diff --git a/26.03.24_48.py b/26.03.24_48.py new file mode 100644 index 0000000..edd55e3 --- /dev/null +++ b/26.03.24_48.py @@ -0,0 +1,18 @@ +import requests +from bs4 import BeautifulSoup +header={'User-Agent':'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36'} +url = 'https://www.douban.com/doulist/3936288/' +response = requests.get(url,headers=header,timeout=10) +response.encoding='utf-8' +print(response.status_code) +soup=BeautifulSoup(response.text,'html.parser') +movies=[] +for a in soup.find_all('a'): + print(a) + href=a.get('href','') + if '/subject' in href: + title = a.get_text(strip=True) + print(title) + movies.append(title) +print('----------') +print(movies) \ No newline at end of file