38 线程池

concurrent.futures模块提供了高度封装的异步调用接口
ThreadPoolExecutor ：线程池，提供异步调用
ProcessPoolExecutor ：进程池，提供异步调用

基本方法：

1. submit(fn,*args,**kwargs) 异步提交任务。
2. map(func,*iterables,timeout = None,chunksize = 1) 取代for循环submit操作。
3. shutdown(wait = True) 相当于进程池pool.close() + pool.join()操作。
   wait = True时，等待池内所有任务执行玩毕回收玩资源后才继续。
   wait = False时，立即返回，并不会等待池内的任务执行完毕。
   但不管wait参数为何值，整个程序都会等到所有的任务执行完毕。
   submit和map必须在shutdown之前。
4. result(timeout = None) 取得结果
5. add_done_callback(fn) 回调函数

例子：

1. shubmit

import time
from concurrent.futures import ThreadPoolExecutor

def func(n):
    time.sleep(2)
    print(n)
tpool = ThreadPoolExecutor(max_workers = 5)  #个数一般起不超过cpu个数*5个
for i in range(20):
    tpool.submit(func,i)
print('主线程')

2. shutdown

import time
from concurrent.futures import ThreadPoolExecutor

def func(n):
    time.sleep(2)
    print(n)
tpool = ThreadPoolExecutor(max_workers = 5)  #个数一般起不超过cpu个数*5个
for i in range(20):
    tpool.submit(func,i)
tpool.shutdown()        #相当于执行close()和join()
print('主线程')

3. result

import time
from concurrent.futures import ThreadPoolExecutor
def func(n):
    time.sleep(2)
    print(n)
    return n * n
tpool = ThreadPoolExecutor(max_workers=5)  # 个数一般起不超过cpu个数*5个
t_list = []
for i in range(20):
    t = tpool.submit(func, i)
    t_list.append(t)
tpool.shutdown()  # 相当于执行close()和join()
print('主线程')

for t in t_list:
    print('****', t.result())

4. map

import time
from concurrent.futures import ThreadPoolExecutor


def func(n):
    time.sleep(2)
    print(n)
    return n * n


tpool = ThreadPoolExecutor(max_workers=5)  # 个数一般起不超过cpu个数*5个

tpool.map(func,range(20))   #拿不到返回值

5. callback

import time
from concurrent.futures import ThreadPoolExecutor
def func(n):
    time.sleep(2)
    print(n)
    return n * n
def call_back(m):
    print('结果是%s' % m.result())
tpool = ThreadPoolExecutor(max_workers=5)  # 个数一般起不超过cpu个数*5个
for i in range(20):
    tpool.submit(func,i).add_done_callback(call_back)