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

Add site msin #3

Browse Source
This commit is contained in:
yoshiko2
2023-07-01 02:20:51 +08:00
parent b69467f288
commit 5cefc85462
1 changed files with 14 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
scrapinglib/msin.py
View File

@@ -12,12 +12,12 @@ 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_title_unsubscribe = '//div[@class="mv_title unsubscribe"]/text()'
expr_studio = '//div[@class="mv_writer"]/text()' expr_studio = '//a[@class="mv_writer"]/text()'
expr_director = '//div[@class="mv_writer"]/text()' expr_director = '//a[@class="mv_writer"]/text()'
expr_actor = '//div[@class="mv_writer"]/text()' expr_actor = '(//div[@class="performer_text"]/a/text())[1]'
expr_label = '//div[@class="mv_mfr"]/text()' expr_label = '//a[@class="mv_mfr"]/text()'
expr_series = '//div[@class="mv_mfr"]/text()' expr_series = '//a[@class="mv_mfr"]/text()'
expr_release = '//div[@class="mv_createDate"]/text()' expr_release = '//a[@class="mv_createDate"]/text()'
expr_cover = '//div[@class="movie_top"]/img/@src' expr_cover = '//div[@class="movie_top"]/img/@src'
expr_tags = '//div[@class="mv_tag"]/label/text()' expr_tags = '//div[@class="mv_tag"]/label/text()'
expr_genres = '//div[@class="mv_genre"]/label/text()' expr_genres = '//div[@class="mv_genre"]/label/text()'
@@ -45,6 +45,14 @@ class Msin(Parser):
result = self.dictformat(htmltree) result = self.dictformat(htmltree)
return result return result
def getActors(self, htmltree):
actors = super().getActors(htmltree)
i = 0
while i < len(actors):
actors[i] = actors[i].replace("（FC2動画）", "")
i = i + 1
return actors
def getTags(self, htmltree) -> list: def getTags(self, htmltree) -> list:
return super().getTags(htmltree) return super().getTags(htmltree)