前言
以前须要异步执行一个工作时,个别是用Thread或者线程池Executor去创立。如果须要返回值,则是调用Executor.submit获取Future。然而多个线程存在依赖组合,咱们又能怎么办?可应用同步组件CountDownLatch、CyclicBarrier等;其实有简略的办法,就是用CompeletableFuture
- 线程工作的创立
- 线程工作的串行执行
- 线程工作的并行执行
- 解决工作后果和异样
- 多任务的简略组合
- 勾销执行线程工作
- 工作后果的获取和实现与否判断
关注公众号,一起交换,微信搜一搜: 潜行前行
github地址,感激star
1 创立异步线程工作
依据supplier创立CompletableFuture工作
//应用内置线程ForkJoinPool.commonPool(),依据supplier构建执行工作
public static <U> CompletableFuture<U> supplyAsync(Supplier<U> supplier)
//指定自定义线程,依据supplier构建执行工作
public static <U> CompletableFuture<U> supplyAsync(Supplier<U> supplier, Executor executor)
依据runnable创立CompletableFuture工作
//应用内置线程ForkJoinPool.commonPool(),依据runnable构建执行工作
public static CompletableFuture<Void> runAsync(Runnable runnable)
//指定自定义线程,依据runnable构建执行工作
public static CompletableFuture<Void> runAsync(Runnable runnable, Executor executor)
-
应用示例
ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<Void> rFuture = CompletableFuture .runAsync(() -> System.out.println("hello siting"), executor); //supplyAsync的应用 CompletableFuture<String> future = CompletableFuture .supplyAsync(() -> { System.out.print("hello "); return "siting"; }, executor); //阻塞期待,runAsync 的future 无返回值,输入null System.out.println(rFuture.join()); //阻塞期待 String name = future.join(); System.out.println(name); executor.shutdown(); // 线程池须要敞开 --------输入后果-------- hello siting null hello siting
常量值作为CompletableFuture返回
//有时候是须要构建一个常量的CompletableFuture public static <U> CompletableFuture<U> completedFuture(U value)
2 线程串行执行
工作实现则运行action,不关怀上一个工作的后果,无返回值
public CompletableFuture<Void> thenRun(Runnable action)
public CompletableFuture<Void> thenRunAsync(Runnable action)
//action用指定线程池执行
public CompletableFuture<Void> thenRunAsync(Runnable action, Executor executor)
-
应用示例
CompletableFuture<Void> future = CompletableFuture .supplyAsync(() -> "hello siting", executor) .thenRunAsync(() -> System.out.println("OK"), executor); executor.shutdown(); --------输入后果-------- OK
工作实现则运行action,依赖上一个工作的后果,无返回值
public CompletableFuture<Void> thenAccept(Consumer<? super T> action)
public CompletableFuture<Void> thenAcceptAsync(Consumer<? super T> action)
//action用指定线程池执行
public CompletableFuture<Void> thenAcceptAsync(Consumer<? super T> action, Executor executor)
-
应用示例
ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<Void> future = CompletableFuture .supplyAsync(() -> "hello siting", executor) .thenAcceptAsync(System.out::println, executor); executor.shutdown(); --------输入后果-------- hello siting
工作实现则运行fn,依赖上一个工作的后果,有返回值
public <U> CompletableFuture<U> thenApply(Function<? super T,? extends U> fn)
public <U> CompletableFuture<U> thenApplyAsync(Function<? super T,? extends U> fn)
//fn用指定线程池执行
public <U> CompletableFuture<U> thenApplyAsync(Function<? super T,? extends U> fn, Executor executor)
-
应用示例
ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<String> future = CompletableFuture .supplyAsync(() -> "hello world", executor) .thenApplyAsync(data -> { System.out.println(data); return "OK"; }, executor); System.out.println(future.join()); executor.shutdown(); --------输入后果-------- hello world OK
thenCompose – 工作实现则运行fn,依赖上一个工作的后果,有返回值
-
相似thenApply(区别是thenCompose的返回值是CompletionStage,thenApply则是返回 U),提供该办法为了和其余CompletableFuture工作更好地配套组合应用
public <U> CompletableFuture<U> thenCompose(Function<? super T, ? extends CompletionStage<U>> fn) public <U> CompletableFuture<U> thenComposeAsync(Function<? super T, ? extends CompletionStage<U>> fn) public <U> CompletableFuture<U> thenComposeAsync(Function<? super T, ? extends CompletionStage<U>> fn, Executor executor)
-
应用示例
//第一个异步工作,常量工作 CompletableFuture<String> f = CompletableFuture.completedFuture("OK"); //第二个异步工作 ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<String> future = CompletableFuture .supplyAsync(() -> "hello world", executor) .thenComposeAsync(data -> { System.out.println(data); return f; //应用第一个工作作为返回 }, executor); System.out.println(future.join()); executor.shutdown(); --------输入后果-------- hello world OK
3 线程并行执行
两个CompletableFuture并行执行完,而后执行action,不依赖上两个工作的后果,无返回值
public CompletableFuture<Void> runAfterBoth(CompletionStage<?> other, Runnable action)
public CompletableFuture<Void> runAfterBothAsync(CompletionStage<?> other, Runnable action)
public CompletableFuture<Void> runAfterBothAsync(CompletionStage<?> other, Runnable action, Executor executor)
-
应用示例
//第一个异步工作,常量工作 CompletableFuture<String> first = CompletableFuture.completedFuture("hello world"); ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<Void> future = CompletableFuture //第二个异步工作 .supplyAsync(() -> "hello siting", executor) // () -> System.out.println("OK") 是第三个工作 .runAfterBothAsync(first, () -> System.out.println("OK"), executor); executor.shutdown(); --------输入后果-------- OK
两个CompletableFuture并行执行完,而后执行action,依赖上两个工作的后果,无返回值
//调用方工作和other并行实现后执行action,action再依赖生产两个工作的后果,无返回值
public <U> CompletableFuture<Void> thenAcceptBoth(CompletionStage<? extends U> other,
BiConsumer<? super T, ? super U> action)
//两个工作异步实现,fn再依赖生产两个工作的后果,无返回值,应用默认线程池
public <U> CompletableFuture<Void> thenAcceptBothAsync(CompletionStage<? extends U> other,
BiConsumer<? super T, ? super U> action)
//两个工作异步实现,fn(用指定线程池执行)再依赖生产两个工作的后果,无返回值
public <U> CompletableFuture<Void> thenAcceptBothAsync(CompletionStage<? extends U> other,
BiConsumer<? super T, ? super U> action, Executor executor)
-
应用示例
//第一个异步工作,常量工作 CompletableFuture<String> first = CompletableFuture.completedFuture("hello world"); ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<Void> future = CompletableFuture //第二个异步工作 .supplyAsync(() -> "hello siting", executor) // (w, s) -> System.out.println(s) 是第三个工作 .thenAcceptBothAsync(first, (s, w) -> System.out.println(s), executor); executor.shutdown(); --------输入后果-------- hello siting
两个CompletableFuture并行执行完,而后执行fn,依赖上两个工作的后果,有返回值
//调用方工作和other并行实现后,执行fn,fn再依赖生产两个工作的后果,有返回值
public <U,V> CompletableFuture<V> thenCombine(CompletionStage<? extends U> other,
BiFunction<? super T,? super U,? extends V> fn)
//两个工作异步实现,fn再依赖生产两个工作的后果,有返回值,应用默认线程池
public <U,V> CompletableFuture<V> thenCombineAsync(CompletionStage<? extends U> other,
BiFunction<? super T,? super U,? extends V> fn)
//两个工作异步实现,fn(用指定线程池执行)再依赖生产两个工作的后果,有返回值
public <U,V> CompletableFuture<V> thenCombineAsync(CompletionStage<? extends U> other,
BiFunction<? super T,? super U,? extends V> fn, Executor executor)
-
应用示例
//第一个异步工作,常量工作 CompletableFuture<String> first = CompletableFuture.completedFuture("hello world"); ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<String> future = CompletableFuture //第二个异步工作 .supplyAsync(() -> "hello siting", executor) // (w, s) -> System.out.println(s) 是第三个工作 .thenCombineAsync(first, (s, w) -> { System.out.println(s); return "OK"; }, executor); System.out.println(future.join()); executor.shutdown(); --------输入后果-------- hello siting OK
4 线程并行执行,谁先执行完则谁触发下一工作(二者选其最快)
上一个工作或者other工作实现, 运行action,不依赖前一工作的后果,无返回值
public CompletableFuture<Void> runAfterEither(CompletionStage<?> other, Runnable action)
public CompletableFuture<Void> runAfterEitherAsync(CompletionStage<?> other, Runnable action)
//action用指定线程池执行
public CompletableFuture<Void> runAfterEitherAsync(CompletionStage<?> other,
Runnable action, Executor executor)
-
应用示例
//第一个异步工作,休眠1秒,保障最晚执行晚 CompletableFuture<String> first = CompletableFuture.supplyAsync(()->{ try{ Thread.sleep(1000); }catch (Exception e){} System.out.println("hello world"); return "hello world"; }); ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<Void> future = CompletableFuture //第二个异步工作 .supplyAsync(() ->{ System.out.println("hello siting"); return "hello siting"; } , executor) //() -> System.out.println("OK") 是第三个工作 .runAfterEitherAsync(first, () -> System.out.println("OK") , executor); executor.shutdown(); --------输入后果-------- hello siting OK
上一个工作或者other工作实现, 运行action,依赖最先实现工作的后果,无返回值
public CompletableFuture<Void> acceptEither(CompletionStage<? extends T> other,
Consumer<? super T> action)
public CompletableFuture<Void> acceptEitherAsync(CompletionStage<? extends T> other,
Consumer<? super T> action, Executor executor)
//action用指定线程池执行
public CompletableFuture<Void> acceptEitherAsync(CompletionStage<? extends T> other,
Consumer<? super T> action, Executor executor)
-
应用示例
//第一个异步工作,休眠1秒,保障最晚执行晚 CompletableFuture<String> first = CompletableFuture.supplyAsync(()->{ try{ Thread.sleep(1000); }catch (Exception e){} return "hello world"; }); ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<Void> future = CompletableFuture //第二个异步工作 .supplyAsync(() -> "hello siting", executor) // data -> System.out.println(data) 是第三个工作 .acceptEitherAsync(first, data -> System.out.println(data) , executor); executor.shutdown(); --------输入后果-------- hello siting
上一个工作或者other工作实现, 运行fn,依赖最先实现工作的后果,有返回值
public <U> CompletableFuture<U> applyToEither(CompletionStage<? extends T> other,
Function<? super T, U> fn)
public <U> CompletableFuture<U> applyToEitherAsync(CompletionStage<? extends T> other,
Function<? super T, U> fn)
//fn用指定线程池执行
public <U> CompletableFuture<U> applyToEitherAsync(CompletionStage<? extends T> other,
Function<? super T, U> fn, Executor executor)
-
应用示例
//第一个异步工作,休眠1秒,保障最晚执行晚 CompletableFuture<String> first = CompletableFuture.supplyAsync(()->{ try{ Thread.sleep(1000); }catch (Exception e){} return "hello world"; }); ExecutorService executor = Executors.newSingleThreadExecutor(); CompletableFuture<String> future = CompletableFuture //第二个异步工作 .supplyAsync(() -> "hello siting", executor) // data -> System.out.println(data) 是第三个工作 .applyToEitherAsync(first, data -> { System.out.println(data); return "OK"; } , executor); System.out.println(future); executor.shutdown(); --------输入后果-------- hello siting OK
5 解决工作后果或者异样
exceptionally-解决异样
public CompletableFuture<T> exceptionally(Function<Throwable, ? extends T> fn)
- 如果之前的解决环节有异样问题,则会触发exceptionally的调用相当于 try…catch
-
应用示例
CompletableFuture<Integer> first = CompletableFuture .supplyAsync(() -> { if (true) { throw new RuntimeException("main error!"); } return "hello world"; }) .thenApply(data -> 1) .exceptionally(e -> { e.printStackTrace(); // 异样捕获解决,后面两个解决环节的日常都能捕捉 return 0; });
handle-工作实现或者异样时运行fn,返回值为fn的返回
-
相比exceptionally而言,即可解决上一环节的异样也能够解决其失常返回值
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)
-
应用示例
CompletableFuture<Integer> first = CompletableFuture .supplyAsync(() -> { if (true) { throw new RuntimeException("main error!"); } return "hello world"; }) .thenApply(data -> 1) .handleAsync((data,e) -> { e.printStackTrace(); // 异样捕获解决 return data; }); System.out.println(first.join()); --------输入后果-------- java.util.concurrent.CompletionException: java.lang.RuntimeException: main error! ... 5 more null
whenComplete-工作实现或者异样时运行action,有返回值
- whenComplete与handle的区别在于,它不参加返回后果的解决,把它当成监听器即可
- 即便异样被解决,在CompletableFuture外层,异样也会再次复现
-
应用whenCompleteAsync时,返回后果则须要思考多线程操作问题,毕竟会呈现两个线程同时操作一个后果
public CompletableFuture<T> whenComplete(BiConsumer<? super T, ? super Throwable> action) public CompletableFuture<T> whenCompleteAsync(BiConsumer<? super T, ? super Throwable> action) public CompletableFuture<T> whenCompleteAsync(BiConsumer<? super T, ? super Throwable> action, Executor executor)
-
应用示例
CompletableFuture<AtomicBoolean> first = CompletableFuture .supplyAsync(() -> { if (true) { throw new RuntimeException("main error!"); } return "hello world"; }) .thenApply(data -> new AtomicBoolean(false)) .whenCompleteAsync((data,e) -> { //异样捕获解决, 然而异样还是会在外层复现 System.out.println(e.getMessage()); }); first.join(); --------输入后果-------- java.lang.RuntimeException: main error! Exception in thread "main" java.util.concurrent.CompletionException: java.lang.RuntimeException: main error! ... 5 more
6 多个工作的简略组合
public static CompletableFuture<Void> allOf(CompletableFuture<?>... cfs)
public static CompletableFuture<Object> anyOf(CompletableFuture<?>... cfs)
-
应用示例
CompletableFuture<Void> future = CompletableFuture .allOf(CompletableFuture.completedFuture("A"), CompletableFuture.completedFuture("B")); //全副工作都须要执行完 future.join(); CompletableFuture<Object> future2 = CompletableFuture .anyOf(CompletableFuture.completedFuture("C"), CompletableFuture.completedFuture("D")); //其中一个工作行完即可 future2.join();
8 勾销执行线程工作
// mayInterruptIfRunning 无影响;如果工作未实现,则返回异样
public boolean cancel(boolean mayInterruptIfRunning)
//工作是否勾销
public boolean isCancelled()
-
应用示例
CompletableFuture<Integer> future = CompletableFuture .supplyAsync(() -> { try { Thread.sleep(1000); } catch (Exception e) { } return "hello world"; }) .thenApply(data -> 1); System.out.println("工作勾销前:" + future.isCancelled()); // 如果工作未实现,则返回异样,须要对应用exceptionally,handle 对后果解决 future.cancel(true); System.out.println("工作勾销后:" + future.isCancelled()); future = future.exceptionally(e -> { e.printStackTrace(); return 0; }); System.out.println(future.join()); --------输入后果-------- 工作勾销前:false 工作勾销后:true java.util.concurrent.CancellationException at java.util.concurrent.CompletableFuture.cancel(CompletableFuture.java:2276) at Test.main(Test.java:25) 0
9 工作的获取和实现与否判断
// 工作是否执行实现
public boolean isDone()
//阻塞期待 获取返回值
public T join()
// 阻塞期待 获取返回值,区别是get须要返回受检异样
public T get()
//期待阻塞一段时间,并获取返回值
public T get(long timeout, TimeUnit unit)
//未实现则返回指定value
public T getNow(T valueIfAbsent)
//未实现,应用value作为工作执行的后果,工作完结。须要future.get获取
public boolean complete(T value)
//未实现,则是异样调用,返回异样后果,工作完结
public boolean completeExceptionally(Throwable ex)
//判断工作是否因产生异样完结的
public boolean isCompletedExceptionally()
//强制地将返回值设置为value,无论该之前工作是否实现;相似complete
public void obtrudeValue(T value)
//强制地让异样抛出,异样返回,无论该之前工作是否实现;相似completeExceptionally
public void obtrudeException(Throwable ex)
-
应用示例
CompletableFuture<Integer> future = CompletableFuture .supplyAsync(() -> { try { Thread.sleep(1000); } catch (Exception e) { } return "hello world"; }) .thenApply(data -> 1); System.out.println("工作实现前:" + future.isDone()); future.complete(10); System.out.println("工作实现后:" + future.join()); --------输入后果-------- 工作实现前:false 工作实现后:10