Home > Article > Java > Introduction to the usage of Java8's CompletableFuture (with examples)

Introduction to the usage of Java8's CompletableFuture (with examples)

不言forward: 2019-04-04 10:04:495662browse

This article brings you an introduction to the usage of Java8's CompletableFuture (with examples). It has certain reference value. Friends in need can refer to it. I hope it will be helpful to you.

Introduced as a Java 8 Concurrency API improvement, this article is an introduction to the functionality and use cases of the CompletableFuture class. At the same time, there are some improvements to CompletableFuture in Java 9, which will be explained later.

Future calculation

Future asynchronous calculation is difficult to operate, and generally we want to treat any calculation logic as a series of steps. But in the case of asynchronous computation, methods represented as callbacks are often scattered throughout the code or deeply nested within each other. But things can get more complicated when we need to handle errors that may occur in one of the steps.

The Future interface is new in Java 5 as asynchronous calculation, but it does not have any method to combine calculations or handle possible errors.

In Java 8, the CompletableFuture class was introduced. Along with the Future interface, it also implements the CompletionStage interface. This interface defines an asynchronous calculation contract that can be combined with other Futures.

CompletableFuture is both a composition and a framework, with about 50 different compositions, combinations, executing asynchronous calculation steps and handling errors.

Such a huge API may be overwhelming. Some important ones will be highlighted below.

Using CompletableFuture as Future implementation

First, the CompletableFuture class implements the Future interface, so you can use it as a Future implementation, but additional completion implementation logic is required.

For example, you can use the parameterless constructor to create an instance of this class, and then use the complete method to complete it. Consumers can use the get method to block the current thread until get() results.

In the example below, we have a method that creates a CompletableFuture instance, then computes and returns the Future immediately in another thread.

After the calculation is completed, the method completes the Future by providing the result to the complete method:

public Future<String> calculateAsync() throws InterruptedException {
    CompletableFuture<String> completableFuture 
      = new CompletableFuture<>();
 
    Executors.newCachedThreadPool().submit(() -> {
        Thread.sleep(500);
        completableFuture.complete("Hello");
        return null;
    });
 
    return completableFuture;
}

To separate the calculation, we use the Executor API, which creates and The completion methods of CompletableFuture can be used with any concurrent package (including the original thread).

Please note that the calculateAsync method returns a Future instance.

We just call the method, receive the Future instance and call its get method when we are ready to block on the result.

Also note that the get method throws some checked exceptions, namely ExecutionException (which encapsulates exceptions that occur during computation) and InterruptedException (Exception indicating that the thread executing the method is interrupted):

Future<String> completableFuture = calculateAsync();
 
// ... 
 
String result = completableFuture.get();
assertEquals("Hello", result);

If you already know the result of the calculation, you can also return the result synchronously.

Future<String> completableFuture = 
  CompletableFuture.completedFuture("Hello");
 
// ...
 
String result = completableFuture.get();
assertEquals("Hello", result);

As in some scenarios, you may want to cancel the execution of Future tasks.

Suppose we find no results and decide to cancel the asynchronous execution of the task entirely. This can be done through the Future's cancellation method. This method mayInterruptIfRunning, but in the case of CompletableFuture, it has no effect because interrupts are not used to control the processing of CompletableFuture.

This is a modified version of the asynchronous method:

public Future<String> calculateAsyncWithCancellation() throws InterruptedException {
    CompletableFuture<String> completableFuture = new CompletableFuture<>();
 
    Executors.newCachedThreadPool().submit(() -> {
        Thread.sleep(500);
        completableFuture.cancel(false);
        return null;
    });
 
    return completableFuture;
}

When we use the Future.get() method to block the result, cancel() means canceling the execution, it will throw CancellationException:

Future<String> future = calculateAsyncWithCancellation();
future.get(); // CancellationException

API introduction

static method description

The above code is very simple. Here are several static methods that use tasks to instantiate a CompletableFuture instance.

CompletableFuture.runAsync(Runnable runnable);
CompletableFuture.runAsync(Runnable runnable, Executor executor);

CompletableFuture.supplyAsync(Supplier<U> supplier);
CompletableFuture.supplyAsync(Supplier<U> supplier, Executor executor)

The runAsync method receives an instance of Runnable, but it does not return a value

The supplyAsync method is a JDK8 functional interface with no parameters and will return a result

These two methods are upgrades of executor, which means that the task is executed in the specified thread pool. If not specified, the task is usually executed in the ForkJoinPool.commonPool() thread pool.

supplyAsync() uses

static methods runAsync and supplyAsync allowing us to create CompletableFuture instances from Runnable and Supplier function types accordingly .

The Runnable interface is an old interface used in threads, which does not allow return values.

The Supplier interface is a general functional interface that has no parameters and returns a single method of a parameterized type.

This allows an instance of Supplier to be provided as a lambda expression that performs a calculation and returns the result:

CompletableFuture<String> future
  = CompletableFuture.supplyAsync(() -> "Hello");
 
// ...
 
assertEquals("Hello", future.get());

thenRun() is used

in two tasks taskA , in task B, if you neither need the value of task A nor want to reference it in task B, then you can pass the Runnable lambda to the thenRun() method. In the following example, after calling the future.get() method, we simply print a single line in the console:

TEMPLATE

CompletableFuture.runAsync(() -> {}).thenRun(() -> {}); 
CompletableFuture.supplyAsync(() -> "resultA").thenRun(() -> {});

第一行用的是 thenRun(Runnable runnable)，任务 A 执行完执行 B，并且 B 不需要 A 的结果。
第二行用的是 thenRun(Runnable runnable)，任务 A 执行完执行 B，会返回resultA，但是 B 不需要 A 的结果。

实战

CompletableFuture<String> completableFuture 
  = CompletableFuture.supplyAsync(() -> "Hello");
 
CompletableFuture<Void> future = completableFuture
  .thenRun(() -> System.out.println("Computation finished."));
 
future.get();

thenAccept()使用

在两个任务任务A，任务B中，如果你不需要在Future中有返回值，则可以用 thenAccept方法接收将计算结果传递给它。最后的future.get（）调用返回Void类型的实例。

模板

CompletableFuture.runAsync(() -> {}).thenAccept(resultA -> {}); 

CompletableFuture.supplyAsync(() -> "resultA").thenAccept(resultA -> {});

第一行中，runAsync不会有返回值，第二个方法thenAccept，接收到的resultA值为null，同时任务B也不会有返回结果

第二行中，supplyAsync有返回值，同时任务B不会有返回结果。

实战

CompletableFuture<String> completableFuture
  = CompletableFuture.supplyAsync(() -> "Hello");
 
CompletableFuture<Void> future = completableFuture
  .thenAccept(s -> System.out.println("Computation returned: " + s));
 
future.get();

thenApply()使用

在两个任务任务A，任务B中，任务B想要任务A计算的结果，可以用thenApply方法来接受一个函数实例，用它来处理结果，并返回一个Future函数的返回值：

模板

CompletableFuture.runAsync(() -> {}).thenApply(resultA -> "resultB");
CompletableFuture.supplyAsync(() -> "resultA").thenApply(resultA -> resultA + " resultB");

第二行用的是 thenApply(Function fn)，任务 A 执行完执行 B，B 需要 A 的结果，同时任务 B 有返回值。

实战

CompletableFuture<String> completableFuture
  = CompletableFuture.supplyAsync(() -> "Hello");
 
CompletableFuture<String> future = completableFuture
  .thenApply(s -> s + " World");
 
assertEquals("Hello World", future.get());

当然，多个任务的情况下，如果任务 B 后面还有任务 C，往下继续调用 .thenXxx() 即可。

thenCompose()使用

接下来会有一个很有趣的设计模式；

CompletableFuture API 的最佳场景是能够在一系列计算步骤中组合CompletableFuture实例。

这种组合结果本身就是CompletableFuture，允许进一步再续组合。这种方法在函数式语言中无处不在，通常被称为monadic设计模式。

简单说，Monad就是一种设计模式，表示将一个运算过程，通过函数拆解成互相连接的多个步骤。你只要提供下一步运算所需的函数，整个运算就会自动进行下去。

在下面的示例中，我们使用thenCompose方法按顺序组合两个Futures。

请注意，此方法采用返回CompletableFuture实例的函数。该函数的参数是先前计算步骤的结果。这允许我们在下一个CompletableFuture的lambda中使用这个值：

CompletableFuture<String> completableFuture 
  = CompletableFuture.supplyAsync(() -> "Hello")
    .thenCompose(s -> CompletableFuture.supplyAsync(() -> s + " World"));
 
assertEquals("Hello World", completableFuture.get());

该thenCompose方法连同thenApply一样实现了结果的合并计算。但是他们的内部形式是不一样的，它们与Java 8中可用的Stream和Optional类的map和flatMap方法是有着类似的设计思路在里面的。

两个方法都接收一个CompletableFuture并将其应用于计算结果，但thenCompose（flatMap）方法接收一个函数，该函数返回相同类型的另一个CompletableFuture对象。此功能结构允许将这些类的实例继续进行组合计算。

thenCombine()

取两个任务的结果

如果要执行两个独立的任务，并对其结果执行某些操作，可以用Future的thenCombine方法：

模板

CompletableFuture<String> cfA = CompletableFuture.supplyAsync(() -> "resultA");
CompletableFuture<String> cfB = CompletableFuture.supplyAsync(() -> "resultB");

cfA.thenAcceptBoth(cfB, (resultA, resultB) -> {});

cfA.thenCombine(cfB, (resultA, resultB) -> "result A + B");

实战

CompletableFuture<String> completableFuture 
  = CompletableFuture.supplyAsync(() -> "Hello")
    .thenCombine(CompletableFuture.supplyAsync(
      () -> " World"), (s1, s2) -> s1 + s2));
 
assertEquals("Hello World", completableFuture.get());

更简单的情况是，当你想要使用两个Future结果时，但不需要将任何结果值进行返回时，可以用thenAcceptBoth，它表示后续的处理不需要返回值，而 thenCombine 表示需要返回值：

CompletableFuture future = CompletableFuture.supplyAsync(() -> "Hello")
  .thenAcceptBoth(CompletableFuture.supplyAsync(() -> " World"),
    (s1, s2) -> System.out.println(s1 + s2));

thenApply()和thenCompose()之间的区别

在前面的部分中，我们展示了关于thenApply()和thenCompose()的示例。这两个API都是使用的CompletableFuture调用，但这两个API的使用是不同的。

thenApply()

此方法用于处理先前调用的结果。但是，要记住的一个关键点是返回类型是转换泛型中的类型，是同一个CompletableFuture。

因此，当我们想要转换CompletableFuture 调用的结果时，效果是这样的：

CompletableFuture<Integer> finalResult = compute().thenApply(s-> s + 1);

thenCompose()

该thenCompose()方法类似于thenApply()在都返回一个新的计算结果。但是，thenCompose()使用前一个Future作为参数。它会直接使结果变新的Future，而不是我们在thenApply()中到的嵌套Future，而是用来连接两个CompletableFuture，是生成一个新的CompletableFuture：

CompletableFuture<Integer> computeAnother(Integer i){
    return CompletableFuture.supplyAsync(() -> 10 + i);
}
CompletableFuture<Integer> finalResult = compute().thenCompose(this::computeAnother);

因此，如果想要继续嵌套链接CompletableFuture 方法，那么最好使用thenCompose()。

并行运行多个任务

当我们需要并行执行多个任务时，我们通常希望等待所有它们执行，然后处理它们的组合结果。

该CompletableFuture.allOf静态方法允许等待所有的完成任务：

API

public static CompletableFuture<Void> allOf(CompletableFuture<?>... cfs){...}

实战

CompletableFuture<String> future1  
  = CompletableFuture.supplyAsync(() -> "Hello");
CompletableFuture<String> future2  
  = CompletableFuture.supplyAsync(() -> "Beautiful");
CompletableFuture<String> future3  
  = CompletableFuture.supplyAsync(() -> "World");
 
CompletableFuture<Void> combinedFuture 
  = CompletableFuture.allOf(future1, future2, future3);
 
// ...
 
combinedFuture.get();
 
assertTrue(future1.isDone());
assertTrue(future2.isDone());
assertTrue(future3.isDone());

请注意，CompletableFuture.allOf()的返回类型是CompletableFuture 。这种方法的局限性在于它不会返回所有任务的综合结果。相反，你必须手动从Futures获取结果。幸运的是，CompletableFuture.join()方法和Java 8 Streams API可以解决：

String combined = Stream.of(future1, future2, future3)
  .map(CompletableFuture::join)
  .collect(Collectors.joining(" "));
 
assertEquals("Hello Beautiful World", combined);

CompletableFuture 提供了 join() 方法，它的功能和 get() 方法是一样的，都是阻塞获取值，它们的区别在于 join() 抛出的是 unchecked Exception。这使得它可以在Stream.map（）方法中用作方法引用。

异常处理

说到这里，我们顺便来说下 CompletableFuture 的异常处理。这里我们要介绍两个方法：

public CompletableFuture<T> exceptionally(Function<Throwable, ? extends T> fn);
public <U> CompletionStage<U> handle(BiFunction<? super T, Throwable, ? extends U> fn);

看下代码

CompletableFuture.supplyAsync(() -> "resultA")
    .thenApply(resultA -> resultA + " resultB")
    .thenApply(resultB -> resultB + " resultC")
    .thenApply(resultC -> resultC + " resultD");

上面的代码中，任务 A、B、C、D 依次执行，如果任务 A 抛出异常（当然上面的代码不会抛出异常），那么后面的任务都得不到执行。如果任务 C 抛出异常，那么任务 D 得不到执行。

那么我们怎么处理异常呢？看下面的代码，我们在任务 A 中抛出异常，并对其进行处理：

CompletableFuture<String> future = CompletableFuture.supplyAsync(() -> {
    throw new RuntimeException();
})
        .exceptionally(ex -> "errorResultA")
        .thenApply(resultA -> resultA + " resultB")
        .thenApply(resultB -> resultB + " resultC")
        .thenApply(resultC -> resultC + " resultD");

System.out.println(future.join());

上面的代码中，任务 A 抛出异常，然后通过 .exceptionally() 方法处理了异常，并返回新的结果，这个新的结果将传递给任务 B。所以最终的输出结果是：

errorResultA resultB resultC resultD

String name = null;
 
// ...
 
CompletableFuture<String> completableFuture  
  =  CompletableFuture.supplyAsync(() -> {
      if (name == null) {
          throw new RuntimeException("Computation error!");
      }
      return "Hello, " + name;
  })}).handle((s, t) -> s != null ? s : "Hello, Stranger!");
 
assertEquals("Hello, Stranger!", completableFuture.get());

当然，它们也可以都为 null，因为如果它作用的那个 CompletableFuture 实例没有返回值的时候，s 就是 null。

Async后缀方法

CompletableFuture类中的API的大多数方法都有两个带有Async后缀的附加修饰。这些方法表示用于异步线程。

没有Async后缀的方法使用调用线程运行下一个执行线程阶段。不带Async方法使用ForkJoinPool.commonPool()线程池的fork / join实现运算任务。带有Async方法使用传递式的Executor任务去运行。

下面附带一个案例，可以看到有thenApplyAsync方法。在程序内部，线程被包装到ForkJoinTask实例中。这样可以进一步并行化你的计算并更有效地使用系统资源。

CompletableFuture<String> completableFuture  
  = CompletableFuture.supplyAsync(() -> "Hello");
 
CompletableFuture<String> future = completableFuture
  .thenApplyAsync(s -> s + " World");
 
assertEquals("Hello World", future.get());

JDK 9 CompletableFuture API

在Java 9中， CompletableFuture API通过以下更改得到了进一步增强：

新工厂方法增加了
支持延迟和超时
改进了对子类化的支持。

引入了新的实例API：

Executor defaultExecutor()
CompletableFuture newIncompleteFuture()

CompletableFuture copy()

CompletionStage minimalCompletionStage()

CompletableFuture completeAsync(Supplier supplier, Executor executor)

CompletableFuture completeAsync(Supplier supplier)

CompletableFuture orTimeout(long timeout, TimeUnit unit)

CompletableFuture completeOnTimeout(T value, long timeout, TimeUnit unit)

还有一些静态实用方法：

Executor delayedExecutor(long delay, TimeUnit unit, Executor executor)
Executor delayedExecutor(long delay, TimeUnit unit)
CompletionStage completedStage(U value)

CompletionStage failedStage(Throwable ex)

CompletableFuture failedFuture(Throwable ex)

最后，为了解决超时问题，Java 9又引入了两个新功能：

orTimeout()
completeOnTimeout()

结论

在本文中，我们描述了CompletableFuture类的方法和典型用例。

【相关推荐：Java视频教程】

The above is the detailed content of Introduction to the usage of Java8's CompletableFuture (with examples). For more information, please follow other related articles on the PHP Chinese website!

Statement：

This article is reproduced at:segmentfault.com. If there is any infringement, please contact admin@php.cn delete

Previous article：Introduction to the usage of ThreadPoolExecutor thread pool in JavaNext article：Introduction to the usage of ThreadPoolExecutor thread pool in Java

See more

Introduction to the usage of Java8's CompletableFuture (with examples)

Related articles