semaphore是一个内置的计数器

每当调用acquire()时，内置计数器-1
每当调用release()时，内置计数器+1

计数器不能小于0，当计数器为0时，acquire()将阻塞线程直到其他线程调用release()。
来看下面的代码：

import time
import threading

def foo():
    time.sleep(2)   #程序休息2秒
    print("ok",time.ctime())

for i in range(20):
    t1=threading.Thread(target=foo,args=()) #实例化一个线程
    t1.start()  #启动线程

执行结果：

ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017
ok Tue Jul 18 20:05:58 2017

可以看到，程序会在很短的时间内生成20个线程来打印一句话。

如果在主机执行IO密集型任务的时候再执行这种类型的程序时，计算机就有很大可能会宕机。
这时候就可以为这段程序添加一个计数器功能，来限制一个时间点内的线程数量。

代码如下：

import time
import threading

s1=threading.Semaphore(5)   #添加一个计数器

def foo():
    s1.acquire()    #计数器获得锁
    time.sleep(2)   #程序休眠2秒
    print("ok",time.ctime())
    s1.release()    #计数器释放锁


for i in range(20):
    t1=threading.Thread(target=foo,args=()) #创建线程
    t1.start()  #启动线程   

执行结果：

ok Tue Jul 18 20:04:38 2017
ok Tue Jul 18 20:04:38 2017
ok Tue Jul 18 20:04:38 2017
ok Tue Jul 18 20:04:38 2017
ok Tue Jul 18 20:04:38 2017
ok Tue Jul 18 20:04:40 2017
ok Tue Jul 18 20:04:40 2017
ok Tue Jul 18 20:04:40 2017
ok Tue Jul 18 20:04:40 2017
ok Tue Jul 18 20:04:40 2017
ok Tue Jul 18 20:04:42 2017
ok Tue Jul 18 20:04:42 2017
ok Tue Jul 18 20:04:42 2017
ok Tue Jul 18 20:04:42 2017
ok Tue Jul 18 20:04:42 2017
ok Tue Jul 18 20:04:44 2017
ok Tue Jul 18 20:04:44 2017
ok Tue Jul 18 20:04:44 2017
ok Tue Jul 18 20:04:44 2017
ok Tue Jul 18 20:04:44 2017