Python多线程抓取代理服务器

Linux224阅读模式

Python作为一门功能强大的脚本语言来说,经常被用来写爬虫程序,下面是Python爬虫多线程抓取代理服务器

年前是用 //linuxeye.com/340.html 来抓取代理服务器的,谁知道过完年回来http://www.cnproxy.com/被墙,导致某网站不能抓取东西,停止更新。

现在改抓取 http://www.proxy.com.ru 的代理服务器,这里的代理服务器更多,更简单。

下面代码可以直接用,代码如下:

  1. #!/usr/bin/env python
  2. #coding:utf-8
  3. #BLOG: linuxeye.com
  4. import urllib2
  5. import re
  6. import threading
  7. import time
  8. import MySQLdb
  9. rawProxyList = []
  10. checkedProxyList = []
  11. #抓取代理网站
  12. targets = []
  13. for i in xrange(1,42):
  14.         target = r"http://www.proxy.com.ru/list_%d.html" % i
  15.         targets.append(target)
  16. #抓取代理服务器正则
  17. p = re.compile(r'''<tr><b><td>(\d+)</td><td>(.+?)</td><td>(\d+)</td><td>(.+?)</td><td>(.+?)</td></b></tr>''')
  18. #获取代理的类
  19. class ProxyGet(threading.Thread):
  20.     def __init__(self,target):
  21.         threading.Thread.__init__(self)
  22.         self.target = target
  23.     def getProxy(self):
  24.         print "代理服务器目标网站: " + self.target
  25.         req = urllib2.urlopen(self.target)
  26.         result = req.read()
  27.         #print chardet.detect(result)
  28.         matchs = p.findall(result)
  29. #       print matchs
  30.         for row in matchs:
  31.             ip=row[1]
  32.             port =row[2]
  33.             addr = row[4].decode("cp936").encode("utf-8")
  34.             proxy = [ip,port,addr]
  35.             print proxy
  36.             rawProxyList.append(proxy)
  37.     def run(self):
  38.         self.getProxy()
  39. #检验代理的类
  40. class ProxyCheck(threading.Thread):
  41.     def __init__(self,proxyList):
  42.         threading.Thread.__init__(self)
  43.         self.proxyList = proxyList
  44.         self.timeout = 5
  45.         self.testUrl = "http://www.baidu.com/"
  46.         self.testStr = "030173"
  47.     def checkProxy(self):
  48.         cookies = urllib2.HTTPCookieProcessor()
  49.         for proxy in self.proxyList:
  50.             proxyHandler = urllib2.ProxyHandler({"http" : r'http://%s:%s' %(proxy[0],proxy[1])})
  51.             #print r'http://%s:%s' %(proxy[0],proxy[1])
  52.             opener = urllib2.build_opener(cookies,proxyHandler)
  53.             opener.addheaders = [('User-agent', 'Mozilla/5.0 (Windows NT 6.2; WOW64; rv:22.0) Gecko/20100101 Firefox/22.0')]
  54.             #urllib2.install_opener(opener)
  55.             t1 = time.time()
  56.             try:
  57.                 #req = urllib2.urlopen("http://www.baidu.com", timeout=self.timeout)
  58.                 req = opener.open(self.testUrl, timeout=self.timeout)
  59.                 #print "urlopen is ok...."
  60.                 result = req.read()
  61.                 #print "read html...."
  62.                 timeused = time.time() - t1
  63.                 pos = result.find(self.testStr)
  64.                 #print "pos is %s" %pos
  65.                 if pos > 1:
  66.                     checkedProxyList.append((proxy[0],proxy[1],proxy[2],timeused))
  67.                     #print "ok ip: %s %s %s %s" %(proxy[0],proxy[1],proxy[2],timeused)
  68.                 else:
  69.                      continue
  70.             except Exception,e:
  71.                 #print e.message
  72.                 continue
  73.     def run(self):
  74.         self.checkProxy()
  75. if __name__ == "__main__":
  76.     getThreads = []
  77.     checkThreads = []
  78. #对每个目标网站开启一个线程负责抓取代理
  79. for i in range(len(targets)):
  80.     t = ProxyGet(targets[i])
  81.     getThreads.append(t)
  82. for i in range(len(getThreads)):
  83.     getThreads[i].start()
  84. for i in range(len(getThreads)):
  85.     getThreads[i].join()
  86. print '.'*10+"总共抓取了%s个代理" %len(rawProxyList) +'.'*10
  87. #开启20个线程负责校验,将抓取到的代理分成20份,每个线程校验一份
  88. for i in range(20):
  89.     t = ProxyCheck(rawProxyList[((len(rawProxyList)+19)/20) * i:((len(rawProxyList)+19)/20) * (i+1)])
  90.     checkThreads.append(t)
  91. for i in range(len(checkThreads)):
  92.     checkThreads[i].start()
  93. for i in range(len(checkThreads)):
  94.     checkThreads[i].join()
  95. print '.'*10+"总共有%s个代理通过校验" %len(checkedProxyList) +'.'*10
  96. #插入数据库,表结构自己创建,四个字段ip,port,speed,address
  97. def db_insert(insert_list):
  98.     try:
  99.         conn = MySQLdb.connect(host="localhost", user="root", passwd="admin",db="m_common",charset='utf8')
  100.         cursor = conn.cursor()
  101.         cursor.execute('delete from proxy')
  102.         cursor.execute('alter table proxy AUTO_INCREMENT=1')
  103.         cursor.executemany("INSERT INTO proxy(ip,port,speed,address) VALUES (%s,%s,%s,%s)",insert_list)
  104.         conn.commit()
  105.         cursor.close()
  106.         conn.close()
  107.     except MySQLdb.Error,e:
  108.         print "Mysql Error %d: %s" % (e.args[0], e.args[1])
  109. #代理排序持久化
  110. proxy_ok = []
  111. f= open("proxy_list.txt",'w+')
  112. for proxy in sorted(checkedProxyList,cmp=lambda x,y:cmp(x[3],y[3])):
  113.     if proxy[3] < 8:
  114.         #print "checked proxy is: %s:%s\t%s\t%s" %(proxy[0],proxy[1],proxy[2],proxy[3])
  115.         proxy_ok.append((proxy[0],proxy[1],proxy[3],proxy[2]))
  116.         f.write("%s:%s\t%s\t%s\n"%(proxy[0],proxy[1],proxy[2],proxy[3]))
  117. f.close()
  118. db_insert(proxy_ok)

测试结果如下:

  1. # python -V
  2. Python 2.6.6
  3. # python proxy.py
  4. ['61.58.94.179', '8088', '\xe5\x8f\xb0\xe6\xb9\xbe\xe7\x9c\x81 \xe5\x8f\xb0\xe6\xb9\xbe\xe5\xae\xbd\xe9\xa2\x91\xe9\x80\x9a\xe8\xae\xaf\xe9\xa1\xbe\xe9\x97\xae\xe8\x82\xa1\xe4\xbb\xbd\xe6\x9c\x89\xe9\x99\x90\xe5\x85\xac\xe5\x8f\xb8']
  5. ['200.84.116.99', '9064', '\xe5\xa7\x94\xe5\x86\x85\xe7\x91\x9e\xe6\x8b\x89 ']
  6. ['183.223.204.8', '8123', '\xe5\x9b\x9b\xe5\xb7\x9d\xe7\x9c\x81\xe8\x87\xaa\xe8\xb4\xa1\xe5\xb8\x82 \xe7\xa7\xbb\xe5\x8a\xa8']
  7. ..........总共抓取了1921个代理..........
  8. ..........总共有524个代理通过校验..........
  1. # more proxy_list.txt
  2. 202.106.169.142:80      北京市 联通ADSL 0.291432857513
  3. 111.13.136.59:80        北京市 移动     0.297957897186
  4. 111.13.136.56:80        北京市 移动     0.373070955276
  5. 111.206.81.248:80       北京市 联通     0.403017997742
  6. 111.13.136.58:80        北京市 移动     0.414332151413
  7. 124.202.217.134:8118    北京市 电信通   0.416817903519
  8. 124.202.183.218:8118    北京市 电信通   0.426618099213
  9. 120.132.71.232:80       北京市 联通     0.440200090408
  10. 61.232.6.164:8081       北京市 铁通     0.469615936279
  11. 118.144.96.253:80       北京市 电信通   0.485229969025
  12. 203.192.10.66:80        北京市 新华社   0.51485991478
  13. 124.202.182.22:8118     北京市 电信通   0.553130865097

数据库:

  1. mysql> select * from m_common.proxy limit 10;  
  2. +----------+-----------------+------+----------+----------------------+---------------------+  
  3. | proxy_id | ip              | port | speed    | address              | create_time         |  
  4. +----------+-----------------+------+----------+----------------------+---------------------+  
  5. |        1 | 202.106.169.142 |   80   | 0.291433 | 北京市 联通ADSL | 2015-02-26 11:29:24 |  
  6. |        2 | 111.13.136.59     |   80   | 0.297958 | 北京市 移动          | 2015-02-26 11:29:24 |  
  7. |        3 | 111.13.136.56     |   80   | 0.373071 | 北京市 移动          | 2015-02-26 11:29:24 |  
  8. |        4 | 111.206.81.248   |   80   | 0.403018 | 北京市 联通          | 2015-02-26 11:29:24 |  
  9. |        5 | 111.13.136.58     |   80   | 0.414332 | 北京市 移动          | 2015-02-26 11:29:24 |  
  10. |        6 | 124.202.217.134 | 8118 | 0.416818 | 北京市 电信通       | 2015-02-26 11:29:24 |  
  11. |        7 | 124.202.183.218 | 8118 | 0.426618 | 北京市 电信通       | 2015-02-26 11:29:24 |  
  12. |        8 | 120.132.71.232   |   80   |   0.4402   | 北京市 联通          | 2015-02-26 11:29:24 |  
  13. |        9 | 61.232.6.164       | 8081 | 0.469616 | 北京市 铁通          | 2015-02-26 11:29:24 |  
  14. |       10 | 118.144.96.253  |   80   |  0.48523  | 北京市 电信通       | 2015-02-26 11:29:24 |  
  15. +----------+-----------------+------+----------+----------------------+---------------------+  
  16. 10 rows in set (0.00 sec)  
Thu Feb 26 11:30:21 CST 2015

 
  • 本文由 yeho 发表于 2015-02-26
  • 转载请务必保留本文链接:https://linuxeye.com/410.html
脚本

腾讯云COS上传、批量删除工具(Python)

腾讯云对象存储COS是类似于阿里云OSS,相比OSS,COS提供每月免费额度:存储空间50G、外网访问流量10G(内网免费)、免费读请求100万次、写请求10万次。对网站备份来说不错,但是,腾讯云提供...
    • 张文龙
      张文龙

      不能用了呀

    匿名

    发表评论

    匿名网友
    确定

    拖动滑块以完成验证