JAVA进阶-多线程(2)

堵塞队列:

1)BlockingQueue该接口提供了:

add()/remove() 假设当队列没有数据,从队列中取数据;或者队列中数据已满,

向队列中加入数据;则会抛出异常.

put()/take() 假设当队列没有数据,从队列中取数据;或者队列中数据已满,

向队列中加入数据;则会形成堵塞.

offer()/poll() 会给调用者返回特殊的值,开发人员能够通过这些值做对应的处理

同一时候还提供了超时版本号. 

2)接口实现

ArrayBlockingQueue>由数组实现的有界队列,默认情况下没有指定公平策略(也就是

一般的FIFO先进先出策略),假设不启动策略,会导致共享资源被贪婪的线程长时间占有,

而无法获取资源的线程可能死掉,这样的情况称为饿死;

LinkedBlockingQueue>将最大的容量变为可选,默认的容量为整型最大值,也就是不存在

生产者生产增加队列时产生堵塞的情况.该队列一般在要求较低的情况下使用.

PriorityBlockingQueue>无界队列,由线程对象的优先级决定获取cpu操作时间,同一时候,

开发着也能够提供自己的比較器,比方同样扩展同样优先级的线程.

DelayedQueue>是用类似栈维护的特殊的优先级队列.

1.检索前指定时间内保持驻留在队列中.

2.依照驻留时间排序,最长驻留时间位于底部.

3.仅仅同意检索过期后的对象,当队列中没有过期对象.poll返回null,peek

 则获取栈顶的对象.

SynchronousQueue>实现了每一个插入操作都必须等待相应的移除操作;队列始终为空,

当,发现队列有东西,就会有相应的消费着瞬间消费这些东西;

TransferQueue>该接口扩展了BlockingQueue.而且LinkedTransferQueue提供了

接口的详细实现;该接口扩展了BlockingQueue的put方法为transfer(),该方法

为超时的非堵塞调用.同一时候,该接口提供了获取等待消费者的数量检測.

---------------

/**
 *
 *
 * 	@author Lean  @date:2014-9-28
 */
public class StockExchange {

	public static void main(String[] args) {
		BlockingQueue<Integer> queue=new LinkedBlockingQueue<Integer>();

		Saller saller=new Saller(queue);
		Buyer buyer=new Buyer(queue);
		Thread[] sallerThreads=new Thread[20];
		Thread[] buyerThreads=new Thread[20];
		for (int i = 0; i <sallerThreads.length; i++) {
			sallerThreads[i]=new Thread(saller);
			sallerThreads[i].start();
			buyerThreads[i]=new Thread(buyer);
			buyerThreads[i].start();
		}
		try {
			Thread.sleep(20);
		} catch (InterruptedException e) {
		}
		System.out.println("all thread interrupt!");
		for (Thread thread : sallerThreads) {
			thread.interrupt();
		}
		for (Thread thread : buyerThreads) {
			thread.interrupt();
		}
	}

	static class Saller implements Runnable{

		private BlockingQueue<Integer> mQueue;
		private boolean shutDownRequest;

		public Saller(BlockingQueue<Integer> queue){
			mQueue=queue;
		}

		@Override
		public void run() {
			while (shutDownRequest==false) {
				int quantity=(int)(Math.random()*100);
				try {
					mQueue.put(quantity);
//					System.out.println("saller order by Thread:"+Thread.currentThread().getName()+"  quantity:"+quantity);
				} catch (InterruptedException e) {
					shutDownRequest=true;
				}
			}
		}

	}

	static class Buyer implements Runnable{

		private BlockingQueue<Integer> mQueue;
		private boolean shutDownRequest;

		public Buyer(BlockingQueue<Integer> queue){
			mQueue=queue;
		}

		@Override
		public void run() {
			while (shutDownRequest==false) {
				try {
					System.out.println("buyer order by Thread:"+Thread.currentThread().getName()+"  quantity:"+mQueue.take());
				} catch (InterruptedException e) {
					shutDownRequest=true;
				}
			}
		}

	}

}

---------------

---------------

/**
 *
 * 	@author Lean  @date:2014-9-28
 */
public class LuckyNumberGenerator {

	public static void main(String[] args) {
		TransferQueue<String> queue=new LinkedTransferQueue<String>();
		Thread producerThread=new Thread(new Producer(queue));
		producerThread.setDaemon(true);
		producerThread.start();

		for (int i = 0; i < 20; i++) {
			Thread comsumerThread=new Thread(new Comsumer(queue));
			comsumerThread.setDaemon(true);
			comsumerThread.start();
			try {
				Thread.sleep(2000);
			} catch (InterruptedException e) {
				e.printStackTrace();
			}
		}
		System.out.println(Thread.currentThread().getThreadGroup().activeCount());
	}

	static class Producer implements Runnable{

		private TransferQueue<String> mQueue;

		public Producer(TransferQueue<String> queue){
			this.mQueue=queue;
		}

		public String product(){
			return "your lucky number is: "+((int)(Math.random()*100));
		}

		@Override
		public void run() {
			while (true) {
				try {
					if (mQueue.hasWaitingConsumer()) {
							mQueue.put(product());
					}
				} catch (InterruptedException e) {
					e.printStackTrace();
				}
			}
		}

	}

	static class Comsumer implements Runnable{

		private TransferQueue<String> mQueue;

		public Comsumer(TransferQueue<String> queue){
			this.mQueue=queue;
		}

		@Override
		public void run() {
			try {
				System.out.println(mQueue.take());
			} catch (InterruptedException e) {
				e.printStackTrace();
			}
		}

	}

}

---------------

同步器:



1)信号量Semaphore>指定代理个数,在某一时间内,查看当前是否有代理处理事情,处理完事件,释放代理;

/**
 *
 *
 * @author Lean
 */
public class Bank {

	private static final int COUNT=100;
	private static final Semaphore semaphore=new Semaphore(2,true);

	public static void main(String[] args) {
		for (int i = 0; i < COUNT; i++) {
			final int count=i;
			new Thread(){
				@Override
				public void run() {
					try {
						if (semaphore.tryAcquire(10, TimeUnit.MILLISECONDS)) {
							try {
								Teller.getService(count);
							}finally{
								semaphore.release();
							}
						}
					} catch (InterruptedException e) {
						e.printStackTrace();
					}
				};
			}.start();
		}
	}

	static class Teller{
		public static void getService(int i){
			System.out.println("serving:"+i);
			try {
				Thread.sleep((long)(Math.random()*10));
			} catch (InterruptedException e) {
				e.printStackTrace();
			}
		}
	}

}

2)屏障CyclicBarrier>指多个线程到达某个点后停止运行(调用CyclicBarrier对象的

wawit()方法)当多个任务(到达构造參数的指定的个数)达到指定的位置后,运行CyclicBarrier构造參数的Runnable;

/**
 *		屏障(会合点)
 *		sample:计算平方和
 * 	@author Lean  @date:2014-9-29
 */
public class CalculateSum {

	public static final int COUNT=3;
	public static int[] tempArray=new int[COUNT];

	public static void main(String[] args) {
		CyclicBarrier barrier=new CyclicBarrier(COUNT,new Runnable() {

			@Override
			public void run() {
				int sum=0;
				for (int i = 0; i < COUNT; i++) {
					sum=sum+tempArray[i];
				}
				System.out.println("the result is:"+sum);
			}
		});
		for (int i = 0; i <COUNT; i++) {
			new Thread(new Square(i,barrier)).start();
		}
		System.out.println("caculate now...");
	}

	static class Square implements Runnable{

		private int initSize;
		private CyclicBarrier barrier;

		public Square(int initSize,CyclicBarrier barrier){
			this.initSize=initSize;
			this.barrier=barrier;
		}

		@Override
		public void run() {
			int result=initSize*initSize;
			tempArray[initSize]=result;
			try {
				barrier.await();
			} catch (InterruptedException e) {
				e.printStackTrace();
			} catch (BrokenBarrierException e) {
				e.printStackTrace();
			}
		}

	}

}

3)倒计数闭锁CountDownLatch>构造CountDownLatch的时候指定倒数个数,调用await()会使其后面的代码堵塞

调用countDown(),倒数-1,当倒数为0时,运行CountDownLatch对象await()后的代码.相比于CyclicBarrier,

CountDownLatch提供了手动控制屏蔽,比較灵活

/**
 *
 * 	@author Lean  @date:2014-9-29
 */
public class EnhancedStockExchange {

	public static void main(String[] args) {
		BlockingQueue<Integer> queue=new LinkedBlockingQueue<Integer>();
		CountDownLatch startLatch=new CountDownLatch(1);
		final CountDownLatch stopLatch=new CountDownLatch(200);
		Producer producer=new Producer(startLatch, stopLatch, queue);
		Saller saller=new Saller(startLatch, stopLatch, queue);
		Thread[] sellerThreads=new Thread[100];
		for (int i = 0; i < sellerThreads.length; i++) {
			sellerThreads[i]=new Thread(saller);
			sellerThreads[i].start();
		}
		Thread[] producerThreads=new Thread[100];
		for (int i = 0; i < producerThreads.length; i++) {
			producerThreads[i]=new Thread(producer);
			producerThreads[i].start();
		}
		//倒数闭锁,当前倒数为1,运行例如以下函数,倒数0;
		startLatch.countDown();

		new Thread(new Runnable() {

			@Override
			public void run() {
				try {
					//运行await(),暂停直至倒数器为0
					stopLatch.await();
				} catch (InterruptedException e) {
					e.printStackTrace();
				}
				System.out.println("all thread countdown!");
			}
		}).start();

		try {
			Thread.sleep(20);
		} catch (InterruptedException e) {
			e.printStackTrace();
		}
		System.out.println("Terminating...");
		//运行interrupt(),运行while语句后的mStopLatch.countDown();倒数为1
		for (Thread thread : sellerThreads) {
			thread.interrupt();
		}
		for (Thread thread : producerThreads) {
			thread.interrupt();
		}
		//倒数为0,运行run()方法内await()后的代码;
		stopLatch.countDown();
	}

	static class Producer implements Runnable{

		public CountDownLatch mStartLatch;
		public CountDownLatch mStopLatch;
		private BlockingQueue<Integer> mQueue;
		private boolean shutDownRequest;

		public Producer(CountDownLatch startLatch,CountDownLatch stopLatch,BlockingQueue<Integer> queue){
			mStartLatch=startLatch;
			mStopLatch=stopLatch;
			mQueue=queue;
		}

		@Override
		public void run() {
			try {
				mStartLatch.await();
			} catch (InterruptedException e) {
				e.printStackTrace();
			}
			while (shutDownRequest==false) {
				try {
					mQueue.put((int)(Math.random()*(100)));
				} catch (InterruptedException e) {
					shutDownRequest=true;
				}
			}
			mStopLatch.countDown();
		}

	}

	static class Saller implements Runnable{

		public CountDownLatch mStartLatch;
		public CountDownLatch mStopLatch;
		private BlockingQueue<Integer> mQueue;
		private boolean shutDownRequest;

		public Saller(CountDownLatch startLatch,CountDownLatch stopLatch,BlockingQueue<Integer> queue){
			mStartLatch=startLatch;
			mStopLatch=stopLatch;
			mQueue=queue;
		}

		@Override
		public void run() {
			try {
				mStartLatch.await();
			} catch (InterruptedException e) {
				e.printStackTrace();
			}
			while (shutDownRequest==false) {
				try {
					System.out.println("saller comsume: "+mQueue.take());
				} catch (InterruptedException e) {
					shutDownRequest=true;
				}
			}
			mStopLatch.countDown();
		}

	}

}

4)移相器Phaser>实现屏障一样的功能,相比于屏障和倒计数闭锁,Phaser实例manager提供了可伸缩的等待数目.

在执行的过程中,动态添加拦截数可调用manager.register();当调用manager.arriveAndDeregister()时,当前全部

等待线程继续运行;在线程运行中,可调用manager.arriveAndAwaitAdvance();

等待其它线程;同一时候我们能够调用manager.getArrivedParties()查看等待线程数;

/**
 *
 * 	@author Lean  @date:2014-9-29
 */
public class HorseRace {

	private final int NUMBER_OF_HORSE=12;
	private static final int INIT_PARTIES=1;
	private static final Phaser manager=new Phaser(INIT_PARTIES);

	public static void main(String[] args) {
		//检查准备就绪的马匹数量
		Thread raceMonitor=new Thread(new RaceMonitor());
		raceMonitor.setDaemon(true);
		raceMonitor.start();

		new HorseRace().managerRace();

	}

	private void managerRace() {
		ArrayList<Horse> horses=new ArrayList<HorseRace.Horse>();
		for (int i = 0; i < NUMBER_OF_HORSE; i++) {
			horses.add(new Horse());
		}
		runRace(horses);
	}

	private void runRace(Iterable<Horse> horses) {
		for (final Horse horse : horses) {
			manager.register();
			new Thread(){
				@Override
				public void run() {
					try {
						Thread.sleep((new Random()).nextInt(1000));
					} catch (InterruptedException e) {
						e.printStackTrace();
					}
					manager.arriveAndAwaitAdvance();
					horse.run();
				};
			}.start();
		}
		try {
			Thread.sleep(1000);
		} catch (InterruptedException e) {
			e.printStackTrace();
		}
		manager.arriveAndDeregister();
	}

	private static class RaceMonitor implements Runnable{

		@Override
		public void run() {
			while (true) {
//				System.out.println("number of horses to run:"+HorseRace.manager.getArrivedParties());
				try {
					Thread.sleep(1);
				} catch (InterruptedException e) {
					e.printStackTrace();
				}
			}
		}

	}

	private static class Horse implements Runnable{

		private static final AtomicInteger idSource=new AtomicInteger();
		private final int id=idSource.incrementAndGet();

		@Override
		public void run() {
			System.out.println(toString()+" is running");
		}

		@Override
		public String toString() {
			return "Horse [id=" + id + "]";
		}

	}

}

5)交换器Exchanger<T>

类型T为两个线程交换的对象,在某些同样操作的批量编程中,当中一类线程

负责生产对象,还有一类编程负责消耗对象,对于线程间共享数据,前面介绍了锁

的定义,当我们使用JAVA提供的Exchanger<T>传输对象,不须要锁的概念.

buffers=ProductExchange.exchanger.exchange(buffers, 1000,TimeUnit.MILLISECONDS);

该对象的exchange方法參数传递了该线程其它线程的数据,并返回了其它线程返回的数据

/**
 *
 * 	@author Lean  @date:2014-9-29
 */
public class ProductExchange {

	public static Exchanger<ArrayList<Integer>> exchanger=new Exchanger<ArrayList<Integer>>();

	public static void main(String[] args) {
		Thread producerThread=new Thread(new Producer());
		Thread comsumeThread=new Thread(new Comsume());
		producerThread.start();
		comsumeThread.start();
		try {
			Thread.sleep(1000);
		} catch (InterruptedException e) {
			e.printStackTrace();
		}

		producerThread.interrupt();
		comsumeThread.interrupt();

	}

	private static class Producer implements Runnable{

		private static ArrayList<Integer> buffers=new ArrayList<Integer>();
		private boolean okToRun=true;

		@Override
		public void run() {
			while (okToRun) {
				try {
					if (buffers.isEmpty()) {
						for (int i = 0; i <10; i++) {
							buffers.add((int)(Math.random()*100));
						}
						Thread.sleep(200);
						for (int i : buffers) {
							System.out.print(i+" ,");
						}
						System.out.println("");
						buffers=ProductExchange.exchanger.exchange(buffers, 1000,TimeUnit.MILLISECONDS);
					}
				} catch (InterruptedException e) {
					okToRun=false;
				} catch (TimeoutException e) {
					System.out.println("produce time out!");
				}
			}
		}

	}

	private static class Comsume implements Runnable{

		private static ArrayList<Integer> buffers=new ArrayList<Integer>();
		private boolean okToRun=true;

		@Override
		public void run() {
			while (okToRun) {
				try {
					if (buffers.isEmpty()) {
						buffers=ProductExchange.exchanger.exchange(buffers);
						for (int i : buffers) {
							System.out.print(i+" ,");
						}
						System.out.println("");
						Thread.sleep(200);
						buffers.clear();
					}
				} catch (InterruptedException e) {
					okToRun=false;
				}
			}
		}

	}

}