fenix/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-12-25 11:50:42 +01:00
Code Issues Projects Releases Wiki Activity

[pornhub] Improve locked videos detection (closes #22449, closes #22780)

Browse Source
This commit is contained in:
Sergey M․ 2019-12-31 23:43:43 +07:00
parent f41347260c
commit 0164cd5dac
No known key found for this signature in database
GPG Key ID: 2C393E0F18A9236D
1 changed files with 8 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
youtube_dl/extractor/pornhub.py
View File

@ -17,6 +17,7 @@ from ..utils import (
determine_ext, determine_ext,
ExtractorError, ExtractorError,
int_or_none, int_or_none,
NO_DEFAULT,
orderedSet, orderedSet,
remove_quotes, remove_quotes,
str_to_int, str_to_int,
@ -227,9 +228,9 @@ class PornHubIE(PornHubBaseIE):
else: else:
thumbnail, duration = [None] * 2 thumbnail, duration = [None] * 2
def extract_js_vars(webpage, pattern, fatal=True): def extract_js_vars(webpage, pattern, default=NO_DEFAULT):
assignments = self._search_regex( assignments = self._search_regex(
pattern, webpage, 'encoded url', fatal=fatal) pattern, webpage, 'encoded url', default=default)
if not assignments: if not assignments:
return {} return {}
@ -270,11 +271,15 @@ class PornHubIE(PornHubBaseIE):
FORMAT_PREFIXES = ('media', 'quality') FORMAT_PREFIXES = ('media', 'quality')
js_vars = extract_js_vars( js_vars = extract_js_vars(
webpage, r'(var\s+(?:%s)_.+)' % '|'.join(FORMAT_PREFIXES), webpage, r'(var\s+(?:%s)_.+)' % '|'.join(FORMAT_PREFIXES),
fatal=False) default=None)
if js_vars: if js_vars:
for key, format_url in js_vars.items(): for key, format_url in js_vars.items():
if any(key.startswith(p) for p in FORMAT_PREFIXES): if any(key.startswith(p) for p in FORMAT_PREFIXES):
add_video_url(format_url) add_video_url(format_url)
if not video_urls and re.search(
r'<[^>]+\bid=["\']lockedPlayer', webpage):
raise ExtractorError(
'Video %s is locked' % video_id, expected=True)
if not video_urls: if not video_urls:
js_vars = extract_js_vars( js_vars = extract_js_vars(