JUC并发编程第二章之CompletableFuture[加强版的线程]
目录
1、 创建线程的几种方式
new thread:
//构造方法给指定的线程指定名称,推荐
Thread thread = new Thread(t1) {
@Override
public void run() {
//要执行的任务
}
};
//启动线程
thread.start();
new runnable:
Runnable runnable = new Runnable() {
@Override
public void run() {
//要执行的任务
}
};
Thread thread = new Thread(runnable);
thread.start();
Future
FutureTask<Integer> task = new FutureTask<>(() -> {
System.out.println(执行的任务);
return 100;
});
new Thread(task,t1).start();
Integer integer = task.get();
System.out.println(结果是+integer);
2、 Future的优缺点
优点: Future配合线程池能够显著提高程序的执行效率
public static void main(String[] args) throws ExecutionException, InterruptedException {
long startTime = System.currentTimeMillis();
ExecutorService executorService = Executors.newFixedThreadPool(3);
FutureTask<String> stringFutureTask1 = new FutureTask<>(() -> {
TimeUnit.SECONDS.sleep(5);
return over;
});
executorService.submit(stringFutureTask1);
FutureTask<String> stringFutureTask2 = new FutureTask<>(() -> {
TimeUnit.SECONDS.sleep(3);
return over;
});
executorService.submit(stringFutureTask2);
FutureTask<String> stringFutureTask3 = new FutureTask<>(() -> {
TimeUnit.SECONDS.sleep(3);
return over;
});
executorService.submit(stringFutureTask3);
System.out.println(stringFutureTask1.get());
System.out.println(stringFutureTask2.get());
System.out.println(stringFutureTask3.get());
executorService.shutdown();
long endTime = System.currentTimeMillis();
System.out.println(耗时了+(endTime-startTime));
}
输出结果:
over
over
over
耗时了5084
可以看到如果是串行输出,结果是5s+3s+3s的耗时;
- 缺点1:Get 方法容易导致堵塞
从上面的程序,我必须得到stringFutureTask1 执行完后,主线程才能执行任务, 输出耗时时间,我们的期望是 stringFutureTask1在耗时5s的时间内,主线程忙其他事情, 并询问下stringFutureTask1 是否执行完毕, 如果执行完毕,则输出耗时时间;
- 缺点2:Get 方法没有最大等待时间
比如我们最多只能等待5s, 但是如果get()方法执行10s的话, 则会影响我们的程序
- 缺点3: isDone 容易导致cpu轮训空转
public static void main(String[] args) throws ExecutionException, InterruptedException {
long startTime = System.currentTimeMillis();
ExecutorService executorService = Executors.newFixedThreadPool(3);
FutureTask<String> stringFutureTask1 = new FutureTask<>(() -> {
TimeUnit.SECONDS.sleep(5);
return over;
});
executorService.submit(stringFutureTask1);
executorService.shutdown();
while (true) {
if (stringFutureTask1.isDone()){
long endTime = System.currentTimeMillis();
System.out.println(耗时了 + (endTime - startTime));
}
}
}
3、CompletableFuture对Future的改进
对于真正的异步处理我们希望可以通过传入回调函数,在Future结束时自动调用该函数;这样我们就不用等待结果了;
3.1、CompletableFuture的基本结构
public class CompletableFuture<T> implements Future<T>, CompletionStage<T>
-
CompletionStage代表计算过程的一个阶段,一个阶段完成以后可能会触发另一个阶段; -
一个阶段的计算,可以是一个function,consumer或者runnable. 比如stage.thenApply(x->square(x)).thenAccept(x->system.out.print(x)).thenReturn(()->system.out.println())
-
一个阶段的执行可能是被单个阶段的完成触发,也可能是由多个阶段一起触发;
3.2、创建CompletableFuture四种方式
- runAsync 无返回值
| 返回值 | 具体方法 |
|---|---|
| static CompletableFuture |
runAsync(Runnable runnable) |
| static CompletableFuture |
runAsync(Runnable runnable, Executor executor) |
代码示例
public static void main(String[] args) throws Exception {
ExecutorService executorService = Executors.newFixedThreadPool(3);
CompletableFuture<Void> voidCompletableFuture = CompletableFuture.runAsync(() -> {
System.out.println(Thread.currentThread().getName());
try {
TimeUnit.MILLISECONDS.sleep(500);
} catch (InterruptedException e) {
e.printStackTrace();
}
},executorService);
System.out.println(voidCompletableFuture.get());
}
- supplyAsync有返回值
| 返回值 | 具体方法 |
|---|---|
| static CompletableFuture | supplyAsync(Supplier supplier) |
| static CompletableFuture | supplyAsync(Supplier supplier, Executor executor) |
public static void main(String[] args) throws Exception {
ExecutorService executorService = Executors.newFixedThreadPool(3);
CompletableFuture<String> stringCompletableFuture = CompletableFuture.supplyAsync(() -> {
System.out.println(Thread.currentThread().getName());
try {
TimeUnit.MILLISECONDS.sleep(500);
} catch (InterruptedException e) {
e.printStackTrace();
}
return over;
}, executorService);
System.out.println(stringCompletableFuture.get());<br></br> System.out.println(stringCompletableFuture.jion());
}
输出结果
pool-1-thread-1
over
- 指的注意的是,join方法和get方法,都可以获取返回值,而join没有处理异常而已;他不会要求抛出异常
3.3、CompletableFuture的流式调用
public static void main(String[] args) throws Exception {
ExecutorService executorService = Executors.newFixedThreadPool(3);
CompletableFuture<String> stringCompletableFuture = CompletableFuture.supplyAsync(() -> {
System.out.println(Thread.currentThread().getName());
try {
TimeUnit.MILLISECONDS.sleep(500);
} catch (InterruptedException e) {
e.printStackTrace();
}
return over;
}, executorService).whenComplete((v, e) -> {
System.out.println(hello +v);
if (e == null) {
System.out.println(没有异常,更新完成);
}
}).exceptionally(s -> {
s.printStackTrace();
System.out.println(异常了,主线程先忙其他事情);
return null;
});
System.out.println(主线程工作);
}
扩展: 函数值编程接口
3.4、CompletableFuture常见的用法
对计算结果进行合并
public static void main(String[] args) throws Exception {
CompletableFuture<Integer> task1 = CompletableFuture.supplyAsync(() -> {
return 1;
});
CompletableFuture<Integer> task2 = CompletableFuture.supplyAsync(() -> {
return 10;
});
CompletableFuture<Integer> integerCompletableFuture = task1.thenCombine(task2, (x, y) -> {
return x + y;
});
System.out.println(integerCompletableFuture.get());
}
