bacow/AV_Data_Capture
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

madou.py:simp by regex

Browse Source
This commit is contained in:
lededev
2022-03-06 20:39:59 +08:00
parent 3117b3a18d
commit 8ad4997342
1 changed files with 1 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
WebCrawler/madou.py
View File

@@ -19,8 +19,7 @@ def getTitle(html, number): # 获取标题
# <title>MD0140-2 / 家有性事EP2 爱在身边-麻豆社</title> # <title>MD0140-2 / 家有性事EP2 爱在身边-麻豆社</title>
# <title>MAD039 机灵可爱小叫花 强诱僧人迫犯色戒-麻豆社</title> # <title>MAD039 机灵可爱小叫花 强诱僧人迫犯色戒-麻豆社</title>
browser_title = str(html.xpath("/html/head/title/text()")[0]) browser_title = str(html.xpath("/html/head/title/text()")[0])
browser_title = browser_title[browser_title.find(' '):].replace('/','').strip() return str(re.findall(r'^.*?( / | )(.*)-麻豆社$', browser_title)[0][1]).strip()
return browser_title[:browser_title.find('-麻豆社')].strip()
def getStudio(html): # 获取厂商 已修改 def getStudio(html): # 获取厂商 已修改