java多线程之join使用

java多线程中join这个方法相信大家都听说过。join这个方法的主要作用就是当前线程等待子线程运行结束。下面是一个join的小示例

public class ThreadJoin {
	public static void main(String[] args) {
		Thread t1 = new Thread(()->{
			IntStream.range(1, 1000)
				.forEach(i -> System.out.println(Thread.currentThread().getName() + "->" + i));
		});
		t1.start();
		try {
			t1.join();
		} catch (InterruptedException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}
		IntStream.range(1, 1000)
			.forEach(i -> System.out.println(Thread.currentThread().getName() + "->" + i));

	}
}

以上代码，当不进行t1.join的时候，t1x线程会和main线程交替输出，当进行t1.join的时候，main线程会先等t1线程执行完以后再执行输出i循环的代码。再看以下代码

public class ThreadJoin {
	public static void main(String[] args) {
		Thread t1 = new Thread(()->{
			IntStream.range(1, 1000)
				.forEach(i -> System.out.println(Thread.currentThread().getName() + "->" + i));
		});
		Thread t2 = new Thread(()->{
			IntStream.range(1, 1000)
				.forEach(i -> System.out.println(Thread.currentThread().getName() + "->" + i));
		});
		t1.start();
		t2.start();
		try {
			t1.join();
			t2.join();
		} catch (InterruptedException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}
		System.out.println("all of tasks finish done");
		IntStream.range(1, 1000)
			.forEach(i -> System.out.println(Thread.currentThread().getName() + "->" + i));

	}
}

以上代码会t1和t2同时执行，等待t1和t2线程都执行结束之后再进行main线程的执行。接着往下看

public class ThreadJoin {
	public static void main(String[] args) {
		Thread t1 = new Thread(()->{
			try {
				System.out.println("t1 is running");
				Thread.sleep(10_000);
				System.out.println("t1 is done");
			} catch (Exception e) {
				e.printStackTrace();
			}
		});
		t1.start();
		try {
			t1.join(100);
		} catch (InterruptedException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}
		System.out.println("all of tasks finish done");
		IntStream.range(1, 100)
			.forEach(i -> System.out.println(Thread.currentThread().getName() + "->" + i));

	}
}

此处t1.join(100)当t1线程执行超过100毫秒之后main不再等待t1线程执行。

下面举一个join的实例，一个a应用要开启三个线程去采集数据，三个线程分别需要10s,15s，30s，计算最终花费的时间。

public class ThreadJoin {
	public static void main(String[] args) throws InterruptedException {
		long startTimestamp = System.currentTimeMillis();
		Thread t1 = new Thread(new CaptureRunnable("M1",10000L));
		Thread t2 = new Thread(new CaptureRunnable("M2",30000L));
		Thread t3 = new Thread(new CaptureRunnable("M3",15000L));
		
		t1.start();
		t2.start();
		t3.start();
		t1.join();
		t2.join();
		t3.join();
		long endTimestamp = System.currentTimeMillis();
		System.out.printf("save data begin timestamp is:%s, end timestamp is%s", startTimestamp, endTimestamp);
	}
	
}

class CaptureRunnable implements Runnable{
	private String machineName;
	private long spendTime;
	public CaptureRunnable(String machineName, long spendTime) {
		this.machineName= machineName;
		this.spendTime = spendTime;
	}
	
	@Override
	public void run() {
		try {
			Thread.sleep(spendTime);
			System.out.println(machineName + "completed data capture and success");
		} catch (InterruptedException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}
	}
	
	public String getResult() {
		return machineName + " finish.";
	}
	
}

此时如果不给三个线程加上join则三个线程还未采集完数据，主线程就输出花费时间。这么显然不准确，因此给三个线程进行join，这样子就会等三个线程都完成采集数据之后再进行花费时间的计算。