Python3简单实现多任务(线程/协程篇)

写在前面

• 上一篇文章[Python3简单实现多任务(多进程篇)]已经介绍了python多进程实现多任务的简单实现方法;
• 这次讲一讲python创建多任务另外两种常见的方式:

• 协程和线程

---

线程多任务实现1:直接使用Thread创建线程

from threading import Thread
import threading
import os
import time
import random

def not_know(thread_num):
    
    print("第%d线程吟唱:不知天上宫阙"%(thread_num))
    time.sleep(random.random())
    print("第%d线程吟唱:今夕是何年"%(thread_num))
    time.sleep(random.random())
    print("第%d号线程:<吟唱古诗>任务结束..."%(thread_num))

def main():
    for i in range(1, 6):
        num = len(threading.enumerate())
        print("当前线程数为:%d"%num)
        t = Thread(target=not_know, args=(i,))
        t.start()
        time.sleep(0.8)

if __name__ == "__main__":
    print("--->主函数开始运行<---")
    main()
    print("--->主函数运行完毕<---")

线程多任务实现2:定义类继承threading.Thread,然后重写run方法(run方法相当于功能函数)

from threading import Thread
import threading
import os
import random
import time

class the_cosmetic(threading.Thread):
    def __init__(self, num):
        self.num = num
        # 一定要记得调用父类构造方法
        threading.Thread.__init__(self)

    def run(self):
        print("-->第%d线程开始执行<--"%self.num)
        time.sleep(random.random())
        print("%d最有效的化妆品是什么?"%self.num)
        time.sleep(random.random())
        print("%dPhotoshop是最好的化妆品!"%self.num)
        time.sleep(random.random())
        print("-->第%d线程执行完毕<--"%self.num)

def main():
    print("-------->开始创建线程<--------")

    for i in range(1, 6):
        t = the_cosmetic(i)
        t.start()

    print("-------->线程创建完毕<--------")

if __name__ == "__main__":
    main()

---

协程多任务实现1:gevent(使用简单,推荐!需要pip安装gevent)

sudo pip3 install gevent

import time
import random
import gevent
from gevent import monkey

monkey.patch_all()

def peach(name):
    for i in range(1, 6):
        start_time = time.time()
        time.sleep(random.random())
        end_time = time.time()
        # 使用 round() 控制小数点位数!
        print("%s产出第%s个桃子,耗时%s"%(name, i, round(end_time - start_time, 2)))

def apple(name):
    for i in range(1, 8):
        start_time = time.time()
        time.sleep(random.random())
        end_time = time.time()
        print("%s产出第%s个苹果,耗时%s"%(name, i, round(end_time - start_time, 2)))

def main():
    # 注意:下面的语句,没有等号! 没有等号! 没有等号!
    gevent.joinall([
        gevent.spawn(peach,"LI"),
        gevent.spawn(apple,"HO"),
        ])

if __name__ == "__main__":
    main()

协程多任务实现2:yield实现协程(yield最底层,最灵活,是python自带的模块)

import time

def to_activate():
    yield
    print("吃早饭")
    print("读文档")
    yield
    print("吃中午饭")
    print("写程序")
    yield
    print("吃晚饭")
    print("解bug")

def to_sleep():
    yield
    print("午睡")
    yield
    print("晚休")

def main():
    print("程序员的一天")
    activate = to_activate()
    sleep = to_sleep()

    # 利用yield开始在两个函数间跳转
    next(activate)
    next(sleep)
    next(activate)
    next(sleep)
    next(activate)

    print("程序员的一天结束了")



if __name__ == "__main__":
    main()

协程多任务实现3:greenlet实现协程(模块须通过pip单独安装,个人感觉这个模块封装的并不够好,所以放到最后,仅供了解)

sudo pip3 install greenlet

import time
from greenlet import greenlet

activate = None
sleep = None

def to_activate():
    print("吃早饭")
    print("读文档")
    sleep.switch()
    print("吃中午饭")
    print("写程序")
    print("吃晚饭")
    print("解bug")

def to_sleep():
    print("午睡")
    activate.switch()
    print("晚休")

def main():
    global activate
    global sleep
    print("程序员的一天")
    activate = greenlet(to_activate)
    sleep = greenlet(to_sleep)

    #从activate开始执行
    activate.switch()
    print("程序员的一天结束了")



if __name__ == "__main__":
    main()

小结

• 线程与进程相比,占用资源更少,但线程依赖于进程,一个进程可以有多个线程,进程完成任务依赖于内部的线程;
• 协程解决了线程之间争用资源引发的资源浪费,所以协程比线程占用的资源更少.