Compare commits

...

6 Commits

Author SHA1 Message Date
dirkf
957d994bbb
Merge e84b4d1c9e into c5098961b0 2024-08-21 22:33:01 -04:00
dirkf
c5098961b0 [Youtube] Rework n function extraction pattern
Now also succeeds with player b12cc44b
2024-08-06 20:59:09 +01:00
dirkf
dbc08fba83 [jsinterp] Improve slice implementation for player b12cc44b
Partly taken from yt-dlp/yt-dlp#10664, thx seproDev
        Fixes #32896
2024-08-06 20:51:38 +01:00
Aiur Adept
71223bff39
[Youtube] Fix nsig extraction for player 20dfca59 (#32891)
* dirkf's patch for nsig extraction
* add generic search per  yt-dlp/yt-dlp/pull/10611 - thx bashonly

---------

Co-authored-by: dirkf <fieldhouse@gmx.net>
2024-08-01 19:18:34 +01:00
dirkf
e84b4d1c9e Sanitise info HTML for yt-dlp's get_elements_by_class() 2022-04-19 04:05:23 +01:00
dirkf
c278f31d62 [Porn91] Update extractor 2022-04-10 15:12:28 +01:00
5 changed files with 137 additions and 28 deletions

View File

@ -425,6 +425,34 @@ class TestJSInterpreter(unittest.TestCase):
self._test(jsi, [''], args=['', '-']) self._test(jsi, [''], args=['', '-'])
self._test(jsi, [], args=['', '']) self._test(jsi, [], args=['', ''])
def test_slice(self):
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice()}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0)}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(5)}', [5, 6, 7, 8])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(99)}', [])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-2)}', [7, 8])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-99)}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0, 0)}', [])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(1, 0)}', [])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0, 1)}', [0])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(3, 6)}', [3, 4, 5])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(1, -1)}', [1, 2, 3, 4, 5, 6, 7])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-1, 1)}', [])
self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-3, -1)}', [6, 7])
self._test('function f(){return "012345678".slice()}', '012345678')
self._test('function f(){return "012345678".slice(0)}', '012345678')
self._test('function f(){return "012345678".slice(5)}', '5678')
self._test('function f(){return "012345678".slice(99)}', '')
self._test('function f(){return "012345678".slice(-2)}', '78')
self._test('function f(){return "012345678".slice(-99)}', '012345678')
self._test('function f(){return "012345678".slice(0, 0)}', '')
self._test('function f(){return "012345678".slice(1, 0)}', '')
self._test('function f(){return "012345678".slice(0, 1)}', '0')
self._test('function f(){return "012345678".slice(3, 6)}', '345')
self._test('function f(){return "012345678".slice(1, -1)}', '1234567')
self._test('function f(){return "012345678".slice(-1, 1)}', '')
self._test('function f(){return "012345678".slice(-3, -1)}', '67')
if __name__ == '__main__': if __name__ == '__main__':
unittest.main() unittest.main()

View File

@ -174,6 +174,14 @@ _NSIG_TESTS = [
'https://www.youtube.com/s/player/5604538d/player_ias.vflset/en_US/base.js', 'https://www.youtube.com/s/player/5604538d/player_ias.vflset/en_US/base.js',
'7X-he4jjvMx7BCX', 'sViSydX8IHtdWA', '7X-he4jjvMx7BCX', 'sViSydX8IHtdWA',
), ),
(
'https://www.youtube.com/s/player/20dfca59/player_ias.vflset/en_US/base.js',
'-fLCxedkAk4LUTK2', 'O8kfRq1y1eyHGw',
),
(
'https://www.youtube.com/s/player/b12cc44b/player_ias.vflset/en_US/base.js',
'keLa5R2U00sR9SQK', 'N1OGyujjEwMnLw',
),
] ]

View File

@ -1,11 +1,21 @@
# coding: utf-8 # coding: utf-8
from __future__ import unicode_literals from __future__ import unicode_literals
import re
from .common import InfoExtractor from .common import InfoExtractor
from ..compat import (
compat_urllib_parse_unquote,
)
from ..utils import ( from ..utils import (
clean_html,
extract_attributes,
get_elements_by_class,
parse_duration, parse_duration,
int_or_none,
ExtractorError, ExtractorError,
str_to_int,
strip_or_none,
unified_strdate,
) )
@ -15,18 +25,21 @@ class Porn91IE(InfoExtractor):
_TEST = { _TEST = {
'url': 'http://91porn.com/view_video.php?viewkey=7e42283b4f5ab36da134', 'url': 'http://91porn.com/view_video.php?viewkey=7e42283b4f5ab36da134',
'md5': '7fcdb5349354f40d41689bd0fa8db05a', 'md5': 'd869db281402e0ef4ddef3c38b866f86',
'info_dict': { 'info_dict': {
'id': '7e42283b4f5ab36da134', 'id': '7e42283b4f5ab36da134',
'title': '18岁大一漂亮学妹水嫩性感再爽一次', 'title': '18岁大一漂亮学妹水嫩性感再爽一次',
'ext': 'mp4', 'ext': 'mp4',
'duration': 431, 'duration': 431,
'age_limit': 18, 'age_limit': 18,
'upload_date': '20150520',
'uploader': '千岁九王爷',
} }
} }
def _real_extract(self, url): def _real_extract(self, url):
video_id = self._match_id(url) video_id = self._match_id(url)
# set language for page to be extracted
self._set_cookie('91porn.com', 'language', 'cn_CN') self._set_cookie('91porn.com', 'language', 'cn_CN')
webpage = self._download_webpage( webpage = self._download_webpage(
@ -35,29 +48,53 @@ class Porn91IE(InfoExtractor):
if '作为游客你每天只可观看10个视频' in webpage: if '作为游客你每天只可观看10个视频' in webpage:
raise ExtractorError('91 Porn says: Daily limit 10 videos exceeded', expected=True) raise ExtractorError('91 Porn says: Daily limit 10 videos exceeded', expected=True)
title = self._search_regex( title = self._html_search_regex(
r'<div id="viewvideo-title">([^<]+)</div>', webpage, 'title') r'(?s)<title\b[^>]*>\s*(.+?)\s*(?:Chinese\s+homemade\s+video\s*)?</title', webpage, 'title')
title = title.replace('\n', '')
video_link_url = self._search_regex( video_link_url = self._search_regex(
r'<textarea[^>]+id=["\']fm-video_link[^>]+>([^<]+)</textarea>', r'''document\s*\.\s*write\s*\(\s*strencode2\s*\((?P<q>"|')(?P<enc_str>[%\da-fA-F]+)(?P=q)\s*\)''',
webpage, 'video link') webpage, 'video link', group='enc_str')
videopage = self._download_webpage(video_link_url, video_id) video_link_url = compat_urllib_parse_unquote(video_link_url)
info_dict = self._parse_html5_media_entries(url, videopage, video_id)[0] info_dict = self._parse_html5_media_entries(url, '<video>%s</video>' % (video_link_url, ), video_id)[0]
duration = parse_duration(self._search_regex( # extract various fields in <tag class=info>Name: value</tag>
r'时长:\s*</span>\s*(\d+:\d+)', webpage, 'duration', fatal=False)) FIELD_MAP = {
# cn_CN name: (yt-dl key, value parser, en name)
'时长': ('duration', parse_duration, 'Runtime', ),
'查看': ('view_count', str_to_int, 'Views', ),
'留言': ('comment_count', str_to_int, 'Comments', ),
'收藏': ('like_count', str_to_int, 'Favorites', ),
'添加时间': ('upload_date', unified_strdate, 'Added', ),
# same as title for en, not description for cn_CN
'__ignore__': ('description', strip_or_none, 'Description', ),
'作者': ('uploader', strip_or_none, 'From', ),
}
# yt-dl's original implementation of get_elements_by_class() uses regex
# yt-dlp uses an actual HTML parser, and can be confused by bad HTML fragments
for elt in get_elements_by_class(
'info',
# concatenate <span>s ...
re.sub(r'(?i)</span>\s*<span\b[^>]*?>', '',
# ... and strip out possibly unbalanced <font> for yt-dlp
re.sub(r'(?i)(?:<font\b[^>]*?>|</font\s*>)', '', webpage))) or []:
elt = re.split(r':\s*', clean_html(elt), 1)
if len(elt) != 2 or elt[1] == '':
continue
parm = FIELD_MAP.get(elt[0].strip())
if parm and elt[1] is not None:
info_dict[parm[0]] = parm[1](elt[1]) if parm[1] else elt[1]
comment_count = int_or_none(self._search_regex( thumbnail = extract_attributes(
r'留言:\s*</span>\s*(\d+)', webpage, 'comment count', fatal=False)) self._search_regex(
r'''(<video\b[^>]*\bid\s*=\s*(?P<q>"|')player_one(?P=q)[^>]*>)''',
webpage, 'poster', default='')).get('poster')
info_dict.update({ info_dict.update({
'id': video_id, 'id': video_id,
'title': title, 'title': title,
'duration': duration,
'comment_count': comment_count,
'age_limit': self._rta_search(webpage), 'age_limit': self._rta_search(webpage),
'thumbnail': thumbnail,
}) })
return info_dict return info_dict

View File

@ -1659,17 +1659,46 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
def _extract_n_function_name(self, jscode): def _extract_n_function_name(self, jscode):
func_name, idx = self._search_regex( func_name, idx = self._search_regex(
# new: (b=String.fromCharCode(110),c=a.get(b))&&c=nfunc[idx](c) # new: (b=String.fromCharCode(110),c=a.get(b))&&c=nfunc[idx](c)
# or: (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)s # or: (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)
# old: .get("n"))&&(b=nfunc[idx](b) # or: (PL(a),b=a.j.n||null)&&(b=nfunc[idx](b)
# older: .get("n"))&&(b=nfunc(b) # or: (b="nn"[+a.D],vL(a),c=a.j[b]||null)&&(c=narray[idx](c),a.set(b,c),narray.length||nfunc("")
# old: (b=a.get("n"))&&(b=nfunc[idx](b)(?P<c>[a-z])\s*=\s*[a-z]\s*
# older: (b=a.get("n"))&&(b=nfunc(b)
r'''(?x) r'''(?x)
(?:\(\s*(?P<b>[a-z])\s*=\s*(?: \((?:[\w$()\s]+,)*?\s* # (
(?P<b>[a-z])\s*=\s* # b=
(?:
(?: # expect ,c=a.get(b) (etc)
String\s*\.\s*fromCharCode\s*\(\s*110\s*\)| String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
"n+"\[\s*\+?s*[\w$.]+\s*] "n+"\[\s*\+?s*[\w$.]+\s*]
)\s*,(?P<c>[a-z])\s*=\s*[a-z]\s*)? )\s*(?:,[\w$()\s]+(?=,))*|
\.\s*get\s*\(\s*(?(b)(?P=b)|"n{1,2}")(?:\s*\)){2}\s*&&\s*\(\s*(?(c)(?P=c)|b)\s*=\s* (?P<old>[\w$]+) # a (old[er])
)\s*
(?(old)
# b.get("n")
(?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
(?:\.\s*n|\[\s*"n"\s*]|\.\s*get\s*\(\s*"n"\s*\))
| # ,c=a.get(b)
,\s*(?P<c>[a-z])\s*=\s*[a-z]\s*
(?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
(?:\[\s*(?P=b)\s*]|\.\s*get\s*\(\s*(?P=b)\s*\))
)
# interstitial junk
\s*(?:\|\|\s*null\s*)?(?:\)\s*)?&&\s*(?:\(\s*)?
(?(c)(?P=c)|(?P=b))\s*=\s* # [c|b]=
# nfunc|nfunc[idx]
(?P<nfunc>[a-zA-Z_$][\w$]*)(?:\s*\[(?P<idx>\d+)\])?\s*\(\s*[\w$]+\s*\) (?P<nfunc>[a-zA-Z_$][\w$]*)(?:\s*\[(?P<idx>\d+)\])?\s*\(\s*[\w$]+\s*\)
''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx')) ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'),
default=(None, None))
# thx bashonly: yt-dlp/yt-dlp/pull/10611
if not func_name:
self.report_warning('Falling back to generic n function search')
return self._search_regex(
r'''(?xs)
(?:(?<=[^\w$])|^) # instead of \b, which ignores $
(?P<name>(?!\d)[a-zA-Z\d_$]+)\s*=\s*function\((?!\d)[a-zA-Z\d_$]+\)
\s*\{(?:(?!};).)+?["']enhanced_except_
''', jscode, 'Initial JS player n function name', group='name')
if not idx: if not idx:
return func_name return func_name

View File

@ -925,9 +925,16 @@ class JSInterpreter(object):
obj.reverse() obj.reverse()
return obj return obj
elif member == 'slice': elif member == 'slice':
assertion(isinstance(obj, list), 'must be applied on a list') assertion(isinstance(obj, (list, compat_str)), 'must be applied on a list or string')
assertion(len(argvals) == 1, 'takes exactly one argument') # From [1]:
return obj[argvals[0]:] # .slice() - like [:]
# .slice(n) - like [n:] (not [slice(n)]
# .slice(m, n) - like [m:n] or [slice(m, n)]
# [1] https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/slice
assertion(len(argvals) <= 2, 'takes between 0 and 2 arguments')
if len(argvals) < 2:
argvals += (None,)
return obj[slice(*argvals)]
elif member == 'splice': elif member == 'splice':
assertion(isinstance(obj, list), 'must be applied on a list') assertion(isinstance(obj, list), 'must be applied on a list')
assertion(argvals, 'takes one or more arguments') assertion(argvals, 'takes one or more arguments')