简单的多线程

批注：由于python GIL破坏了python使用cpu执行并发的可能性，

多线程不纯粹，但至少保证能够跑多个任务，共享全局变量不出错

1、最简单的多线程

import  threading
import time
#简单的并发
def run(n):
    print("task ",n)
    time.sleep(2)
start_time = time.time()
t_thread = []
for i in range(50):
    t = threading.Thread(target=run,args=("task-%s" %i,))
    t.start()
    t_thread.append(t)

for t in t_thread:
    t.join()

print("main thread .....",threading.current_thread())
print("cost time:",time.time()-start_time)

View Code

2、守护线程

import  threading
import time
#简单的并发
def run(n):
    print("task ",n)
    time.sleep(1)
    print("task-%s...done",threading.current_thread())
start_time = time.time()
t_thread = []
for i in range(50):
    t = threading.Thread(target=run,args=("task-%s" %i,))
    t.setDaemon(True) #把当前线程设置为守护线程
    t.start()
    t_thread.append(t)

#
# for t in t_thread:
#     t.join()
time.sleep(1)
print("main thread .....",threading.current_thread())
print("cost time:",time.time()-start_time)

注意这里和上面的差别，运行时间能看出来，因为主线程已经不需要等待子线程运行完成

不知道用的多不多，以后遇上了再更新使用

3、线程上锁(互斥)

import  threading
import time
#简单的并发
def run(n):
    print("task ",n)
    lock.acquire()
    time.sleep(2)
    global num
    num += 1
    lock.release()
    print("task-%s...done",threading.current_thread())
num = 0
# lock = threading.Lock()
lock = threading.RLock()#在多重锁的情况使用，递归锁，否则死锁
start_time = time.time()
t_thread = []

for i in range(10):
    t = threading.Thread(target=run,args=("task-%s" %i,))
    t.start()
    t_thread.append(t)

for t in t_thread:
    t.join()
while threading.active_count() !=1:
    print(threading.active_count())
else:
    print("main thread .....",threading.current_thread())
    print("num ",num)
    print("cost time:",time.time()-start_time)

View Code

多个线程同时访问全局变量需要锁住临界区

4、简单的信号量使用

import threading,os
import time
def run(n):
    semaphore.acquire()
    print("run ",n)
    time.sleep(1)
    semaphore.release()

semaphore = threading.BoundedSemaphore(5)
for i in range(10):
    t = threading.Thread(target=run,args=("task-%s" %i,))
    t.start()

while threading.active_count() != 1:
    pass
else:
    print("main thread....",threading.current_thread())

View Code

5、Event的使用(同步)

import  threading
import time

def light():
    count = 0
    while True:
        if count >5 and count <10:
            #红灯
            event.clear()
            print("\033[41;1mred light is on...\033[0m")
        elif count >=10:
            #标志位
            event.set()
            count = 0
        else:
            print("\033[42;1mgreen light is on...\033[0m")
        time.sleep(1)
        count += 1

def car(name):
    while True:
        if event.is_set():
            print("%s running..." %name)
            time.sleep(1)
        else:
            print("red light on....")
            event.wait()
event = threading.Event()
event.set()
lighter = threading.Thread(target=light)
lighter.start()

car = threading.Thread(target=car,args=("aotu",))
car.start()

View Code

6、生产者、消费者模型

扫描二维码关注公众号，回复： 2762762 查看本文章

import queue,time
import threading
q = queue.Queue(maxsize=10)

def producer(name):

    count = 1
    while True:
        print("%s produce %d" %(name,count))
        q.put("product %s" %count)
        time.sleep(0.5)
        count += 1

def consumer(name):
    while  True:
        if q.qsize() > 0:
            print("%s is eating %s" %(name,q.get()))

        else:
            continue

p = threading.Thread(target=producer,args=("qu",))

c = threading.Thread(target=consumer,args=("Li",))
c1 = threading.Thread(target=consumer,args=("wang",))
p.start()
c.start()
c1.start()

View Code

线程中可以串接各种操作，使用面也很广，因此很重要！

猜你喜欢