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

Add site msin #2

Browse Source
This commit is contained in:
yoshiko2
2023-06-30 20:30:46 +08:00
parent 3a4fe159db
commit abd6503ae4
1 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
scrapinglib/msin.py
View File

@@ -11,6 +11,7 @@ class Msin(Parser):
expr_number = '//div[@class="mv_fileName"]/text()' expr_number = '//div[@class="mv_fileName"]/text()'
expr_title = '//div[@class="mv_title"]/text()' expr_title = '//div[@class="mv_title"]/text()'
expr_title_unsubscribe = '//div[@class="mv_title unsubscribe"]/text()'
expr_studio = '//div[@class="mv_writer"]/text()' expr_studio = '//div[@class="mv_writer"]/text()'
expr_director = '//div[@class="mv_writer"]/text()' expr_director = '//div[@class="mv_writer"]/text()'
expr_actor = '//div[@class="mv_writer"]/text()' expr_actor = '//div[@class="mv_writer"]/text()'
@@ -35,10 +36,12 @@ class Msin(Parser):
session = request_session(cookies=self.cookies, proxies=self.proxies, verify=self.verify) session = request_session(cookies=self.cookies, proxies=self.proxies, verify=self.verify)
htmlcode = session.get(self.detailurl).text htmlcode = session.get(self.detailurl).text
htmltree = etree.HTML(htmlcode) htmltree = etree.HTML(htmlcode)
# if title are null, use unsubscribe title
if super().getTitle(htmltree) == "":
self.expr_title = self.expr_title_unsubscribe
# if tags are null, use genres # if tags are null, use genres
if len(super().getTags(htmltree)) == 0: if len(super().getTags(htmltree)) == 0:
self.expr_tags = self.expr_genres self.expr_tags = self.expr_genres
result = self.dictformat(htmltree) result = self.dictformat(htmltree)
return result return result