concurrent:并发
Python标准库为我们提供了threading和multiprocessing模块编写相应的多线程/多进程代码。从Python3.2开始,标准库为我们提供了concurrent.futures
模块,它提供了ThreadPoolExecutor
和ProcessPoolExecutor
两个类,实现了对threading
和multiprocessing
的更高级的抽象,对编写线程池/进程池提供了直接的支持。
concurrent.futures基础模块是executor和future。
Executor
Executor是一个抽象类,它不能被直接使用。它为具体的异步执行定义了一些基本的方法。 ThreadPoolExecutor和ProcessPoolExecutor继承了Executor,分别被用来创建线程池和进程池的代码。
submit方法
Executor中定义了submit()
方法,这个方法的作用是提交一个可执行的回调task
,并返回一个future实例。future对象代表的就是给定的调用。
我们使用submit方法来往线程池中加入一个task,submit返回一个Future对象,对于Future对象可以简单地理解为一个在未来完成的操作。
map方法
Exectuor还为我们提供了map方法,和内建的map用法类似。映射。
future
Future实例是由Executor.submit()
创建的。可以理解为一个在未来完成的操作,这是异步编程的基础。通常情况下,我们执行io操作,访问url时(如下)在等待结果返回之前会产生阻塞,cpu不能做其他事情,而Future的引入帮助我们在等待的这段时间可以完成其他的操作。
示例:
from concurrent.futures import ProcessPoolExecutor,ThreadPoolExecutor import os,time,random def foo(i): print('%s is running %s'%(os.getpid(),i)) time.sleep(random.randint(1, 3)) return i**2 if __name__ == '__main__': print('cpu_num:',os.cpu_count()) executor=ProcessPoolExecutor() print('executor',executor,type(executor)) # futures=[] # for i in range(10): # future=executor.submit(foo,i) # futures.append(future) futures=[executor.submit(foo,i) for i in range(10)] executor.shutdown() #程序运行到这里有明显的时间间隔,可见是在shutdown存在的情况下,程序将future全部执行完,才继续往下走的 print('主') print(futures) for future in futures: print(future.result())
输出:
cpu_num: 8 executor <concurrent.futures.process.ProcessPoolExecutor object at 0x00000276745AA978> <class 'concurrent.futures.process.ProcessPoolExecutor'> 11740 is running 0 3156 is running 1 9928 is running 2 2208 is running 3 2324 is running 4 13080 is running 5 1892 is running 6 2964 is running 7 2208 is running 8 2324 is running 9 主 [<Future at 0x27674900e10 state=finished returned int>, <Future at 0x27674949dd8 state=finished returned int>, <Future at 0x27674949e80 state=finished returned int>, <Future at 0x27674949f28 state=finished returned int>, <Future at 0x27674949fd0 state=finished returned int>, <Future at 0x2767495a0b8 state=finished returned int>, <Future at 0x2767495a198 state=finished returned int>, <Future at 0x2767495a278 state=finished returned int>, <Future at 0x2767495a358 state=finished returned int>, <Future at 0x2767495a438 state=finished returned int>] 0 1 4 9 16 25 36 49 64 81
利用ThreadProcessExecutor爬虫
from concurrent.futures import ProcessPoolExecutor,ThreadPoolExecutor import requests def get(url): r=requests.get(url) return {'url':url,'text':r.text} def parse(future): dic=future.result() #future对象调用result方法取其值、 f=open('db.text','a') date='url:%s '%len(dic['text']) f.write(date) f.close() if __name__ == '__main__': executor=ThreadPoolExecutor() url_l = ['http://cn.bing.com/', 'http://www.cnblogs.com/wupeiqi/', 'http://www.cnblogs.com/654321cc/', 'https://www.cnblogs.com/', 'http://society.people.com.cn/n1/2017/1012/c1008-29581930.html', 'http://www.xilu.com/news/shaonianxinzangyou5gedong.html', ] futures=[] for url in url_l: executor.submit(get,url).add_done_callback(parse) #与Pool进程池回调函数接收的是A函数的返回值(对象ApplyResult.get()得到的值)。 executor.shutdown() #这里回调函数parse,接收的参数是submit生成的 Future对象。 print('主')
输出: