Java 多线程循环打印
北城酒卿 人气:0三个线程T1、T2、T3轮流打印ABC，打印n次，如ABCABCABCABC…
N个线程循环打印1-100…
wait-notify
循环打印问题可以通过设置目标值，每个线程想打印目标值，如果拿到锁后这次轮到的数不是它想要的就进入wait
class Wait_Notify_ABC {
private int num;
private static final Object Lock = new Object();
private void print_ABC(int target) {
synchronized (Lock) {
//循环打印
for (int i = 0; i < 10; i++) {
while (num % 3 != target) {
try {
Lock.wait();
} catch (InterruptedException e) {
e.printStackTrace();
}
}
num++;
System.out.print(Thread.currentThread().getName());
Lock.notifyAll();
}
}
}
public static void main(String[] args) {
Wait_Notify_ABC wait_notify_abc = new Wait_Notify_ABC();
new Thread(() -> {
wait_notify_abc.print_ABC(0);
}, "A").start();
new Thread(() -> {
wait_notify_abc.print_ABC(1);
}, "B").start();
new Thread(() -> {
wait_notify_abc.print_ABC(2);
}, "C").start();
}
}
打印1-100问题可以理解为有个全局计数器记录当前打印到了哪个数，其它就和循环打印ABC问题相同。
class Wait_Notify_100 {
private int num;
private static final Object LOCK = new Object();
private int maxnum = 100;
private void printABC(int targetNum) {
while (true) {
synchronized (LOCK) {
while (num % 3 != targetNum) {
if (num >= maxnum) {
break;
}
try {
LOCK.wait();
} catch (InterruptedException e) {
e.printStackTrace();
}
}
if (num >= maxnum) {
break;
}
num++;
System.out.println(Thread.currentThread().getName() + ": " + num);
LOCK.notifyAll();
}
}
}
public static void main(String[] args) {
Wait_Notify_100 wait_notify_100 = new Wait_Notify_100();
new Thread(() -> {
wait_notify_100.printABC(0);
}, "thread1").start();
new Thread(() -> {
wait_notify_100.printABC(1);
}, "thread2").start();
new Thread(() -> {
wait_notify_100.printABC(2);
}, "thread3").start();
}
}
join方式
一个线程内调用另一个线程的join()方法可以让另一个线程插队执行，比如Main方法里调用了A.join()，那么此时cpu会去执行A线程中的任务，执行完后再看Main是否能抢到运行权。所以对于ABC，我们可以对B说让A插队，对C说让B插队
class Join_ABC {
static class printABC implements Runnable {
private Thread beforeThread;
public printABC(Thread beforeThread) {
this.beforeThread = beforeThread;
}
@Override
public void run() {
if (beforeThread != null) {
try {
beforeThread.join();
} catch (InterruptedException e) {
e.printStackTrace();
}
}
System.out.print(Thread.currentThread().getName());
}
}
public static void main(String[] args) throws InterruptedException {
for (int i = 0; i < 10; i++) {
Thread t1 = new Thread(new printABC(null), "A");
Thread t2 = new Thread(new printABC(t1), "B");
Thread t3 = new Thread(new printABC(t2), "C");
t1.start();
t2.start();
t3.start();
Thread.sleep(100);
}
}
}
ReentrantLock
同理，synchronized和reentrantlock都是我们常用的加锁方式，不过后者可以中断，可以实现公平锁，可以使用condition…但是需要我们手动释放锁。jdk8后二者性能差不多，毕竟synchronized有锁升级的过程嘛。
class ReentrantLock_ABC {
private int num;
private Lock lock = new ReentrantLock();
private void printABC(int targetNum) {
for (int i = 0; i < 100; ) {
lock.lock();
if (num % 3 == targetNum) {
num++;
i++;
System.out.print(Thread.currentThread().getName());
}
lock.unlock();
}
}
public static void main(String[] args) {
Lock_ABC lockABC = new Lock_ABC();
new Thread(() -> {
lockABC.printABC(0);
}, "A").start();
new Thread(() -> {
lockABC.printABC(1);
}, "B").start();
new Thread(() -> {
lockABC.printABC(2);
}, "C").start();
}
}
ReentrantLock+Condition
以上方式如果线程抢到锁后发现自己无法执行任务，那么就释放，然后别的线程再抢占再看是不是自己的…这种方式比较耗时，如果我们能实现精准唤醒锁呢，即A完成任务后唤醒它的下一个即B，这就用到我们的Condition啦
class ReentrantLock_Condition_ABC {
private int num;
private static Lock lock = new ReentrantLock();
private static Condition c1 = lock.newCondition();
private static Condition c2 = lock.newCondition();
private static Condition c3 = lock.newCondition();
private void printABC(int targetNum, Condition currentThread, Condition nextThread) {
for (int i = 0; i < 100; ) {
lock.lock();
try {
while (num % 3 != targetNum) {
currentThread.await(); //阻塞当前线程
}
num++;
i++;
System.out.print(Thread.currentThread().getName());
nextThread.signal(); //唤醒下一个线程
} catch (Exception e) {
e.printStackTrace();
} finally {
lock.unlock();
}
}
}
public static void main(String[] args) {
ReentrantLock_Condition_ABC reentrantLockConditionAbc = new ReentrantLock_Condition_ABC();
new Thread(() -> {
reentrantLockConditionAbc.printABC(0, c1, c2);
}, "A").start();
new Thread(() -> {
reentrantLockConditionAbc.printABC(1, c2, c3);
}, "B").start();
new Thread(() -> {
reentrantLockConditionAbc.printABC(2, c3, c1);
}, "C").start();
}
}
Semaphore
小伙伴们有没有想到过，在生产者消费者模型中我们有哪几种实现方式呢？wait\notify，ReentrantLock，Semaphone，阻塞队列，管道输入输出流。
对的就是Semaphone。
Semaphore有acquire方法和release方法。 当调用acquire方法时线程就会被阻塞，直到获得许可证为止。 当调用release方法时将向Semaphore中添加一个许可证。如果没有获取许可证的线程， Semaphore只是记录许可证的可用数量。
使用Semaphore也可以实现精准唤醒。
class SemaphoreABC {
private static Semaphore s1 = new Semaphore(1); //因为先执行线程A，所以这里设s1的计数器为1
private static Semaphore s2 = new Semaphore(0);
private static Semaphore s3 = new Semaphore(0);
private void printABC(Semaphore currentThread, Semaphore nextThread) {
for (int i = 0; i < 10; i++) {
try {
currentThread.acquire(); //阻塞当前线程，即信号量的计数器减1为0
System.out.print(Thread.currentThread().getName());
nextThread.release(); //唤醒下一个线程，即信号量的计数器加1
} catch (InterruptedException e) {
e.printStackTrace();
}
}
}
public static void main(String[] args) throws InterruptedException {
SemaphoreABC printer = new SemaphoreABC();
new Thread(() -> {
printer.printABC(s1, s2);
}, "A").start();
Thread.sleep(100);
new Thread(() -> {
printer.printABC(s2, s3);
}, "B").start();
Thread.sleep(100);
new Thread(() -> {
printer.printABC(s3, s1);
}, "C").start();
}
}
加载全部内容