JUC组件扩展（一）：FutureTask理解

一、概述

FutureTask包装器是一种非常便利的机制，同时实现了Future和Runnable接口。

类图如下：

FutureTask是一种可以取消的异步的计算任务。它的计算是通过Callable实现的，它等价于可以携带结果的Runnable，并且有三个状态：等待、运行和完成。完成包括所有计算以任意的方式结束，包括正常结束、取消和异常。

Future有个get方法而获取结果只有在计算完成时获取，否则会一直阻塞直到任务转入完成状态，然后会返回结果或者抛出异常。

FutureTask有下面几个重要的方法：

1.get()

阻塞一直等待执行完成拿到结果

2.get(int timeout, TimeUnit timeUnit)

阻塞一直等待执行完成拿到结果，如果在超时时间内，没有拿到抛出异常

3.isCancelled()

是否被取消

4.isDone()

是否已经完成

5.cancel(boolean mayInterruptIfRunning)

试图取消正在执行的任务

二、FutureTask的状态转换过程

* NEW -> COMPLETING -> NORMAL
 * NEW -> COMPLETING -> EXCEPTIONAL
 * NEW -> CANCELLED
 * NEW -> INTERRUPTING -> INTERRUPTED

三、FutureTask的执行过程

创建一个futureTask对象task
提交task到调度器executor等待调度或者在另外一个线程中执行task

等待调度中...

如果此时currentThread调取执行结果task.get(),会有几种情况
if task 还没有被executor调度或正在执行中
    阻塞当前线程，并加入到一个阻塞链表中waitNode
else if task被其它Thread取消，并取消成功 或task处于中断状态
    throw exception
else if task执行完毕，返回执行结果，或执行存在异常，返回异常信息

如果此时有另外一个线程调用task.get()

执行过程同上

四、应用场景

1. Future用于异步获取执行结果或者取消任务。

2. 在高并发场景下确保任务只执行一次。

五、源码分析

1.核心状态

  /**
  * The run state of this task, initially NEW.  The run state
  * transitions to a terminal state only in methods set,
  * setException, and cancel.  During completion, state may take on
  * transient values of COMPLETING (while outcome is being set) or
  * INTERRUPTING (only while interrupting the runner to satisfy a
  * cancel(true)). Transitions from these intermediate to final
  * states use cheaper ordered/lazy writes because values are unique
  * and cannot be further modified.
  *
  * Possible state transitions:
  * NEW -> COMPLETING -> NORMAL
  * NEW -> COMPLETING -> EXCEPTIONAL
  * NEW -> CANCELLED
  * NEW -> INTERRUPTING -> INTERRUPTED
  */
 private volatile int state;
 private static final int NEW          = 0;
 private static final int COMPLETING   = 1;
 private static final int NORMAL       = 2;
 private static final int EXCEPTIONAL  = 3;
 private static final int CANCELLED    = 4;
 private static final int INTERRUPTING = 5;
 private static final int INTERRUPTED  = 6;

2.构造函数

 public FutureTask(Callable<V> callable) {
     if (callable == null)
         throw new NullPointerException();
     this.callable = callable;
     this.state = NEW;       // ensure visibility of callable
 }

 public FutureTask(Runnable runnable, V result) {
     this.callable = Executors.callable(runnable, result);
     this.state = NEW;       // ensure visibility of callable
 }

3.获取执行结果

 public V get() throws InterruptedException, ExecutionException {
     int s = state;
     if (s <= COMPLETING)
         s = awaitDone(false, 0L);
     return report(s);
 }

 public V get(long timeout, TimeUnit unit)
     throws InterruptedException, ExecutionException, TimeoutException {
     if (unit == null)
         throw new NullPointerException();
     int s = state;
     if (s <= COMPLETING &&
         (s = awaitDone(true, unit.toNanos(timeout))) <= COMPLETING)
         throw new TimeoutException();
     return report(s);
 }

4.执行方法

 public void run() {
     if (state != NEW ||
         !UNSAFE.compareAndSwapObject(this, runnerOffset,
                                      null, Thread.currentThread()))
         return;
     try {
         Callable<V> c = callable;
         if (c != null && state == NEW) {
             V result;
             boolean ran;
             try {
                 result = c.call();
                 ran = true;
             } catch (Throwable ex) {
                 result = null;
                 ran = false;
                 setException(ex);
             }
             if (ran)
                 set(result);
         }
     } finally {
         // runner must be non-null until state is settled to
         // prevent concurrent calls to run()
         runner = null;
         // state must be re-read after nulling runner to prevent
         // leaked interrupts
         int s = state;
         if (s >= INTERRUPTING)
             handlePossibleCancellationInterrupt(s);
     }
 }

5.设置状态

 protected void set(V v) {
     if (UNSAFE.compareAndSwapInt(this, stateOffset, NEW, COMPLETING)) {
         outcome = v;
         UNSAFE.putOrderedInt(this, stateOffset, NORMAL); // final state
         finishCompletion();
     }
 }

 protected void setException(Throwable t) {
     if (UNSAFE.compareAndSwapInt(this, stateOffset, NEW, COMPLETING)) {
         outcome = t;
         UNSAFE.putOrderedInt(this, stateOffset, EXCEPTIONAL); // final state
         finishCompletion();
     }
 }

六、代码示例

1. FutureTask执行多任务计算的使用场景

利用FutureTask和ExecutorService，可以用多线程的方式提交计算任务，主线程继续执行其他任务，当主线程需要子线程的计算结果时，在异步获取子线程的执行结果。

 public class FutureTaskForMultiCompute {
     public static void main(String[] args) throws InterruptedException, ExecutionException {
         // 创建任务集合
         List<FutureTask<Integer>> taskList = new ArrayList<FutureTask<Integer>>();
         // 创建线程池
         ExecutorService exec = Executors.newFixedThreadPool(5);
         for (int i = 0; i < 10; i++) {
             FutureTask<Integer> ft = new FutureTask<>(new ComputeTask(i, ""+i));
             taskList.add(ft);
             // 提交给线程池执行任务，也可以通过exec.invokeAll(taskList)一次性提交所有任务;
             exec.submit(ft);
         }

         System.out.println("所有计算任务提交完毕, 主线程接着干其他事情！");  

         // 开始统计各计算线程计算结果
         Integer totalResult = 0;
         for (FutureTask<Integer> ft : taskList) {
             totalResult += ft.get();
         }

         // 关闭线程池
         exec.shutdown();
         System.out.println("多任务计算后的总结果是:" + totalResult);
     }

 }
 class ComputeTask implements Callable<Integer> {

     private Integer result = 0;
     private String taskName = "";  

     public String getTaskName(){
         return this.taskName;
     }

     public ComputeTask(Integer iniResult, String taskName){
         result = iniResult;
         this.taskName = taskName;
         System.out.println("生成子线程计算任务: "+taskName);
     }  

     @Override
     public Integer call() throws Exception {
         for (int i = 0; i < 100; i++) {
             result =+ i;
         }
         // 休眠5秒钟，观察主线程行为，预期的结果是主线程会继续执行，到要取得FutureTask的结果是等待直至完成。
         Thread.sleep(5000);
         System.out.println("子线程计算任务: "+taskName+" 执行完成!");
         return result;
     }

 }

2. FutureTask在高并发环境下确保任务只执行一次

在很多高并发的环境下，往往我们只需要某些任务只执行一次。这种使用情景FutureTask的特性恰能胜任。举一个例子，假设有一个带key的连接池，当key存在时，即直接返回key对应的对象；当key不存在时，则创建连接。对于这样的应用场景，通常采用的方法为使用一个Map对象来存储key和连接池对应的对应关系，典型的代码如下面所示：

 public class FutureTaskTest {
     private Map<String, Connection> connectionPool = new HashMap<String, Connection>();
     private ReentrantLock lock = new ReentrantLock();  

     public Connection getConnection(String key){
         try {
             lock.lock();
             if(connectionPool.containsKey(key)){
                 return connectionPool.get(key);
             }else{
                 //创建 Connection
                 Connection conn = createConnection();
                 connectionPool.put(key, conn);
                 return conn;
             }
         } finally{
             lock.unlock();
         }

     }

     //创建Connection
     private Connection createConnection(){
         return null;
     }  

 }

在上面的例子中，我们通过加锁确保高并发环境下的线程安全，也确保了connection只创建一次，然而确牺牲了性能。改用ConcurrentHash的情况下，几乎可以避免加锁的操作，性能大大提高，但是在高并发的情况下有可能出现Connection被创建多次的现象。这时最需要解决的问题就是当key不存在时，创建Connection的动作能放在connectionPool之后执行，这正是FutureTask发挥作用的时机，基于ConcurrentHashMap和FutureTask的改造代码如下：

 public class FutureTaskTest {
     private ConcurrentHashMap<String,FutureTask<Connection>>connectionPool = new ConcurrentHashMap<String, FutureTask<Connection>>();  

     public Connection getConnection(String key) throws Exception{
         FutureTask<Connection> connectionTask = connectionPool.get(key);
         if(connectionTask!=null){
             return connectionTask.get();
         }
         else{
             Callable<Connection> callable = new Callable<Connection>(){
                 @Override
                 public Connection call() throws Exception {
                     // TODO Auto-generated method stub
                     return createConnection();
                 }
             };
             FutureTask<Connection> newTask = new FutureTask<Connection>(callable);
             connectionTask = connectionPool.putIfAbsent(key, newTask);
             if(connectionTask==null){
                 connectionTask = newTask;
                 connectionTask.run();
             }
             return connectionTask.get();
         }
     }  

     //创建Connection
     private Connection createConnection(){
         return null;
     }  

 }

经过这样的改造，可以避免由于并发带来的多次创建连接及锁的出现。