Thread、Runnable和Callable

1. 关于Thread的Runnable和Callable接口

其实非常简单：其实他们的区别就是Callable有返回值并且可以抛出异常。

/**
 * Represents a command that can be executed. Often used to run code in a
 * different {@link Thread}.
 */
public interface Runnable {


    /**
     * Starts executing the active part of the class' code. This method is
     * called when a thread is started that has been created with a class which
     * implements {@code Runnable}.
     */
    public void run();
}

/**
 * A task that returns a result and may throw an exception.
 * Implementors define a single method with no arguments called
 * <tt>call</tt>.
 *
 * <p>The <tt>Callable</tt> interface is similar to {@link
 * java.lang.Runnable}, in that both are designed for classes whose
 * instances are potentially executed by another thread.  A
 * <tt>Runnable</tt>, however, does not return a result and cannot
 * throw a checked exception.
 *
 * <p> The {@link Executors} class contains utility methods to
 * convert from other common forms to <tt>Callable</tt> classes.
 *
 * @see Executor
 * @since 1.5
 * @author Doug Lea
 * @param <V> the result type of method <tt>call</tt>
 */
public interface Callable<V> {
    /**
     * Computes a result, or throws an exception if unable to do so.
     *
     * @return computed result
     * @throws Exception if unable to compute a result
     */
    V call() throws Exception;
}

2. Thread类

　1、Runnable接口源码：

1 public interface Runnable {
2     public void run();
3 }

　　2、Thread类与Runnable接口的继承关系

1 public class Thread implements Runnable{
2 
3 }

　　Runnable接口仅有一个run()方法，Thread类实现了Runnable接口。

　　3、构造函数

1 public Thread() {
2     init(null, null, "Thread-" + nextThreadNum(), 0);
3 }

1 public Thread(Runnable target) {
2     init(null, target, "Thread-" + nextThreadNum(), 0);
3 }

1 public Thread(ThreadGroup group, Runnable target) {
2     init(group, target, "Thread-" + nextThreadNum(), 0);
3 }

1 public Thread(String name) {
2     init(null, null, name, 0);
3 }
　　　　　　　　　　　　　　　　　　还有其它的构造方法，此处省略。。。

　　这里的第三个参数是设置线程的名称，从下面的代码中可以看出，生成名称的规则是：”Thread-”加上创建的线程的个数（第几个）。

继续查看init方法：

 1 /**
 2      * Initializes a Thread.
 3      *
 4      * @param g the Thread group
 5      * @param target the object whose run() method gets called
 6      * @param name the name of the new Thread
 7      * @param stackSize the desired stack size for the new thread, or
 8      *        zero to indicate that this parameter is to be ignored.
 9      */
　　　　//ThreadGroup：线程组表示一个线程的集合。此外，线程组也可以包含其他线程组。线程组构成一棵树，在树中，除了初始线程组外，每个线程组都有一个父线程组。 
10     private void init(ThreadGroup g, Runnable target, String name,
11                       long stackSize) {
12     Thread parent = currentThread();
13     SecurityManager security = System.getSecurityManager();
14     if (g == null) {
15         /* Determine if it's an applet or not */
16         
17         /* If there is a security manager, ask the security manager
18            what to do. */
19         if (security != null) {
20         g = security.getThreadGroup();
21         }
22 
23         /* If the security doesn't have a strong opinion of the matter
24            use the parent thread group. */
25         if (g == null) {
26         g = parent.getThreadGroup();
27         }
28     }
29 
30     /* checkAccess regardless of whether or not threadgroup is
31            explicitly passed in. */
32     g.checkAccess();
33 
34     /*
35      * Do we have the required permissions?
36      */
37     if (security != null) {
38         if (isCCLOverridden(getClass())) {
39             security.checkPermission(SUBCLASS_IMPLEMENTATION_PERMISSION);
40         }
41     }
42 
43 
44         g.addUnstarted();
45 
46     this.group = g;

　　　　//每个线程都有一个优先级，高优先级线程的执行优先于低优先级线程。每个线程都可以或不可以标记为一个守护程序。当某个线程中运行的代码创建一个新 Thread 对象时，该新线程的初始优先级被设定为创建线程的优先级，并且当且仅当创建线程是守护线程时，新线程才是守护程序。 

47     this.daemon = parent.isDaemon();
48     this.priority = parent.getPriority();
49     this.name = name.toCharArray();
50     if (security == null || isCCLOverridden(parent.getClass()))
51         this.contextClassLoader = parent.getContextClassLoader();
52     else
53         this.contextClassLoader = parent.contextClassLoader;
54     this.inheritedAccessControlContext = AccessController.getContext();
55     this.target = target;
56     setPriority(priority);
57         if (parent.inheritableThreadLocals != null)
58         this.inheritableThreadLocals =
59         ThreadLocal.createInheritedMap(parent.inheritableThreadLocals);
60         /* Stash the specified stack size in case the VM cares */
61         this.stackSize = stackSize;
62 
63         /* Set thread ID */
64         tid = nextThreadID();
65     }

　　初始化时设置了是否为守护线程，优先级，初始化名称。

　　4、Thread的start方法的实现：

 1 public synchronized void start() {
 2         /**
 3      * This method is not invoked for the main method thread or "system"
 4      * group threads created/set up by the VM. Any new functionality added 
 5      * to this method in the future may have to also be added to the VM.
 6      *
 7      * A zero status value corresponds to state "NEW".
 8          */
 9         if (threadStatus != 0)
10             throw new IllegalThreadStateException();
11         group.add(this);
12         start0();
13         if (stopBeforeStart) {
14         stop0(throwableFromStop);
15     }
16 }

　　这里主要的是start0方法；查看其实现：

 1 private native void start0();

　　这里使用了本地调用，通过C代码初始化线程需要的系统资源。可见，线程底层的实现是通过C代码去完成的。

4、Thread的run方法的实现

1 public void run() {
2     if (target != null) {
3         target.run();
4     }
5 }

　　这里的target实际上要保存的是一个Runnable接口的实现的引用：

1 private Runnable target;

　　所以使用继承Thread创建线程类时，需要重写run方法，因为默认的run方法什么也不干。

　　而当我们使用Runnable接口实现线程类时，为了启动线程，需要先把该线程类实例初始化一个Thread，实际上就执行了如下构造函数：

1 public Thread(Runnable target) {
2     init(null, target, "Thread-" + nextThreadNum(), 0);
3 }

　　即是把线程类的引用保存到target中。这样，当调用Thread的run方法时，target就不为空了，而是继续调用了target的run方法，所以我们需要实现Runnable的run方法。这样通过Thread的run方法就调用到了Runnable实现类中的run方法。

　　这也是Runnable接口实现的线程类需要这样启动的原因。

另外java中的Executor 类也是非常简单的接口，只有一个execute（）方法

   public interface Executor {


    /**
     * Executes the given command at some time in the future.  The command
     * may execute in a new thread, in a pooled thread, or in the calling
     * thread, at the discretion of the <tt>Executor</tt> implementation.
     *
     * @param command the runnable task
     * @throws RejectedExecutionException if this task cannot be
     * accepted for execution.
     * @throws NullPointerException if command is null
     */
    void execute(Runnable command);
}

     Excutor执行已提交的 Runnable 任务的对象。此接口提供一种将任务提交与每个任务将如何运行的机制（包括线程使用的细节、调度等）分离开来的方法。通常使用 Executor 而不是显式地创建线程。例如，可能会使用以下方法，而不是为一组任务中的每个任务调用 new Thread(new(RunnableTask())).start()：

 Executor executor = anExecutor;
 executor.execute(new RunnableTask1());
 executor.execute(new RunnableTask2());
 ...
 

不过， Executor  接口并没有严格地要求执行是异步的。在最简单的情况下，执行程序可以在调用方的线程中立即运行已提交的任务：

 class DirectExecutor implements Executor {
     public void execute(Runnable r) {
         r.run();
     }
 }

更常见的是，任务是在某个不是调用方线程的线程中执行的。以下执行程序将为每个任务生成一个新线程。

 class ThreadPerTaskExecutor implements Executor {
     public void execute(Runnable r) {
         new Thread(r).start();
     }
 }

许多  Executor  实现都对调度任务的方式和时间强加了某种限制。以下执行程序使任务提交与第二个执行程序保持连续，这说明了一个复合执行程序。

 class SerialExecutor implements Executor {
     final Queue<Runnable> tasks = new LinkedBlockingQueue<Runnable>();
     final Executor executor;
     Runnable active;

     SerialExecutor(Executor executor) {
         this.executor = executor;
     }

     public synchronized void execute(final Runnable r) {
         tasks.offer(new Runnable() {
             public void run() {
                 try {
                     r.run();
                 } finally {
                     scheduleNext();
                 }
             }
         });
         if (active == null) {
             scheduleNext();
         }
     }

     protected synchronized void scheduleNext() {
         if ((active = tasks.poll()) != null) {
             executor.execute(active);
         }
     }
 }