[pbs] Fix multi part videos extraction

This commit is contained in:
Sergey M․ 2016-02-11 22:02:37 +06:00
parent a7ab46375b
commit 6b3fbd3425

View File

@ -366,10 +366,14 @@ def _extract_webpage(self, url):
webpage, 'upload date', default=None)) webpage, 'upload date', default=None))
# tabbed frontline videos # tabbed frontline videos
tabbed_videos = re.findall( MULTI_PART_REGEXES = (
r'<div[^>]+class="videotab[^"]*"[^>]+vid="(\d+)"', webpage) r'<div[^>]+class="videotab[^"]*"[^>]+vid="(\d+)"',
if tabbed_videos: r'<a[^>]+href=["\']#video-\d+["\'][^>]+data-coveid=["\'](\d+)',
return tabbed_videos, presumptive_id, upload_date )
for p in MULTI_PART_REGEXES:
tabbed_videos = re.findall(p, webpage)
if tabbed_videos:
return tabbed_videos, presumptive_id, upload_date
MEDIA_ID_REGEXES = [ MEDIA_ID_REGEXES = [
r"div\s*:\s*'videoembed'\s*,\s*mediaid\s*:\s*'(\d+)'", # frontline video embed r"div\s*:\s*'videoembed'\s*,\s*mediaid\s*:\s*'(\d+)'", # frontline video embed