java并发计算的几种基本使用示例

通常我们使用程序处理计算，常见的有单线程，JUC下单的多线程,还有优秀的ForkJoin框架（与递归形式上相似却又优于递归），直接上代码吧

package com.fork;

import java.util.concurrent.ExecutionException;
import java.util.concurrent.ForkJoinPool;
import java.util.concurrent.ForkJoinTask;
import java.util.concurrent.RecursiveTask;
10 
public class ForkDemo extends RecursiveTask<Long>{
    private long end;
    private final long THRELOD = 500000000L;
    private  long start;
    public ForkDemo(long start ,long end) {
        this.end = end;
        this.start = start;
    }

    @Override
    protected  Long compute() {
        long v = 0;
        if( end - start <= THRELOD) {
            for (long i = start; i <= end; i++) {
                v += i;
            }
            return v;
        }else {
            long middle = (end + start)/2;
            ForkDemo demo1 = new ForkDemo(start, middle);
            ForkDemo demo2 = new ForkDemo(middle+1, end);
            demo1.fork();
             demo2.fork();
            return demo1.join() + demo2.join();
            
        }        
    }
}

 这是ForkJoin的基本示例使用，以下是三种计算的比较；

package com.fork;

import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.concurrent.Callable;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.ForkJoinPool;
import java.util.concurrent.ForkJoinTask;
import java.util.concurrent.Future;
import java.util.concurrent.locks.Lock;
import java.util.concurrent.locks.ReentrantLock;




public class LockDemo {
//	 单线程方式
	public static long execSingle(long n) {
		long sum = 0;
		for (long i = 1; i <= n; i++) {
			sum += i;
		}
		System.out.println("the sum = " + sum);
		return sum;
	}
//	 多线程方式
	 public static long execMult(long n) throws ExecutionException {
		 long value = 0L;
	     ExecutorService exect = Executors.newFixedThreadPool(3);
	     ArrayList<Callable<Long>> list =new ArrayList<Callable<Long>>();
	     long total;
	     list.add(()->{
	        long sum = 0L;
	    	for(Long i=1L;i<=n;i+=3) {
	    		sum += i;
	    	}
	    	return sum;
	    });
	     list.add(()->{
		        long sum = 0L;
		    	for(Long i=2L;i<=n;i+=3) {
		    		sum += i;
		    	}
		    	return sum;
		    });
	     list.add(()->{
		        long sum = 0L;
		    	for(Long i=3L;i<=n;i+=3) {
		    		sum += i;
		    	}
		    	return sum;
		    });
	   List<Future<Long>> end =null;
	     try {
			end = exect.invokeAll(list);
			for(Future<Long> t:end) {
				value += t.get();
			}
		} catch (InterruptedException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}
	     exect.shutdownNow();
	     System.out.println("this end is " + value);
	     return value;  
	 }
	 

	 public static void main(String[] args) throws ExecutionException {
		 Long start1 =  System.currentTimeMillis();
		 execSingle(3800000000L);
		 Long end1 =  System.currentTimeMillis();
		 System.out.println("单线程下：end1 - start1 = " +(end1-start1) );
		 Long start2 =  System.currentTimeMillis();
		 execMult(3800000000L);
		 Long end2 =  System.currentTimeMillis();
		 System.out.println("多线程下：end2 - start2 = " +(end2-start2) );
		 Long start3 =  System.currentTimeMillis();
		 ForkJoinPool pool = new ForkJoinPool();
		 ForkDemo demo = new ForkDemo(0L, 3800000000L);
		 ForkJoinTask<Long> task = pool.submit(demo);
		 Long end3 =  System.currentTimeMillis();
		 System.out.println("forkjoin框架下：end3 - start3 = " +(end3-start3) );
		 try {
			System.out.println(task.get());
		} catch (InterruptedException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}
		 
	}
}

　　最后控制台输出：（ForkJoin 优秀啊.....此处向大神Doug Lea模拜......）

注意ForkJoin需要使用合理的细分程度，这里多线程计算时间夸张了点.....具体情境还是要具体分析，有些情况多线程不一定比多线程快（有线程切换的代价），多线程需要互不依赖，如爬取图片，以此我也不知道....多线程花了这么多时间，哪位大佬给分下.....