[extractor/common] Speed-up media tags regex (closes #11979)
This commit is contained in:
parent
815482d4eb
commit
2aec7256ae
1 changed files with 6 additions and 1 deletions
|
@ -1959,7 +1959,12 @@ class InfoExtractor(object):
|
|||
media_tags = [(media_tag, media_type, '')
|
||||
for media_tag, media_type
|
||||
in re.findall(r'(?s)(<(video|audio)[^>]*/>)', webpage)]
|
||||
media_tags.extend(re.findall(r'(?s)(<(?P<tag>video|audio)[^>]*>)(.*?)</(?P=tag)>', webpage))
|
||||
media_tags.extend(re.findall(
|
||||
# We only allow video|audio followed by a whitespace or '>'.
|
||||
# Allowing more characters may end up in significant slow down (see
|
||||
# https://github.com/rg3/youtube-dl/issues/11979, example URL:
|
||||
# http://www.porntrex.com/maps/videositemap.xml).
|
||||
r'(?s)(<(?P<tag>video|audio)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>', webpage))
|
||||
for media_tag, media_type, media_content in media_tags:
|
||||
media_info = {
|
||||
'formats': [],
|
||||
|
|
Loading…
Reference in a new issue