commit 754e70cf3e74218ae5d840985fbf07bbe274332a
parent 84bc4dcb0f678f0a8c9f993e101b9769e3959f76
Author: George Brighton <george@gebn.co.uk>
Date: Sun, 2 Aug 2015 19:21:10 +0100
[pornhub] Fix video url regular expression.
PornHub seems to have subtly changed their JavaScript. Before, video URL strings were embedded directly in the video's `flashvars_*` object, but they are now assigned to variables of the form `player_quality_*`, which are then added to this object later under the relevant quality key.
Diffstat:
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/youtube_dl/extractor/pornhub.py b/youtube_dl/extractor/pornhub.py
@@ -81,7 +81,7 @@ class PornHubIE(InfoExtractor):
comment_count = self._extract_count(
r'All Comments\s*<span>\(([\d,.]+)\)', webpage, 'comment')
- video_urls = list(map(compat_urllib_parse_unquote, re.findall(r'"quality_[0-9]{3}p":"([^"]+)', webpage)))
+ video_urls = list(map(compat_urllib_parse_unquote, re.findall(r"var player_quality_[0-9]{3}p = '([^']+)'", webpage)))
if webpage.find('"encrypted":true') != -1:
password = compat_urllib_parse_unquote_plus(
self._search_regex(r'"video_title":"([^"]+)', webpage, 'password'))