Callable和Runnable有什么區(qū)別？

首先，Runnable是出自jdk1.0，Callable出自jdk1.5，那么，后出的類肯定對(duì)于前者有增強(qiáng)。
再看Runnable的run方法與Callable的call方法進(jìn)行對(duì)比

// Runnable
public abstract void run();

// Callable
V call() throws Exception;

可以看出，run方法沒有返回值，call方法有返回值并能夠拋出異常。

在對(duì)Callable原理進(jìn)行解析之前，首先先看一下Callable的基本用法

方式一：通過FutureTask，交給線程池執(zhí)行，阻塞等待結(jié)果返回

    public static void main(String[] args) throws Exception {
        FutureTask<String> futureTask = new FutureTask<String>(new Callable<String>() {
            public String call() throws InterruptedException {
                Thread.sleep(3000L);
                System.out.println("當(dāng)前線程:" + Thread.currentThread().getName());
                return "hello world";
            }
        });
        ExecutorService executorService = Executors.newFixedThreadPool(1);
        System.out.println("開始時(shí)間戳為:" + System.currentTimeMillis());
        executorService.submit(futureTask);
        String result = futureTask.get();
        System.out.println("結(jié)束時(shí)間戳為:" + System.currentTimeMillis() + "，result = " + result);
    }
結(jié)果打印：
開始時(shí)間戳為:1639275517192
當(dāng)前線程:pool-1-thread-1
結(jié)束時(shí)間戳為:1639275520202，result = hello world

可以看出，callable的邏輯是在異步線程中處理的，并且，主線程通過阻塞式接受異步線程返回的內(nèi)容。
那么，如果直接調(diào)用futureTask的run方法會(huì)怎樣？

    public static void main(String[] args) throws Exception {
        FutureTask<String> futureTask = new FutureTask<String>(new Callable<String>() {
            public String call() throws InterruptedException {
                Thread.sleep(3000L);
                System.out.println("當(dāng)前線程:" + Thread.currentThread().getName());
                return "hello world";
            }
        });
        System.out.println("開始時(shí)間戳為:" + System.currentTimeMillis());
        futureTask.run();
        String result = futureTask.get();
        System.out.println("結(jié)束時(shí)間戳為:" + System.currentTimeMillis() + "，result = " + result);
    }
結(jié)果打?。?開始時(shí)間戳為:1639275759794
當(dāng)前線程:main
結(jié)束時(shí)間戳為:1639275762796，result = hello world

結(jié)果和直接調(diào)用Runnable的方法是一樣的，最終是由主線程完成執(zhí)行

方式二：通過線程池執(zhí)行，返回Future對(duì)象

    public static void main(String[] args) throws Exception {
        ExecutorService executorService = Executors.newFixedThreadPool(1);
        System.out.println("開始時(shí)間戳為:" + System.currentTimeMillis());
        Future<String> future = executorService.submit(new Callable<String>() {
            public String call() throws InterruptedException {
                Thread.sleep(3000L);
                System.out.println("當(dāng)前線程:" + Thread.currentThread().getName());
                return "hello world";
            }
        });
        String result = future.get();
        System.out.println("結(jié)束時(shí)間戳為:" + System.currentTimeMillis() + "，result = " + result);
    }
結(jié)果打印：
開始時(shí)間戳為:1639276055569
當(dāng)前線程:pool-1-thread-1
結(jié)束時(shí)間戳為:1639276058583，result = hello world

實(shí)現(xiàn)效果與方式一相同，這里返回的Future，實(shí)際上也是FutureTask

原理分析

首先分析下線程池的submit方法

    // FutureTask傳入方式調(diào)用的submit方法
    public Future<?> submit(Runnable task) {
        if (task == null) throw new NullPointerException();
        RunnableFuture<Void> ftask = newTaskFor(task, null);
        execute(ftask);
        return ftask;
    }

    // Callable傳入方式調(diào)用的submit方法
    public <T> Future<T> submit(Callable<T> task) {
        if (task == null) throw new NullPointerException();
        RunnableFuture<T> ftask = newTaskFor(task);
        execute(ftask);
        return ftask;
    }

1.最終都會(huì)調(diào)用newTaskFor方法生成一個(gè)RunnableFuture方法
2.調(diào)用execute方法
3.返回RunnableFuture對(duì)象
線程池相關(guān)的原理不在本文章中具體展開，有興趣的小伙伴可以自行研究或者在之后的文章中推出詳解

再分析下future.get方法如何阻塞等待異步線程執(zhí)行結(jié)果

    public V get() throws InterruptedException, ExecutionException {
        int s = state;
        if (s <= COMPLETING)
            // 如果狀態(tài)還在進(jìn)行中，或者剛創(chuàng)建，就阻塞等待
            s = awaitDone(false, 0L);
        // 調(diào)用Report，返回結(jié)果或者拋出異常
        return report(s);
    }

    private V report(int s) throws ExecutionException {
        Object x = outcome;
        if (s == NORMAL)
            // 狀態(tài)正常，返回結(jié)果
            return (V)x;
        if (s >= CANCELLED)
            // 狀態(tài)取消，拋出取消異常
            throw new CancellationException();
        // 拋出程序執(zhí)行異常
        throw new ExecutionException((Throwable)x);
    }

FutureTask內(nèi)部維護(hù)了任務(wù)進(jìn)行的狀態(tài)，當(dāng)異步任務(wù)完成時(shí)，會(huì)將狀態(tài)碼設(shè)置為已完成，如果發(fā)生異常，會(huì)將狀態(tài)碼設(shè)置成對(duì)應(yīng)的異常狀態(tài)碼。

如何自己實(shí)現(xiàn)一個(gè)類似的能夠有返回值的Runnable

    public static class MyRunnable implements Runnable {

        private String result;

        private AtomicBoolean finished = new AtomicBoolean();

        public void run() {
            try {
                Thread.sleep(3000L);
                System.out.println("當(dāng)前線程:" + Thread.currentThread().getName());
                this.result = "hello world";
                finished.compareAndSet(false, true);
            } catch (InterruptedException e) {
                e.printStackTrace();
            }
        }

        public String get() {
            while (true) {
                if (finished.get()) {
                    return result;
                }
            }
        }

    }

    public static void main(String[] args) throws Exception {
        System.out.println("開始時(shí)間戳為:" + System.currentTimeMillis());
        MyRunnable myRunnable = new MyRunnable();
        new Thread(myRunnable).start();
        String result = myRunnable.get();
        System.out.println("結(jié)束時(shí)間戳為:" + System.currentTimeMillis() + "，result = " + result);
    }
結(jié)果打?。?開始時(shí)間戳為:1639280212975
當(dāng)前線程:異步線程
結(jié)束時(shí)間戳為:1639280215984，result = hello world

本質(zhì)上就是調(diào)用線程持有了異步線程的對(duì)象引用，jdk給我們封裝好了對(duì)應(yīng)的邏輯，無須額外處理。