Python爬虫要经历爬虫、爬虫被限制、爬虫反限制的过程。当然后续还要网页爬虫限制优化，爬虫再反限制的一系列道高一尺魔高一丈的过程。爬虫的初级阶段，添加headers和ip代理可以解决很多问题。
本人自己在爬取豆瓣读书的时候,就以为爬取次数过多,直接被封了IP.后来就研究了代理IP的问题.
(当时不知道什么情况,差点心态就崩了…),下面给大家介绍一下我自己代理IP爬取数据的问题,请大家指出不足之处.
问题
这是我的IP被封了,一开始好好的,我还以为是我的代码问题了
从网上查找了一些关于爬虫代理IP的资料,得到下面的思路
爬取一些IP,过滤掉不可用.
在requests的请求的proxies参数加入对应的IP.
继续爬取.
收工
好吧,都是废话,理论大家都懂,上面直接上代码…
思路有了,动手起来.
运行环境
Python 3.7, Pycharm
这些需要大家直接去搭建好环境…
准备工作
爬取IP地址的网站(国内高匿代理)
校验IP地址的网站
你之前被封IP的py爬虫脚本…
上面的网址看个人的情况来选取
爬取IP的完整代码
PS:简单的使用bs4获取IP和端口号,没有啥难度,里面增加了一个过滤不可用IP的逻辑
关键地方都有注释了
#!/usr/bin/env python3
- - coding: utf-8 - -
@Time : 2018/11/22
@Author : liangk
@Site :
@File : auto_archive_ios.py
@Software: PyCharm
import requests
from bs4 import BeautifulSoup
import json
class GetIp(object):
“”“抓取代理IP”""
def
init
(self):
“”“初始化变量”""
self.url = ‘http://www.xicidaili.com/nn/’
self.check_url = ‘https://www.ip.cn/’
self.ip_list = []
@staticmethod
def get_html(url):
“”“请求html页面信息”""
header = {
‘User-Agent’: ‘Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.77 Safari/537.36’
}
try:
request = requests.get(url=url, headers=header)
request.encoding = ‘utf-8’
html = request.text
return html
except Exception as e:
return ‘’
def get_available_ip(self, ip_address, ip_port):
“”“检测IP地址是否可用”""
header = {
‘User-Agent’: ‘Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.77 Safari/537.36’
}
ip_url_next = ‘/’ + ip_address + ‘:’ + ip_port
proxies = {‘http’: ‘http’ + ip_url_next, ‘https’: ‘https’ + ip_url_next}
try:
r = requests.get(self.check_url, headers=header, proxies=proxies, timeout=3)
html = r.text
except:
print(‘fail-%s’ % ip_address)
else:
print(‘success-%s’ % ip_address)
soup = BeautifulSoup(html, ‘lxml’)
div = soup.find(class_=‘well’)
if div:
print(div.text)
ip_info = {‘address’: ip_address, ‘port’: ip_port}
self.ip_list.append(ip_info)
def main(self):
“”“主方法”""
web_html = self.get_html(self.url)
soup = BeautifulSoup(web_html, ‘lxml’)
ip_list = soup.find(id=‘ip_list’).find_all(‘tr’)
for ip_info in ip_list:
td_list = ip_info.find_all(‘td’)
if len(td_list) > 0:
ip_address = td_list[1].text
ip_port = td_list[2].text
# 检测IP地址是否有效
self.get_available_ip(ip_address, ip_port)
# 写入有效文件
with open(‘ip.txt’, ‘w’) as file:
json.dump(self.ip_list, file)
print(self.ip_list)
程序主入口
if
name
== ‘
main
’:
get_ip = GetIp()
get_ip.main()
使用方法完整代码
PS: 主要是通过使用随机的IP来爬取,根据request_status来判断这个IP是否可以用.
为什么要这样判断?
主要是虽然上面经过了过滤,但是不代表在你爬取的时候是可以用的,所以还是得多做一个判断.
#!/usr/bin/env python3
- - coding: utf-8 - -
@Time : 2018/11/22
@Author : liangk
@Site :
@File : get_douban_books.py
@Software: PyCharm
from bs4 import BeautifulSoup
import datetime
import requests
import json
import random
ip_random = -1
article_tag_list = []
article_type_list = []
def get_html(url):
header = {
‘User-Agent’: ‘Mozilla/5.0 (Macintosh; Intel Mac OS X 10_14_0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.102 Safari/537.36’
}
global ip_random
ip_rand, proxies = get_proxie(ip_random)
print(proxies)
try:
request = requests.get(url=url, headers=header, proxies=proxies, timeout=3)
except:
request_status = 500
else:
request_status = request.status_code
print(request_status)
while request_status != 200:
ip_random = -1
ip_rand, proxies = get_proxie(ip_random)
print(proxies)
try:
request = requests.get(url=url, headers=header, proxies=proxies, timeout=3)
except:
request_status = 500
else:
request_status = request.status_code
print(request_status)
ip_random = ip_rand
request.encoding = ‘gbk’
html = request.content
print(html)
return html
def get_proxie(random_number):
with open(‘ip.txt’, ‘r’) as file:
ip_list = json.load(file)
if random_number == -1:
random_number = random.randint(0, len(ip_list) - 1)
ip_info = ip_list[random_number]
ip_url_next = ‘/’ + ip_info[‘address’] + ‘:’ + ip_info[‘port’]
proxies = {‘http’: ‘http’ + ip_url_next, ‘https’: ‘https’ + ip_url_next}
return random_number, proxies
程序主入口
if
name
== ‘
main
’:
“”“只是爬取了书籍的第一页,按照评价排序”""
start_time = datetime.datetime.now()
url = ‘https://book.douban.com/tag/?view=type&icn=index-sorttags-all’
base_url = ‘https://book.douban.com/tag/’
html = get_html(url)
soup = BeautifulSoup(html, ‘lxml’)
article_tag_list = soup.find_all(class_=‘tag-content-wrapper’)
tagCol_list = soup.find_all(class_=‘tagCol’)
for table in tagCol_list:
“”" 整理分析数据 “”"
sub_type_list = []
a = table.find_all(‘a’)
for book_type in a:
sub_type_list.append(book_type.text)
article_type_list.append(sub_type_list)
for sub in article_type_list:
for sub1 in sub:
title = ‘
’ + sub1 + '
’
print(title)
print(base_url + sub1 + ‘?start=0’ + ‘&type=S’)
with open(‘book.text’, ‘a’, encoding=‘utf-8’) as f:
f.write(’\n’ + title + ‘\n’)
f.write(url + ‘\n’)
for start in range(0, 2):
# (start * 20) 分页是0 20 40 这样的
# type=S是按评价排序
url = base_url + sub1 + ‘?start=%s’ % (start * 20) + ‘&type=S’
html = get_html(url)
soup = BeautifulSoup(html, ‘lxml’)
li = soup.find_all(class_=‘subject-item’)
for div in li:
info = div.find(class_=‘info’).find(‘a’)
img = div.find(class_=‘pic’).find(‘img’)
content = ‘书名:<%s>’ % info[‘title’] + ’ 书本图片:’ + img[‘src’] + ‘\n’
print(content)
with open(‘book.text’, ‘a’, encoding=‘utf-8’) as f:
f.write(content)
end_time = datetime.datetime.now()
print('耗时: ', (end_time - start_time).seconds)
为什么选择国内高匿代理，因为使用这样简单的代理IP,基本上就可以应付在爬爬爬着被封IP的情况了.而且没有使用自己的IP,间接的保护
