[迎风奔雨]进程+携程

#-*- coding=utf-8 -*-  
import requests  
from multiprocessing import Process  
import gevent  
from gevent import monkey; monkey.patch_all()  
  
import sys  
reload(sys)  
sys.setdefaultencoding(‘utf8‘)  
def fetch(url):  
    try:  
        s = requests.Session()  
        r = s.get(url,timeout=1)#在这里抓取页面  
    except Exception,e:  
        print e   
    return ‘‘  
   
def process_start(url_list):  
    tasks = []  
    for url in url_list:  
        tasks.append(gevent.spawn(fetch,url))  
    gevent.joinall(tasks)#使用协程来执行  
  
def task_start(filepath,flag = 100000):#每10W条url启动一个进程  
    with open(filepath,‘r‘) as reader:#从给定的文件中读取url  
        url = reader.readline().strip()  
        url_list = []#这个list用于存放协程任务  
        i = 0 #计数器，记录添加了多少个url到协程队列  
        while url!=‘‘:  
            i += 1  
            url_list.append(url)#每次读取出url，将url添加到队列  
            if i == flag:#一定数量的url就启动一个进程并执行  
                p = Process(target=process_start,args=(url_list,))  
                p.start()  
                url_list = [] #重置url队列  
                i = 0 #重置计数器  
            url = reader.readline().strip()  
        if url_list not []:#若退出循环后任务队列里还有url剩余  
            p = Process(target=process_start,args=(url_list,))#把剩余的url全都放到最后这个进程来执行  
            p.start()  
    
if __name__ == ‘__main__‘:  
    task_start(‘./testData.txt‘)#读取指定文件