http://zulko.github.io/blog/2013/09/19/a-basic-example-of-threads-synchronization-in-python/

We will see how to use threading Events to have functions in different Python threads start at the same time.

I recently coded a method to view movies in Python : it plays the video, and in the same time, in a parralel thread, it renders the audio. The difficult part is that the audio and video should be exactly synchronized. The pseudo-code looks like this:

1
2
3
4
def view(movie):

    new_thread( play_audio( movie ) )
play_video( movie )

In this code, play_audio() and play_video() will start at approximately the same time and will run parallely, but these functions need some preparation before actually starting playing stuff. Their code looks like that:

1
2
3
4
5
6
7
8
9
10
def play_audio(movie):

    audio = prepare_audio( movie )
audio.start_playing() def play_video(movie): video = prepare_video( movie )
video.start_playing()

To have a well-synchronized movie we need the internal functions audio.start_playing() and video.start_playing(), which are run in two separate threads, to start at exactly the same time. How do we do that ?

The solution seems to be using threading.Event objects. An Event is an object that can be accessed from all the threads and allows very basic communication between them : each thread can set or unset an Event, or check whether this event has already been been set (by another thread).

For our problem we will use two events video_ready and audio_ready which will enable our two threads to scream at each other “I am ready ! Are you ?”. Here is the Python fot that:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
import threading

def play_audio(movie, audio_ready, video_ready):

    audio = prepare_audio( movie )

    audio_ready.set() # Say "I'm ready" to play_video()
video_ready.wait() # Wait for play_video() to say "I'm ready" audio.start_playing() def play_video(movie, audio_ready, video_ready): video = prepare_video( movie ) video_ready.set() # Say "I'm ready" to play_audio()
audio_ready.wait() # Wait for play_audio() to say "I'm ready" video.start_playing()

and finally the code for view(movie):

1
2
3
4
5
6
7
8
9
10
11
12
def view(movie):

    audio_ready = threading.Event()
video_ready = threading.Event() # launch the parrallel audio thread
audiothread = threading.Thread(target=play_audio,
args = (movie, audio_ready, video_ready))
audiothread.start() play_video(movie, audio_ready, video_ready)

A few tips tips to go further:

  • Here I am using the module threading, and the two threads will be played in parrallel on the same processor. If you have a computer with several processors you can also use the multiprocessing module to have your threads played on two different processors (which can be MUCH faster). Nicely enough the two modules have the same syntax: simply replace threading by multiprocessing and Thread by Process in the example above and it should work.
  • In my original program, I also use an Event to terminate play_video and play_audio at the same time: when the video playing is exited, play_video unsets that Event. In play_audio, this event is regularly checked, and when it is seen to be unset, play_audio exits too.
  • Instead of using wait to wait for an Event to be set, you can use a loop to you decide at which frequency you want to check the Event. Only do that if don’t mind a lag of a few milliseconds between your processes :
1
2
3
import time
while not audio_ready.is_set():
time.sleep(0.002) # sleep 2 milliseconds

Posted by Zulko Sep 19th, 2013 MoviePy, Python,, threads,

A Basic Example of Threads Synchronization in Python, python中的线程同步示例的更多相关文章

  1. Python程序中的线程操作(线程池)-concurrent模块

    目录 Python程序中的线程操作(线程池)-concurrent模块 一.Python标准模块--concurrent.futures 二.介绍 三.基本方法 四.ProcessPoolExecut ...

  2. Python并发编程-进程 线程 同步锁 线程死锁和递归锁

    进程是最小的资源单位,线程是最小的执行单位 一.进程 进程:就是一个程序在一个数据集上的一次动态执行过程. 进程由三部分组成: 1.程序:我们编写的程序用来描述进程要完成哪些功能以及如何完成 2.数据 ...

  3. 30、Python程序中的线程操作(oncurrent模块)

    进程是cpu资源分配的最小单元,一个进程中可以有多个线程. 线程是cpu计算的最小单元. 对于Python来说他的进程和线程和其他语言有差异,是有GIL锁. GIL锁 GIL锁保证一个进程中同一时刻只 ...

  4. Python程序中的线程操作-锁

    目录 一.同步锁 1.1 多个线程抢占资源的情况 1.1.1 对公共数据的操作 1.2 同步锁的引用 1.3 互斥锁与join的区别 二.死锁与递归锁 2.1 死锁 2.2 递归锁RLock 三.典型 ...

  5. Python程序中的线程操作-concurrent模块

    目录 一.Python标准模块--concurrent.futures 二.介绍 三.基本方法 四.ProcessPoolExecutor 五.ThreadPoolExecutor 六.map的用法 ...

  6. Python程序中的线程操作-创建多线程

    目录 一.python线程模块的选择 二.threading模块 三.通过threading.Thread类创建线程 3.1 创建线程的方式一 3.2 创建线程的方式二 四.多线程与多进程 4.1 p ...

  7. Python程序中的线程操作-线程队列

    目录 一.线程队列 二.先进先出 三.后进先出 四.存储数据时可设置优先级的队列 4.1 优先级队列 4.2 更多方法说明 一.线程队列 queue队列:使用import queue,用法与进程Que ...

  8. [b0032] python 归纳 (十七)_线程同步_信号量Semaphore

    代码: # -*- coding: utf-8 -*- """ 多线程并发同步 ,使用信号量threading.Semaphore 逻辑: 多个线程,对同一个共享变量 , ...

  9. Python程序中的线程操作-守护线程

    目录 一.守护线程 1.1 详细解释 1.2 守护线程例1 1.3 守护线程例2 一.守护线程 无论是进程还是线程,都遵循:守护xx会等待主xx运行完毕后被销毁.需要强调的是:运行完毕并非终止运行. ...

随机推荐

  1. json格式的javascript对象用法分析

    格式: objectName = {  property1:value1,  property2:value2,  …,  propertyN:valueN } property是对象的属性 ,val ...

  2. Buff系统

    BUFF状态可以通过游戏道具.使用技能.被攻击技能.NPC.宠物等等实现.BUFF状态,有很多技能在释放后,会对目标产生一定时间的额外影响,这些影响有的是增益的,有的是减免的.比如法师的“熔岩地”,会 ...

  3. 静态变量数组实现LRU算法

    LRU算法的解释详情请见 https://baike.baidu.com/item/LRU/1269842 这里百度百科给出的比较详细,然后后面有一个例子 说 LRU(least recently u ...

  4. 什么原因接触接触impala的

    最近一个项目,关于大数据的改造项目,底层选择Impala还是sparkSQL呢? 最后选择Impala.这样就开启了我的Impala学习之旅.我大部分负责Imapa接口开发工作. 我是控制不住的想整个 ...

  5. Unity鼠标点击Collider

    void OnGUI() { if (Event.current != null && Event.current.type == EventType.mouseDown) { )) ...

  6. Android储存

    Android储存一共5种方法 一: 手机内置,外部储存 1.获取本地存储 (Android的读写文件及权限设置) getFilesDir()   data/data/包名/File getCache ...

  7. STM32学习之路之MDK安装篇

  8. 设计模式之单例模式(JAVA实现)

    单例模式之自我介绍 我,单例模式(Singleton Pattern)是一个比较简单的模式,我的定义如下: Ensure a class has only one instance,and provi ...

  9. Git介绍和基本原理

    官方文档:http://git-scm.com/doc 1.1 起步 - 关于版本控制 本章关于开始学习 Git. 我们从介绍有关版本控制工具的一些背景知识开始,然后讲解如何在你的系统运行 Git,最 ...

  10. N小时改变一次url时间戳的方法

    //为url添加时间戳//time 为多长时间改变一次时间戳,以小时为单位function setTimeStamp(url, time){    var time = time || 4,      ...