mirror of
https://github.com/l1ving/youtube-dl
synced 2024-12-31 02:12:52 +08:00
[imdb] Fix duplicated entries bug
This commit is contained in:
parent
075911d48e
commit
a3978a6159
@ -69,12 +69,9 @@ class ImdbListIE(InfoExtractor):
|
|||||||
list_id = mobj.group('id')
|
list_id = mobj.group('id')
|
||||||
|
|
||||||
webpage = self._download_webpage(url, list_id)
|
webpage = self._download_webpage(url, list_id)
|
||||||
list_code = self._search_regex(
|
|
||||||
r'(?s)<div\s+class="list\sdetail">(.*?)class="see-more"',
|
|
||||||
webpage, 'list code')
|
|
||||||
entries = [
|
entries = [
|
||||||
self.url_result('http://www.imdb.com' + m, 'Imdb')
|
self.url_result('http://www.imdb.com' + m, 'Imdb')
|
||||||
for m in re.findall(r'href="(/video/imdb/vi[^"]+)"', webpage)]
|
for m in re.findall(r'href="(/video/imdb/vi[^"]+)"\s+data-type="playlist"', webpage)]
|
||||||
|
|
||||||
list_title = self._html_search_regex(
|
list_title = self._html_search_regex(
|
||||||
r'<h1 class="header">(.*?)</h1>', webpage, 'list title')
|
r'<h1 class="header">(.*?)</h1>', webpage, 'list title')
|
||||||
|
Loading…
Reference in New Issue
Block a user