Python中进行并发编程一般使用threading和multiprocessing模块,不过大部分的并发编程任务都是派生一系列线程,从队列中收集资源,然后用队列收集结果。在这些任务中,往往需要生成线程池,concurrent.futures模块对threading和multiprocessing模块进行了进一步的包装,可以很方便地实现池的功能。
python3中concurrent.futures是标准库,在python2中还需要自己安装futures:
concurrent.futures供了ThreadPoolExecutor和ProcessPoolExecutor两个类,都继承自Executor,分别被用来创建线程池和进程池,接受max_workers参数,代表创建的线程数或者进程数。ProcessPoolExecutor的max_workers参数可以为空,程序会自动创建基于电脑cpu数目的进程数。
from concurrent.futures import ThreadPoolExecutor, ProcessPoolExecutor import requests def load_url(url): return requests.get(url) url = 'http://httpbin.org' executor = ThreadPoolExecutor(max_workers=1) future = executor.submit(load_url, url)
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能够使用done()方法判断该任务是否结束,done()方法是不阻塞的,使用result()方法可以获取任务的返回值,这个方法是阻塞的。
print future.done() print future.result().status_code
Future类似于js中的Promise,可以添加回调函数:
future.add_done_callback(fn)
回调函数fn在future取消或者完成后运行,参数是future本身。
submit()方法只能进行单个任务,用并发多个任务,需要使用map与as_completed。
URLS = ['http://httpbin.org', 'http://example.com/', 'https://api.github.com/'] def load_url(url): return requests.get(url) with ThreadPoolExecutor(max_workers=3) as executor: for url, data in zip(URLS, executor.map(load_url, URLS)): print('%r page status_code %s' % (url, data.status_code))
结果:
'http://httpbin.org' page status_code 200 'http://example.com/' page status_code 200 'https://api.github.com/' page status_code 200
map方法接收两个参数,第一个为要执行的函数,第二个为一个序列,会对序列中的每个元素都执行这个函数,返回值为执行结果组成的生成器。
由上面可以看出返回结果与序列结果的顺序是一致的
as_completed()方法返回一个Future组成的生成器,在没有任务完成的时候,会阻塞,在有某个任务完成的时候,会yield这个任务,直到所有的任务结束。
def load_url(url): return url, requests.get(url).status_code with ThreadPoolExecutor(max_workers=3) as executor: tasks = [executor.submit(load_url, url) for url in URLS] for future in as_completed(tasks): print future.result()
结果:
('http://example.com/', 200) ('http://httpbin.org', 200) ('https://api.github.com/', 200)
可以看出,结果与序列顺序不一致,先完成的任务会先通知主线程。
wait方法可以让主线程阻塞,直到满足设定的要求。有三种条件ALL_COMPLETED, FIRST_COMPLETED,FIRST_EXCEPTION。
from concurrent.futures import ThreadPoolExecutor, ProcessPoolExecutor, wait, ALL_COMPLETED, FIRST_COMPLETED from concurrent.futures import as_completed import requests URLS = ['http://httpbin.org', 'http://example.com/', 'https://api.github.com/'] def load_url(url): requests.get(url) print url with ThreadPoolExecutor(max_workers=3) as executor: tasks = [executor.submit(load_url, url) for url in URLS] wait(tasks, return_when=ALL_COMPLETED) print 'all_cone'
返回:
http://example.com/ http://httpbin.org https://api.github.com/ all_cone
可以看出阻塞到任务全部完成。
使用ProcessPoolExecutor与ThreadPoolExecutor方法基本一致,注意文档中有一句:
The __main__
module must be importable by worker subprocesses. This means that ProcessPoolExecutor
will not work in the interactive interpreter.
需要__main__模块。
def main(): with ProcessPoolExecutor() as executor: tasks = [executor.submit(load_url, url) for url in URLS] for f in as_completed(tasks): ret = f.done() if ret: print f.result().status_code if __name__ == '__main__': main()