random.nextInt(1000),random.nextInt(1000));
        return map;
    }

    public static void main(String[] args) throws Exception {
        CompletableFuture<Map<String,String>> future = CompletableFuture.supplyAsync(MainTest::getMoreData);
        Future<Map<String,String>> f = future.whenComplete((v, e) -> {
            logger.info("v="+v);
            logger.info("e="+e);
        });
        logger.info("get="+f.get());
    }
}

     exceptionally方法返回一个新的CompletableFuture，

当原始的CompletableFuture抛出异常的时候，就会触发这个CompletableFuture的计算，调用function计算值，

否则如果原始的CompletableFuture正常计算完后，这个新的CompletableFuture也计算完成，它的值和原始的CompletableFuture的计算的值相同（？？）。

也就是这个exceptionally方法用来处理异常的情况。

        CompletableFuture<Map<String,String>> futureException = future.exceptionally(new Function<Throwable, Map<String, String>>() {
            @Override
            public Map<String, String> apply(Throwable throwable) {
                logger.error("error="+throwable.getMessage());
                Map map=new HashMap();
                map.put(random.nextInt(1000),random.nextInt(1000));
                return map;
            }
        });

        logger.info("get.exception="+futureException.get());

下面一组方法虽然也返回CompletableFuture对象，但是对象的值和原来的CompletableFuture计算的值不同。

当原先的CompletableFuture的值计算完成或者抛出异常的时候，会触发这个CompletableFuture对象的计算，结果由BiFunction参数计算而得。

因此这组方法兼有whenComplete和转换的两个功能。

public <U> CompletableFuture<U>     handle(BiFunction<? super T,Throwable,? extends U> fn)
public <U> CompletableFuture<U>     handleAsync(BiFunction<? super T,Throwable,? extends U> fn)
public <U> CompletableFuture<U>     handleAsync(BiFunction<? super T,Throwable,? extends U> fn, Executor executor)

同样，不以Async结尾的方法由原来的线程计算，以Async结尾的方法由默认的线程池ForkJoinPool.commonPool()或者指定的线程池executor运行。

4）转换

    CompletableFuture可以作为monad(单子)和functor。由于回调风格的实现，我们不必因为等待一个计算完成而阻塞着调用线程，

而是告诉CompletableFuture当计算完成的时候请执行某个function。而且我们还可以将这些操作串联起来，或者将CompletableFuture组合起来。

三 ThreadPoolTaskExecutor

包路径：org.springframework.scheduling.concurrent