下面的代码片段将在单独的线程中启动一个操作，然后等待最多10秒以使操作完成。如果操作没有及时完成，代码将尝试取消该操作，然后继续其愉快的方式。即使操作不能轻易取消，父线程也不会等待子线程终止。

ExecutorService executorService = getExecutorService();
Future<SomeClass> future = executorService.submit(new Callable<SomeClass>() {
    public SomeClass call() {
        // Perform long-running task, return result. The code should check
        // interrupt status regularly, to facilitate cancellation.
    }
});
try {
    // Real life code should define the timeout as a constant or
    // retrieve it from configuration
    SomeClass result = future.get(10, TimeUnit.SECONDS);
    // Do something with the result
} catch (TimeoutException e) {
    future.cancel(true);
    // Perform other error handling, e.g. logging, throwing an exception
}

getExecutorService()方法可以通过多种方式实现。如果您没有任何特殊的要求，您可以简单地调用Executors.newCachedThreadPool()进行线程池，没有线程数量的上限。

下面的代码片段将在单独的线程中启动一个操作，然后等待最多10秒以使操作完成。如果操作没有及时完成，代码将尝试取消该操作，然后继续其愉快的方式。即使操作不能轻易取消，父线程也不会等待子线程终止。

ExecutorService executorService = getExecutorService();
Future<SomeClass> future = executorService.submit(new Callable<SomeClass>() {
    public SomeClass call() {
        // Perform long-running task, return result. The code should check
        // interrupt status regularly, to facilitate cancellation.
    }
});
try {
    // Real life code should define the timeout as a constant or
    // retrieve it from configuration
    SomeClass result = future.get(10, TimeUnit.SECONDS);
    // Do something with the result
} catch (TimeoutException e) {
    future.cancel(true);
    // Perform other error handling, e.g. logging, throwing an exception
}

getExecutorService()方法可以通过多种方式实现。如果您没有任何特殊的要求，您可以简单地调用Executors.newCachedThreadPool()进行线程池，没有线程数量的上限。

不久前，我为此创建了一个helper类。伟大的工作:

import java.util.concurrent.BrokenBarrierException;
import java.util.concurrent.CyclicBarrier;
/**
 * TimeOut class - used for stopping a thread that is taking too long
 * @author Peter Goransson
 *
 */
public class TimeOut {

    Thread interrupter;
    Thread target;
    long timeout;
    boolean success;
    boolean forceStop;

    CyclicBarrier barrier;

    /**
     * 
     * @param target The Runnable target to be executed
     * @param timeout The time in milliseconds before target will be interrupted or stopped
     * @param forceStop If true, will Thread.stop() this target instead of just interrupt() 
     */
    public TimeOut(Runnable target, long timeout, boolean forceStop) {      
        this.timeout = timeout;
        this.forceStop = forceStop;

        this.target = new Thread(target);       
        this.interrupter = new Thread(new Interrupter());

        barrier = new CyclicBarrier(2); // There will always be just 2 threads waiting on this barrier
    }

    public boolean execute() throws InterruptedException {  

        // Start target and interrupter
        target.start();
        interrupter.start();

        // Wait for target to finish or be interrupted by interrupter
        target.join();  

        interrupter.interrupt(); // stop the interrupter    
        try {
            barrier.await(); // Need to wait on this barrier to make sure status is set
        } catch (BrokenBarrierException e) {
            // Something horrible happened, assume we failed
            success = false;
        } 

        return success; // status is set in the Interrupter inner class
    }

    private class Interrupter implements Runnable {

        Interrupter() {}

        public void run() {
            try {
                Thread.sleep(timeout); // Wait for timeout period and then kill this target
                if (forceStop) {
                  target.stop(); // Need to use stop instead of interrupt since we're trying to kill this thread
                }
                else {
                    target.interrupt(); // Gracefully interrupt the waiting thread
                }
                System.out.println("done");             
                success = false;
            } catch (InterruptedException e) {
                success = true;
            }


            try {
                barrier.await(); // Need to wait on this barrier
            } catch (InterruptedException e) {
                // If the Child and Interrupter finish at the exact same millisecond we'll get here
                // In this weird case assume it failed
                success = false;                
            } 
            catch (BrokenBarrierException e) {
                // Something horrible happened, assume we failed
                success = false;
            }

        }

    }
}

它的名称是这样的:

long timeout = 10000; // number of milliseconds before timeout
TimeOut t = new TimeOut(new PhotoProcessor(filePath, params), timeout, true);
try {                       
  boolean sucess = t.execute(); // Will return false if this times out
  if (!sucess) {
    // This thread timed out
  }
  else {
    // This thread ran completely and did not timeout
  }
} catch (InterruptedException e) {}  

下面是我非常简单的使用helper类来运行或调用一段Java代码:-)

这是基于BalusC的精彩回答

package com.mycompany.util.concurrent;

import java.util.concurrent.Callable;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.Future;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;

/**
 * Calling {@link Callable#call()} or Running {@link Runnable#run()} code
 * with a timeout based on {@link Future#get(long, TimeUnit))}
 * @author pascaldalfarra
 *
 */
public class CallableHelper
{

    private CallableHelper()
    {
    }

    public static final void run(final Runnable runnable, int timeoutInSeconds)
    {
        run(runnable, null, timeoutInSeconds);
    }

    public static final void run(final Runnable runnable, Runnable timeoutCallback, int timeoutInSeconds)
    {
        call(new Callable<Void>()
        {
            @Override
            public Void call() throws Exception
            {
                runnable.run();
                return null;
            }
        }, timeoutCallback, timeoutInSeconds); 
    }

    public static final <T> T call(final Callable<T> callable, int timeoutInSeconds)
    {
        return call(callable, null, timeoutInSeconds); 
    }

    public static final <T> T call(final Callable<T> callable, Runnable timeoutCallback, int timeoutInSeconds)
    {
        ExecutorService executor = Executors.newSingleThreadExecutor();
        try
        {
            Future<T> future = executor.submit(callable);
            T result = future.get(timeoutInSeconds, TimeUnit.SECONDS);
            System.out.println("CallableHelper - Finished!");
            return result;
        }
        catch (TimeoutException e)
        {
            System.out.println("CallableHelper - TimeoutException!");
            if(timeoutCallback != null)
            {
                timeoutCallback.run();
            }
        }
        catch (InterruptedException e)
        {
            e.printStackTrace();
        }
        catch (ExecutionException e)
        {
            e.printStackTrace();
        }
        finally
        {
            executor.shutdownNow();
            executor = null;
        }

        return null;
    }

}

我认为你应该看看适当的并发处理机制(线程运行到无限循环本身听起来不太好，顺便说一句)。确保你阅读了一些关于“杀死”或“停止”线程主题的内容。

你所描述的，听起来很像一个“会合”，所以你可能想看看CyclicBarrier。

可能有其他结构(例如使用CountDownLatch)可以解决您的问题(一个线程等待闩锁超时，另一个线程应该在完成工作后倒数闩锁，这将在超时后或闩锁倒计时被调用时释放您的第一个线程)。

我通常推荐这方面的两本书:《Java并发编程》和《Java并发实践》。

现在，我遇到了这样的问题。它恰好解码图片。解码过程耗时太长，导致屏幕黑屏。l添加一个时间控制器:当时间太长时，从当前线程中弹出。 差异如下:

   ExecutorService executor = Executors.newSingleThreadExecutor();
   Future<Bitmap> future = executor.submit(new Callable<Bitmap>() {
       @Override
       public Bitmap call() throws Exception {
       Bitmap bitmap = decodeAndScaleBitmapFromStream(context, inputUri);// do some time consuming operation
       return null;
            }
       });
       try {
           Bitmap result = future.get(1, TimeUnit.SECONDS);
       } catch (TimeoutException e){
           future.cancel(true);
       }
       executor.shutdown();
       return (bitmap!= null);