python模块学习（threading模块中的Semaphore类和BoundedSemaphore类）

学习版本3.5.2

Semaphore类和BoundedSemaphore类都是信号量类，每次有线程获得信号量的时候（即acquire()）计数器－1，释放信号量时候(release())计数器＋1，计数器为0的时候其它线程就被阻塞无法获得信号量。当计数器为设定好的上限的时候BoundedSemaphore就无法进行release()操作了，Semaphore没有这个限制检查。

#/usr/bin/python3# -*- utf-8 -*-#学习版本3.5.2import threadingimport timenum = 2def test():    global num    n = se.acquire()    num -= 1    print("Semaphore acquire:",num)    time.sleep(1)    se.release()    num += 1    print("Semaphore release:",num)if __name__ == "__main__":    se = threading.BoundedSemaphore(num)    tlist = []    for i in range(5):        t = threading.Thread(target=test)        tlist.append(t)    for i in tlist:        i.start()    print("main")

运行结果

Semaphore acquire: 1Semaphore acquire: 0mainSemaphore release: 1Semaphore release: 2Semaphore acquire: 1Semaphore acquire: 0Semaphore release: 1Semaphore release: 2Semaphore acquire: 1Semaphore release: 2

当使用Semaphore时候，可以直接调用release()使得内置的计数器就直接加1，BoundedSemaphore这样的操作就会抛出ValueError。

#/usr/bin/python3# -*- utf-8 -*-#学习版本3.5.2import threadingimport timenum = 2def test():    global num    n = se.acquire()    num -= 1    print("Semaphore acquire:",num)    time.sleep(1)    se.release()    num += 1    print("Semaphore release:",num)if __name__ == "__main__":    se = threading.Semaphore(num)    se.release()    tlist = []    for i in range(5):        t = threading.Thread(target=test)        tlist.append(t)    for i in tlist:        i.start()    print("main")

运行结果就是3个线程3个线程一起跑，就像threading.Semaphore(3)的效果一样

Semaphore acquire: 1Semaphore acquire: 0Semaphore acquire: -1mainSemaphore release: 0Semaphore release: 1Semaphore acquire: 1Semaphore release: 2Semaphore acquire: 0Semaphore release: 1Semaphore release: 2