mirrors/youtube-dl
1
0
Fork 0
mirror of https://github.com/ytdl-org/youtube-dl.git synced 2024-12-22 06:15:08 +00:00
Code Issues Projects Releases Wiki Activity

[generic] Improve KVS (etc) extraction

Browse Source 
* detect kt_player('kt_player', 'https://.../kt_player.swf?v=5...
* detect age limit if 18 USC 2257 is mentioned
* test with shooshtime.com

Partially resolves #31332.
This commit is contained in:
dirkf 2022-11-13 14:59:30 +00:00 committed by GitHub
parent 1a4fbe8462
commit fc2beab0e7
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 22 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
youtube_dl/extractor/generic.py
View File

@ -36,6 +36,7 @@ from ..utils import (
unsmuggle_url,
UnsupportedError,
url_or_none,
urljoin,
xpath_attr,
xpath_text,
xpath_with_ns,
@ -2308,6 +2309,17 @@ class GenericIE(InfoExtractor):
'height': 720,
'age_limit': 18,
},
}, {
'url': 'https://shooshtime.com/videos/284002/just-out-of-the-shower-joi/',
'md5': 'e2f0a4c329f7986280b7328e24036d60',
'info_dict': {
'id': '284002',
'display_id': 'just-out-of-the-shower-joi',
'ext': 'mp4',
'title': 'Just Out Of The Shower JOI - Shooshtime',
'height': 720,
'age_limit': 18,
},
},
]
@ -2477,7 +2489,7 @@ class GenericIE(InfoExtractor):
format_id = flashvars.get(key + '_text', key)
formats.append(merge_dicts(
parse_resolution(format_id) or parse_resolution(flashvars[key]), {
'url': getrealurl(flashvars[key], flashvars['license_code']),
'url': urljoin(url, getrealurl(flashvars[key], flashvars['license_code'])),
'format_id': format_id,
'ext': 'mp4',
'http_headers': {'Referer': url},
@ -2704,6 +2716,7 @@ class GenericIE(InfoExtractor):
AGE_LIMIT_MARKERS = [
r'Proudly Labeled <a href="http://www\.rtalabel\.org/" title="Restricted to Adults">RTA</a>',
r'>[^<]*you acknowledge you are at least (\d+) years old',
r'>\s*(?:18\s+U(?:\.S\.C\.|SC)\s+)?(?:§+\s*)?2257\b',
]
for marker in AGE_LIMIT_MARKERS:
m = re.search(marker, webpage)
@ -3559,13 +3572,15 @@ class GenericIE(InfoExtractor):
return info_dict
# Look for generic KVS player (before ld+json for tests)
found = re.search(
r'<script\b[^>]+?\bsrc\s*=\s*(["\'])https?://(?:\S+?/)+kt_player\.js\?v=(?P<ver>(?P<maj_ver>\d+)(\.\d+)+)\1[^>]*>',
webpage)
found = self._search_regex(
(r'<script\b[^>]+?\bsrc\s*=\s*(["\'])https?://(?:\S+?/)+kt_player\.js\?v=(?P<ver>\d+(?:\.\d+)+)\1[^>]*>',
# kt_player('kt_player', 'https://i.shoosh.co/player/kt_player.swf?v=5.5.1', ...
r'kt_player\s*\(\s*(["\'])(?:(?!\1)[\w\W])+\1\s*,\s*(["\'])https?://(?:\S+?/)+kt_player\.swf\?v=(?P<ver>\d+(?:\.\d+)+)\2\s*,',
), webpage, 'KVS player', group='ver', default=False)
if found:
self.report_extraction('KVS Player')
if found.group('maj_ver') not in ('4', '5', '6'):
self.report_warning('Untested major version (%s) in player engine - download may fail.' % (found.group('ver'), ))
self.report_extraction('%s: KVS Player' % (video_id, ))
if found.split('.')[0] not in ('4', '5', '6'):
self.report_warning('Untested major version (%s) in player engine - download may fail.' % (found, ))
return merge_dicts(
self._extract_kvs(url, webpage, video_id),
info_dict)