Pythoh—7.2、线程池

概念简介

场景：进程中的线程在执行完成后就会销毁，下次执行任务时再次创建；线程的频繁创建涉及和操作系统的交互，有一定的时间和资源成本，线程池应用而生；

Python 官方文档：https://docs.python.org/zh-cn/3/library/concurrent.futures.html?

代码实现

显示多线程

• 代码实现

import time
from datetime import datetime
import threading

list = ["aa", "bb", "cc", "dd", "ee", "ff", "gg", "hh", "ii", "jj", "kk"]
lock = threading.Lock()

def deleteElement():
    global list
    while True:
        lock.acquire()
        if not list:
            lock.release()
            break
        else:
            ele = list.pop()
            print("{}-->{}-->{}".format(threading.current_thread().name,datetime.now(), ele))
            lock.release()
        time.sleep(1)


if __name__ == "__main__":
    task_a = threading.Thread(target=deleteElement, name="task-A")
    task_b = threading.Thread(target=deleteElement, name="task-B")
    task_a.start()
    task_b.start()
    task_a.join()
    task_b.join()
    print("mainThread现场执行完成")

注意：该代码中当列表为空退出时，需要同时释放 lock，否则将陷入死锁状态；

线程池引入

• 方式一

import time
from datetime import datetime
import threading
from concurrent.futures import ThreadPoolExecutor

list = ["aa", "bb", "cc", "dd", "ee", "ff", "gg", "hh", "ii", "jj", "kk"]


def outputList(ele):
    print("{}-->{}-->{}".format(threading.current_thread().name, datetime.now(), ele))
    time.sleep(1)


if __name__ == "__main__":
    pool = ThreadPoolExecutor(max_workers=3, thread_name_prefix='task')
    for x in list:
        future = pool.submit(outputList, x)
    future.result()
    print("mainThread现场执行完成")

说明：此处的 future.result() 阻塞式的获取结果，可以确保主线程的内容最后执行；

• 方式二

结果获取

from concurrent.futures import as_completed

    for o in as_completed(future_list):
        o.result()

回调函数

实现非阻塞式的结果获取

================================ over ========================================