写这篇文章的目的，是看到群里有人在实现延迟的时候，用如下的第四种方法，个人感觉有点不妥，为了防止更多的人有这种想法，所以自己抽空深入分析，就分析的结果，写下此文，希望对部分人有启示作用。

1.实现延迟的几种方法？

答：

public void schedule(TimerTask task, long delay) {        if (delay < 0)            throw new IllegalArgumentException("Negative delay.");        sched(task, System.currentTimeMillis()+delay, 0);    }

2.android.os.Handler类：

public final boolean postDelayed(Runnable r, long delayMillis)    {        return sendMessageDelayed(getPostMessage(r), delayMillis);    }

3.android.app.AlarmManager类:

@SystemApi    @RequiresPermission(android.Manifest.permission.UPDATE_DEVICE_STATS)    public void set(@AlarmType int type, long triggerAtMillis, long windowMillis,            long intervalMillis, OnAlarmListener listener, Handler targetHandler,            WorkSource workSource) {        setImpl(type, triggerAtMillis, windowMillis, intervalMillis, 0, null, listener, null,                targetHandler, workSource, null);    }

4.Thread.sleep()然后在一定时间之后再执行想执行的代码：

new Thread(new Runnable(){    Thead.sleep(4*1000);    doTask();}).start()

2.他们的各自的实现原理？

答：

1.Timer的实现，是通过内部开启一个TimerThread：

private void mainLoop() {        while (true) {            try {                TimerTask task;                boolean taskFired;                synchronized(queue) {                    // Wait for queue to become non-empty                    while (queue.isEmpty() && newTasksMayBeScheduled)                        queue.wait();                    if (queue.isEmpty())                        break; // Queue is empty and will forever remain; die                    // Queue nonempty; look at first evt and do the right thing                    long currentTime, executionTime;                    task = queue.getMin();                    synchronized(task.lock) {                        if (task.state == TimerTask.CANCELLED) {                            queue.removeMin();                            continue;  // No action required, poll queue again                        }                        currentTime = System.currentTimeMillis();                        executionTime = task.nextExecutionTime;                        if (taskFired = (executionTime<=currentTime)) {                            if (task.period == 0) { // Non-repeating, remove                                queue.removeMin();                                task.state = TimerTask.EXECUTED;                            } else { // Repeating task, reschedule                                queue.rescheduleMin(                                  task.period<0 ? currentTime   - task.period                                                : executionTime + task.period);                            }                        }                    }                    if (!taskFired) // Task hasn't yet fired; wait                        queue.wait(executionTime - currentTime);                }                if (taskFired)  // Task fired; run it, holding no locks                    task.run();            } catch(InterruptedException e) {            }        }    }

是通过wait和延迟时间到达的时候，调用notify来唤起线程继续执行，这样来实现延迟的话，我们可以回开启一个新的线程，貌似为了个延迟没必要这样吧，定时，频繁执行的任务，再考虑这个吧。

2.Handler的postDelay是通过设置Message的when为delay的时间，我们知道当我们的应用开启的时候，会同步开启Looper.loop()方法循环的，不停的通过MeassgeQueue的next方法：

Message next() {        ......        int nextPollTimeoutMillis = 0;        for (;;) {            if (nextPollTimeoutMillis != 0) {                Binder.flushPendingCommands();            }            nativePollOnce(ptr, nextPollTimeoutMillis);            synchronized (this) {                // Try to retrieve the next message.  Return if found.                final long now = SystemClock.uptimeMillis();                Message prevMsg = null;                Message msg = mMessages;                if (msg != null && msg.target == null) {                    // Stalled by a barrier.  Find the next asynchronous message in the queue.                    do {                        prevMsg = msg;                        msg = msg.next;                    } while (msg != null && !msg.isAsynchronous());                }                if (msg != null) {                    if (now < msg.when) {                        // Next message is not ready.  Set a timeout to wake up when it is ready.                        nextPollTimeoutMillis = (int) Math.min(msg.when - now, Integer.MAX_VALUE);                    } else {                        // Got a message.                        mBlocked = false;                        if (prevMsg != null) {                            prevMsg.next = msg.next;                        } else {                            mMessages = msg.next;                        }                        msg.next = null;                        if (DEBUG) Log.v(TAG, "Returning message: " + msg);                        msg.markInUse();                        return msg;                    }                } else {                    // No more messages.                    nextPollTimeoutMillis = -1;                }        ......        }    }

当我们向MessageQueue插入一条延迟的Message的时候，Looper在执行loop方法，底层会调用epoll_wait(mEpollFd, eventItems, EPOLL_MAX_EVENTS, timeoutMillis);其中的timeoutMillis参数指定了在没有事件发生的时候epoll_wait调用阻塞的毫秒数(milliseconds)。这样我们在之前的时间内这个时候阻塞了是会释放cpu的资源，等到延迟的时间到了时候，再监控到事件发生。在这里可能有人会有疑问，一直阻塞，那我接下来的消息应该怎么执行呢？我们可以看到当我们插入消息的时候的方法：

boolean enqueueMessage(Message msg, long when) {        if (msg.target == null) {            throw new IllegalArgumentException("Message must have a target.");        }        if (msg.isInUse()) {            throw new IllegalStateException(msg + " This message is already in use.");        }        synchronized (this) {            if (mQuitting) {                IllegalStateException e = new IllegalStateException(                        msg.target + " sending message to a Handler on a dead thread");                Log.w(TAG, e.getMessage(), e);                msg.recycle();                return false;            }            msg.markInUse();            msg.when = when;            Message p = mMessages;            boolean needWake;            if (p == null || when == 0 || when < p.when) {                msg.next = p;                mMessages = msg;                needWake = mBlocked;            } else {                needWake = mBlocked && p.target == null && msg.isAsynchronous();                Message prev;                for (;;) {                    prev = p;                    p = p.next;                    if (p == null || when < p.when) {                        break;                    }                    if (needWake && p.isAsynchronous()) {                        needWake = false;                    }                }                msg.next = p; // invariant: p == prev.next                prev.next = msg;            }            mQuitting is false.            if (needWake) {                nativeWake(mPtr);            }        }        return true;    }

阻塞了有两种方式唤醒，一种是超时了，一种是被主动唤醒了，在上面我们可以看到当有消息进入的时候，我们会唤醒继续执行，所以我们的即时消息在延迟消息之后插入是没有关系的。然后在延迟时间到了的时候，我们也会被唤醒，执行对应的消息send，以达到延迟时间执行某个任务的目的。

3.AlarmManager的延迟的实现原理，是通过一个AlarmManager的set方法:

IAlarmManager mService.set(mPackageName, type, triggerAtMillis, windowMillis, intervalMillis, flags,                    operation, recipientWrapper, listenerTag, workSource, alarmClock);

这里是通过aidl与AlarmManagerService的所在进程进行通信，具体的实现是在AlarmManagerService类里面：

private final IBinder mService = new IAlarmManager.Stub() {        @Override        public void set(String callingPackage,                int type, long triggerAtTime, long windowLength, long interval, int flags,                PendingIntent operation, IAlarmListener directReceiver, String listenerTag,                WorkSource workSource, AlarmManager.AlarmClockInfo alarmClock) {            final int callingUid = Binder.getCallingUid();            if (interval != 0) {                if (directReceiver != null) {                    throw new IllegalArgumentException("Repeating alarms cannot use AlarmReceivers");                }            }            if (workSource != null) {                getContext().enforcePermission(                        android.Manifest.permission.UPDATE_DEVICE_STATS,                        Binder.getCallingPid(), callingUid, "AlarmManager.set");            }            // No incoming callers can request either WAKE_FROM_IDLE or            // ALLOW_WHILE_IDLE_UNRESTRICTED -- we will apply those later as appropriate.            flags &= ~(AlarmManager.FLAG_WAKE_FROM_IDLE                    | AlarmManager.FLAG_ALLOW_WHILE_IDLE_UNRESTRICTED);            // Only the system can use FLAG_IDLE_UNTIL -- this is used to tell the alarm            // manager when to come out of idle mode, which is only for DeviceIdleController.            if (callingUid != Process.SYSTEM_UID) {                flags &= ~AlarmManager.FLAG_IDLE_UNTIL;            }            if (windowLength == AlarmManager.WINDOW_EXACT) {                flags |= AlarmManager.FLAG_STANDALONE;            }            if (alarmClock != null) {                flags |= AlarmManager.FLAG_WAKE_FROM_IDLE | AlarmManager.FLAG_STANDALONE;            } else if (workSource == null && (callingUid < Process.FIRST_APPLICATION_UID                    || Arrays.binarySearch(mDeviceIdleUserWhitelist,                            UserHandle.getAppId(callingUid)) >= 0)) {                flags |= AlarmManager.FLAG_ALLOW_WHILE_IDLE_UNRESTRICTED;                flags &= ~AlarmManager.FLAG_ALLOW_WHILE_IDLE;            }            setImpl(type, triggerAtTime, windowLength, interval, operation, directReceiver,                    listenerTag, flags, workSource, alarmClock, callingUid, callingPackage);        }    }}

虽然有人觉得用AlarmManager能够在应用关闭的情况下，定时器还能再唤起，经过自己的测试，当杀掉应用程序的进程，AlarmManager的receiver也是接收不到消息的，但是我相信在这里定时器肯定是发送了，但是作为接收方的应用程序进程被杀掉了，执行不了对应的代码。不过有人也觉得AlarmManager更耗电，是因为我们执行定时任务的情况会频繁唤起cpu，但是如果只是用来只是执行延迟任务的话，个人觉得和Handler.postDelayed()相比应该也不会耗电多的。

2.在上面的第四种方法，达到的延迟会一直通过Thread.sleep来达到延迟的话，会一直占用cpu的资源，这种方法不赞同使用。

3.总结

如上面我们看到的这样，如果是单纯的实现一个任务的延迟的话，我们可以用Handler.postDelayed()和AlarmManager.set()来实现，用（4）的方法Thread.sleep()的话，首先开启一个新的线程，然后会持有cpu的资源，用（1）的方法，Timer，会开启一个死循环的线程，这样在资源上面都有点浪费。