| 123456789101112131415161718192021222324252627282930 |
- # -*- coding: utf-8 -*-
- import re
- import time
- import random
- import httpx
- all_urls = []
- for i in range(151, 201):
- url = 'https://www.flaticon.com/stickers-packs/{}'.format(i)
- headers = {
- 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.169 Safari/537.36'
- }
- resp = httpx.get(url, headers=headers)
- resp.encoding = 'utf-8'
- page = resp.text
- url_list = re.findall('data-pin-url="(.*?)"', page)
- for u in url_list:
- if '{{link}}' not in u:
- all_urls.append(u)
- time.sleep(random.uniform(1, 2))
- with open('url_file_0_flaticon_urls.txt', 'w', encoding='utf-8') as file:
- for url in all_urls:
- file.write(url + '\n')
|