java 线程池工具类,线程池工具类测试可以用

import java.util.concurrent.Executors;

import java.util.concurrent.LinkedBlockingDeque;

import java.util.concurrent.ThreadPoolExecutor;

import java.util.concurrent.TimeUnit;

public class ThreadPoolManager {

public static ThreadPool instance;

// 获取单例的线程池对象

public static ThreadPool getInstance() {

if (instance == null) {

synchronized (ThreadPoolManager.class) {

if (instance == null) {

int cpuNum = Runtime.getRuntime().availableProcessors();// 获取处理器数量

int threadNum = cpuNum * 2 + 1;// 根据cpu数量,计算出合理的线程并发数

instance = new ThreadPool(threadNum - 1, threadNum, Integer.MAX_VALUE);//默认是双核的cpu 每个核心走一个线程 一个等待线程

}

}

}

return instance;

}

public static class ThreadPool {

private ThreadPoolExecutor mExecutor;

private int corePoolSize;

private int maximumPoolSize;

private long keepAliveTime;

private ThreadPool(int corePoolSize, int maximumPoolSize, long keepAliveTime) {

this.corePoolSize = corePoolSize;

this.maximumPoolSize = maximumPoolSize;

this.keepAliveTime = keepAliveTime;

}

public void execute(Runnable runnable) {

if (runnable == null) {

return;

}

if (mExecutor == null) {

mExecutor = new ThreadPoolExecutor(corePoolSize,// 核心线程数

maximumPoolSize, // 最大线程数

keepAliveTime, // 闲置线程存活时间

TimeUnit.MILLISECONDS,// 时间单位

new LinkedBlockingDeque(Integer.MAX_VALUE),// 线程队列

Executors.defaultThreadFactory(),// 线程工厂

new ThreadPoolExecutor.AbortPolicy() {// 队列已满,而且当前线程数已经超过最大线程数时的异常处理策略

@Override

public void rejectedExecution(Runnable r, ThreadPoolExecutor e) {

super.rejectedExecution(r, e);

}

}

);

}

mExecutor.execute(runnable);

}

// 从线程队列中移除对象

public void cancel(Runnable runnable) {

if (mExecutor != null) {

mExecutor.getQueue().remove(runnable);

}

}

}

}