From 70a1c3f4254e46077b76828f40af6e3fa24b0e4b Mon Sep 17 00:00:00 2001 From: Jun Zhou Date: Sun, 13 Sep 2015 19:29:36 -0700 Subject: [PATCH] [youku] support multi-page playlists --- src/you_get/extractors/youku.py | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/src/you_get/extractors/youku.py b/src/you_get/extractors/youku.py index 448feeb5..3bf4435f 100644 --- a/src/you_get/extractors/youku.py +++ b/src/you_get/extractors/youku.py @@ -74,6 +74,11 @@ class Youku(VideoExtractor): video_page = get_content('http://www.youku.com/playlist_show/id_%s' % playlist_id) videos = set(re.findall(r'href="(http://v\.youku\.com/[^?"]+)', video_page)) + + for extra_page_url in set(re.findall('href="(http://www\.youku\.com/playlist_show/id_%s_[^?"]+)' % playlist_id, video_page)): + extra_page = get_content(extra_page_url) + videos |= set(re.findall(r'href="(http://v\.youku\.com/[^?"]+)', extra_page)) + self.title = re.search(r'