爬IP代码

import requests

import re

import  dauk

from bs4 import BeautifulSoup

import time

def daili():

      print('[+]极速爬取代理IP，默认为99页')

      for b in range(1,99):

        url="http://www.xicidaili.com/nt/{}".format(b)

        header={'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; WOW64; rv:58.0) Gecko/20100101 Firefox/48.0'}

        r=requests.get(url,headers=header)

        gsx=BeautifulSoup(r.content,'html.parser')

        for line in gsx.find_all('td'):

            sf=line.get_text()

            dailix=re.findall('(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)',str(sf))

            for g in dailix:

                po=".".join(g)

                print(po)

                with open ('采集到的IP.txt','a') as l:

                    l.write(po+'\n')

daili()

def dailigaoni():

    print('[+]极速爬取代理IP，默认为99页')

    for i in range(1,99):

      url="http://www.xicidaili.com/nn/{}".format(i)

      header={'User-Agent':'Mozilla/5.0 (Windows NT 6.1 Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Safari/537.36'}

      r=requests.get(url,headers=header)

      bks=r.content

      luk=re.findall('(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)',str(bks))

      for g in luk:

          vks=".".join(g)

          print(vks)

          with open('采集到的IP.txt','a') as b:

              b.write(vks+'\n')

dailigaoni()

def dailihtp():

    print('[+]极速爬取代理IP，默认为99页')

    for x in range(1,99):

        header="{'User-Agent':'Mozilla/5.0 (Windows NT 6.1 Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Safari/537.36'}"

        url="http://www.xicidaili.com/wn/{}".format(x)

        r=requests.get(url,headers=header)

        gs=r.content

        bs=re.findall('(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)',gs)

        for kl in bs:

            kgf=".".join(kl)

            print(kgf)

            with open ('采集到的IP.txt','a') as h:

                h.write(kgf)

dailihtp()

def dailihttps():

    print('[+]极速爬代理IP,默认为99页')

    for s in range(1,99):

        url="http://www.xicidaili.com/wt/{}".format(s)

        header={'User-Agent':'Mozilla/5.0 (Windows NT 6.1 Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Safari/537.36'}

        r=requests.get(url,headers=header)

        kl=r.content

        lox=re.findall('(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)\.(25[0-5]|2[0-4]\d|[0-1]\d{2}|[1-9]?\d)',kl)

        for lk in lox:

            los=".".join(lk)

            print(los)

            with open('采集到的IP.txt','a') as lp:

                lp.write(los)

dailihttps()

　端口代码

import requests

import re

from bs4 import BeautifulSoup

def daili():

    print('[+]极速爬取代理IP端口，默认为99页')

    for b in range(1, 99):

        url = "http://www.xicidaili.com/nt/{}".format(b)

        header = {'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; WOW64; rv:58.0) Gecko/20100101 Firefox/48.0'}

        r = requests.get(url, headers=header)

        gsx = BeautifulSoup(r.content, 'html.parser')

        for line in gsx.find_all('td'):

            sf = line.get_text()

            dailix = re.findall(

                '<td>([0-9]|[1-9]\d{1,3}|[1-5]\d{4}|6[0-5]{2}[0-3][0-5])</td>',

                str(sf))

            for g in dailix:

                po = ".".join(g)

                print(po )

                with open('采集到的端口.txt.txt', 'a') as l:

                    l.write(po + '\n')

daili()

def dailigaoni():

    print('[+]极速爬取代理IP的端口，默认为99页')

    for i in range(1, 99):

        url = "http://www.xicidaili.com/nn/{}".format(i)

        header = {

            'User-Agent': 'Mozilla/5.0 (Windows NT 6.1 Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Safari/537.36'}

        r = requests.get(url, headers=header)

        bks = r.content

        luk = re.findall(

            '<td>([0-9]|[1-9]\d{1,3}|[1-5]\d{4}|6[0-5]{2}[0-3][0-5])</td>',

            str(bks))

        for g in luk:

            vks = ".".join(g)

            print(vks)

            with open('采集到的端口.txt.txt', 'a') as b:

                b.write(vks + '\n')

dailigaoni()

def dailihtp():

    print('[+]极速爬取代理IP，默认为99页')

    for x in range(1, 99):

        header = "{'User-Agent':'Mozilla/5.0 (Windows NT 6.1 Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Safari/537.36'}"

        url = "http://www.xicidaili.com/wn/{}".format(x)

        r = requests.get(url, headers=header)

        gs = r.content

        bs = re.findall(

            '<td>([0-9]|[1-9]\d{1,3}|[1-5]\d{4}|6[0-5]{2}[0-3][0-5])</td>',

            gs)

        for kl in bs:

            kgf = ".".join(kl)

            print(kgf)

            with open('采集到的端口.txt.txt', 'a') as h:

                h.write(kgf)

dailihtp()

def dailihttps():

    print('[+]极速爬代理IP的端口,默认为99页')

    for s in range(1, 99):

        url = "http://www.xicidaili.com/wt/{}".format(s)

        header = {

            'User-Agent': 'Mozilla/5.0 (Windows NT 6.1 Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Safari/537.36'}

        r = requests.get(url, headers=header)

        kl = r.content

        lox = re.findall(

            '<td>([0-9]|[1-9]\d{1,3}|[1-5]\d{4}|6[0-5]{2}[0-3][0-5])</td>',

            kl)

        for lk in lox:

            los = ".".join(lk)

            print(los)

            with open('采集到的端口.txt', 'a') as lp:

                lp.write(los)

dailihttps()

　　调用代码

print('''

                   _ooOoo_

                  o8888888o

                  88" . "88

                  (| -_- |)

                  O\  =  /O

               ____/`---'\____

             .'  \\|     |//  `.

            /  \\|||  :  |||//  \

           /  _||||| -:- |||||-  \

           |   | \\\  -  /// |   |

           | \_|  ''\---/''  |   |

           \  .-\__  `-`  ___/-. /

         ___`. .'  /--.--\  `. . __

      ."" '<  `.___\_<|>_/___.'  >'"".

     | | :  `- \`.;`\ _ /`;.`/ - ` : | |

     \  \ `-.   \_ __\ /__ _/   .-` /  /

======`-.____`-.___\_____/___.-`____.-'======

                   `=---='

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

            佛祖保佑       永无BUG

            ''')

print('[!]爬虫速度过快，导致IP被封请更换IP')

print('[*]极速爬取代理IP')

print('1.普通代理IP')

print('2.高匿代理IP')

print('3.http代理IP')

print('4.https代理IP')

bk=input('请选择：')

def xs():

  import 代理.daili

  import 代理.dauk

  if bk=='1':

        代理.daili.daili.daili()

        代理.dauk.daili()

        exit()

  elif bk=='2':

      代理.daili.daili.dailigaoni()

      代理.dauk.dailigaoni()

      exit()

  elif bk=='3':

      代理.daili.daili.dailihtp()

      代理.dauk.dailihtp()

      exit()

  elif bk=='4':

      代理.daili .daili.dailihttps()

      代理.dauk.dailihttps()

      exit()

  elif bk=='q':

      exit()

  else:

      print('[-]没有找到你要的选项')

xs()

2018-02-17

最新文章

  1. 【JavaWeb】Spring+SpringMVC+MyBatis+SpringSecurity+EhCache+JCaptcha 完整Web基础框架（五）
  2. [译]Object.getPrototypeOf
  3. Html限制input输入框只能输入数字
  4. ssh项目java.lang.ClassNotFoundException: org.springframework.web.context.ContextLoade错误
  5. 一起来学习DOJO吧--序
  6. PHP的学习--生成器Generators
  7. 管理故事&mdash;&mdash;和尚挑水的故事
  8. 年度十佳 DevOps 博客文章（后篇）
  9. 微软原版SQLHelper类
  10. pecl/mongo is already installed
  11. Square spiral
  12. Effective C++ Item 40 明智而审慎地使用多重继承
  13. [Apio2012]dispatching
  14. 深度学习之tensorflow （一）
  15. 100Mbps和100Mb/s有什么不同
  16. django基础 -- 10.form , ModelForm ,modelformset
  17. SpringMVC不能引入静态资源
  18. [转帖]IBM POWER系列处理器的前世今生
  19. concat layer
  20. 4.1 C++多态的概念及前提条件

热门文章

  1. java多线程补：充原子性和可见性
  2. LeetCode OJ：First Bad Version（首个坏版本）
  3. 【tensorflow:Google】三、tensorflow入门
  4. toString 和 valueOf 总结
  5. tomcat8启动慢原因及解决办法
  6. flask中cookie和session介绍
  7. ASP.NET中服务器控件的生命周期
  8. 十、python沉淀之路--高阶函数初识
  9. JUnit测试，获取Spring MVC环境
  10. 如何设计并使用FireMonkeyStyle