mirror of
https://gh.wpcy.net/https://github.com/SoloSynth1/wordpress-scraper.git
synced 2026-04-30 05:36:18 +08:00
18 lines
647 B
Python
18 lines
647 B
Python
from legacy.crawler import crawler
|
|
import os
|
|
|
|
url = "https://www.scmagazine.com"
|
|
output_dir = os.path.join('.', 'data', 'www.scmagazine.com')
|
|
|
|
headers = {
|
|
'Accept':'application/json, text/javascript, */*; q=0.01',
|
|
'Accept-Language':'zh-CN,zh;q=0.8',
|
|
'Connection':'keep-alive',
|
|
'User-Agent':'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.81 Safari/537.36',
|
|
'X-Requested-With':'XMLHttpRequest'
|
|
}
|
|
|
|
if __name__ == "__main__":
|
|
from legacy import common_crawl
|
|
wpc = crawler.WordPressCrawler(url, headers, output_dir)
|
|
common_crawl(wpc)
|
