在上一篇文章中写道用一个静态的变量保存线程的执行状态,并用时间等待的方法后来仔细考虑,其实是
线程不安全的.多个线程同时执行这个类时,这个静态变量的值就不能保证了.
用一个线程同步的Map保存这个值,勉强能实现[每个线程生产一个不重复的map的key]
但是这样很麻烦.
java. util. concurrent.CountDownLatch
却能完美的实现.能线程安全的计数,因为每个实现的主线程在并发的情况下java.util.concurrent.CountDownLatch; 是新的实例 不会像上一篇一样需要处理计数器的线程安全.
具体代码如下
package org.masque.effective;import java.util.concurrent.CountDownLatch;import java.util.concurrent.TimeUnit;/** * * @author masque.java@gmail.com * */public class ThreadTest { public static void main(String[] args) throws InterruptedException { CountDownLatch latch=new CountDownLatch(3);//数量标示线程的个数 (new Thread1(latch)).start(); (new Thread2(latch)).start(); (new Thread3(latch)).start(); latch.await();//等待线程执行完成.还有一个重载的方法可以设置超时时间 System.out.println("over!"); }}class Thread1 extends Thread{ CountDownLatch latch; public Thread1(CountDownLatch latch){ this.latch = latch; } @Override public void run() { for (int i=10;i<20;i++) { System.out.println("Thread1===========:"+i); try { TimeUnit.SECONDS.sleep(1); } catch (InterruptedException e) { e.printStackTrace(); } } latch.countDown();//完成后计数 }}class Thread2 extends Thread{ CountDownLatch latch; public Thread2(CountDownLatch latch){ this.latch = latch; } @Override public void run() { for (int i=10;i<20;i++) { System.out.println("Thread2===========:"+i); try { TimeUnit.SECONDS.sleep(1); } catch (InterruptedException e) { e.printStackTrace(); } } latch.countDown();//完成后计数 }}class Thread3 extends Thread{ CountDownLatch latch; public Thread3(CountDownLatch latch){ this.latch = latch; } @Override public void run() { for (int i=10;i<20;i++) { System.out.println("Thread3===========:"+i); try { TimeUnit.SECONDS.sleep(1); } catch (InterruptedException e) { e.printStackTrace(); } } latch.countDown();//完成后计数 }}