github/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-21 20:46:36 -05:00
Code Issues Projects Releases Packages Wiki Activity

[GenericIE] Outsource embedded blip.tv player video id extraction to BlipTVIE and fix minor errors in RegEx

Browse source
This commit is contained in:
rzhxeo 2013-12-30 06:15:02 +01:00
parent c0f9969b9e
commit 19dab5e6cc
1 changed files with 4 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
youtube_dl/extractor/generic.py
View file

@ -271,16 +271,12 @@ def _real_extract(self, url):
}
# Look for embedded blip.tv player
mobj = re.search(r'<meta\s[^>]*https?://api.blip.tv/\w+/redirect/\w+/(\d+)', webpage)
mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
if mobj:
return self.url_result('http://blip.tv/seo/-'+mobj.group(1), 'BlipTV')
mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*https?://(?:\w+\.)?blip.tv/(?:play/|api\.swf#)([a-zA-Z0-9]+)', webpage)
return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
if mobj:
player_url = 'http://blip.tv/play/%s.x?p=1' % mobj.group(1)
player_page = self._download_webpage(player_url, mobj.group(1))
blip_video_id = self._search_regex(r'data-episode-id="(\d+)', player_page, u'blip_video_id', fatal=False)
if blip_video_id:
return self.url_result('http://blip.tv/seo/-'+blip_video_id, 'BlipTV')
return self.url_result(mobj.group(1), 'BlipTV')
# Look for Bandcamp pages with custom domain
mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)