IP 프록시 풀

매번 인터넷에서 코드를 찾아야 하는데, 여기에 먼저 저장해 두어라.

# encoding=utf8
import urllib2
from bs4 import BeautifulSoup
import urllib
import socket

User_Agent = 'Mozilla/5.0 (Windows NT 6.3; WOW64; rv:43.0) Gecko/20100101 Firefox/43.0'
header = {}
header['User-Agent'] = User_Agent

'''
      IP  
'''


def getProxyIp():
    proxy = []
    for i in range(1, 2):
        try:
            url = 'http://www.xicidaili.com/nn/' + str(i)
            req = urllib2.Request(url, headers=header)
            res = urllib2.urlopen(req).read()
            soup = BeautifulSoup(res)
            ips = soup.findAll('tr')
            for x in range(1, len(ips)):
                ip = ips[x]
                tds = ip.findAll("td")
                ip_temp = tds[1].contents[0] + "\t" + tds[2].contents[0]
                proxy.append(ip_temp)
        except:
            continue
    return proxy


'''
       IP      
'''


def validateIp(proxy):
    url = "http://ip.chinaz.com/getip.aspx"
    f = open("C:\ip.txt", "w")
    socket.setdefaulttimeout(3)
    for i in range(0, len(proxy)):
        try:
            ip = proxy[i].strip().split("\t")
            proxy_host = "http://" + ip[0] + ":" + ip[1]
            proxy_temp = {"http": proxy_host}
            res = urllib.urlopen(url, proxies=proxy_temp).read()
            f.write(proxy[i] + '
')
            print proxy[i]
        except Exception, e:
            continue
    f.close()


if __name__ == '__main__':
    proxy = getProxyIp()
    validateIp(proxy)