Semapher信号量也是Java中的一个同步器，与CountDownLatch和CycleBarrier不同的是，它内部的计数器是递增的，并且在一开始初始化Semaphoer时可以指定一个初始值，但是并不需要知道需要同步的线程个数，而是在需要同步的地方调用acquire方法时指定需要同步的线程个数。
我们通过下面一个例子来看一下Semapher效果：
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.Semaphore;
public class SemaphoreTest {
private static Semaphore semaphore = new Semaphore(0);
public static void main(String[] args) throws InterruptedException{
ExecutorService executorService = Executors.newFixedThreadPool(2);
executorService.submit(new Runnable() {
@Override
public void run() {
try {
System.out.println(Thread.currentThread() + "over");
semaphore.release();
} catch (Exception e) {
e.printStackTrace();
}
}
});
executorService.submit(new Runnable() {
@Override
public void run() {
try {
System.out.println(Thread.currentThread() + "over");
semaphore.release();
} catch (Exception e) {
e.printStackTrace();
}
}
});
semaphore.acquire(2);
System.out.println("all child thread over!");
executorService.shutdown();
}
}
如上代码首先创建了一个信号量实例，构造函数的入参为0，说明当前信号量计数器的值为0。然后，main函数向线程池添加两个线程任务，在每个线程内部调用信号量的acquire方法，传参为2说明调用acquire方法的线程会一直阻塞，知道信号量的技术变为2才会返回。如果构造Semaphore时，传递的参数为N，并在M个线程中调用了该信号量的release方法，那么在调用acquire使M个线程同步时传递的参数应该是M+N。
下面举例子来模拟CycliBarrier复用的功能，代码如下：
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.Semaphore;
public class Semaphoer {
private static Semaphore semaphore = new Semaphore(0);
public static void main(String[] args) throws InterruptedException {
ExecutorService executorService = Executors.newFixedThreadPool(2);
executorService.submit(new Runnable() {
@Override
public void run() {
try {
System.out.println(Thread.currentThread() + "A task over");
semaphore.release();
} catch (Exception e) {
e.printStackTrace();
}
}
});
executorService.submit(new Runnable() {
@Override
public void run() {
try {
System.out.println(Thread.currentThread() + "A task over");
semaphore.release();
} catch (Exception e) {
e.printStackTrace();
}
}
});
semaphore.acquire(2);
System.out.println("task A is over");
executorService.submit(new Runnable() {
@Override
public void run() {
try {
System.out.println(Thread.currentThread() + "B task over");
semaphore.release();
} catch (Exception e) {
e.printStackTrace();
}
}
});
executorService.submit(new Runnable() {
@Override
public void run() {
try {
System.out.println(Thread.currentThread() + "B task over");
semaphore.release();
} catch (Exception e) {
e.printStackTrace();
}
}
});
semaphore.acquire(2);
System.out.println("task B is over");
executorService.shutdown();
}
}
如上代码首先将线程A和线程B加入到线程池。主线程执行代码(1)后被阻塞。线程A和线程B调用release方法后信号量的值变为了2，这时候主线程的aquire方法会在获取到2个信号量后返回(返回后当前信号量值为0)。然后主线程添加线程C和线程D到线程池，之后主线程执行代码(2)后被阻塞(因为主线程要获取2个信号量，而当前信号量个数为0)。当线程C和线程D执行完release 方法后，主线程才返回。从本例子可以看出，Semaphore 在某种程度上实现了CyclicBarrier 的复用功能。
