Python爬虫之一键保存全部必应高清1080P壁纸

ay2021 2021-08-07 原文

源码：

 1 import requests
 2 from lxml import etree
 3 
 4 for i in range(1, 152):
 5     print(\'page:\t\', i)
 6     url = \'https://bing.ioliu.cn/?p={}\'.format(i)
 7 
 8     headers = {
 9         \'Host\': \'bing.ioliu.cn\',
10         \'Connection\': \'keep-alive\',
11         \'Cache-Control\': \'max-age=0\',
12         \'DNT\': \'1\',
13         \'Upgrade-Insecure-Requests\': \'1\',
14         \'User-Agent\': \'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.163 Safari/537.36\',
15         \'Sec-Fetch-Dest\': \'document\',
16         \'Accept\': \'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9\',
17         \'Sec-Fetch-Site\': \'none\',
18         \'Sec-Fetch-Mode\': \'navigate\',
19         \'Sec-Fetch-User\': \'?1\',
20         \'Accept-Encoding\': \'gzip, deflate, br\',
21         \'Accept-Language\': \'zh-CN,zh;q=0.9\',
22         \'Cookie\': \'_ga=GA1.2.1389080226.1586346124; _gid=GA1.2.1179718529.1586346124; Hm_lvt_667639aad0d4654c92786a241a486361=1586346124; likes=; Hm_lpvt_667639aad0d4654c92786a241a486361=1586347115\',
23         \'If-None-Match\': \'W/"5ae9-A6K6aP64lqd/8LCoQ4XYnQ"\'
24     }
25     res = requests.get(url, headers=headers, verify=False)
26     # print(res.text)
27     parseHtml = etree.HTML(res.text)
28     picList = parseHtml.xpath(\'//img/@src\')
29     # print(picList)
30     for pic in picList:
31         try:
32             # http://h1.ioliu.cn/bing/SantoriniAerial_ZH-CN9367767863_640x480.jpg?imageslim
33             picUrl = pic.split(\'_640\')[0] + \'_1920x1080.jpg\'
34             picName = pic.split(\'bing/\')[-1].split(\'_\')[0] + \'.jpg\'
35             picRes = requests.get(picUrl)
36             with open(picName, \'wb\') as f:
37                 f.write(picRes.content)
38 
39         except Exception as e:
40             print(i, pic, e)