Compare commits

...

12 Commits

Author SHA1 Message Date
Daenges
774e5d2b9b
Merge 4232427a8d into c5098961b0 2024-08-21 22:32:46 -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
4232427a8d
Allow numbers and upper-case letters in ID 2023-07-05 20:51:58 +01:00
dirkf
2804216e81
Improve fallback 2022-07-22 15:54:30 +01:00
dirkf
e8aca87278
Improve fallback 2022-07-22 15:48:26 +01:00
Daenges
cf4a829c13 Implement _search_json_ld() 2022-05-23 19:21:56 +02:00
Daenges
73771cd768 Commit suggested changes.
- Verify description through md5
- Implement robust detection of description
- Remove format attribute to allow auto detection
- Allow conditioning of URLs
2022-05-21 16:39:46 +02:00
Daenges
041abb92b3 Apply coding conventions 2022-05-16 22:32:47 +02:00
Daenges
2b128c729e Add further comments 2022-05-16 21:37:35 +02:00
Daenges
3bffd20f33 Add extractor for MegaCartoons 2022-05-16 21:28:19 +02:00
6 changed files with 196 additions and 13 deletions

View File

@ -425,6 +425,34 @@ class TestJSInterpreter(unittest.TestCase):
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__':
unittest.main()

View File

@ -174,6 +174,14 @@ _NSIG_TESTS = [
'https://www.youtube.com/s/player/5604538d/player_ias.vflset/en_US/base.js',
'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

@ -707,6 +707,7 @@ from .mixcloud import (
MixcloudUserIE,
MixcloudPlaylistIE,
)
from .megacartoons import MegaCartoonsIE
from .mlb import (
MLBIE,
MLBVideoIE,

View File

@ -0,0 +1,110 @@
# coding: utf-8
from __future__ import unicode_literals
import re
from ..utils import (
bug_reports_message,
JSON_LD_RE,
merge_dicts,
NO_DEFAULT,
RegexNotFoundError,
try_get,
url_or_none,
)
from ..compat import compat_str
from .common import InfoExtractor
class MegaCartoonsIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?megacartoons\.net/(?P<id>[a-zA-Z\d-]+)/'
_TESTS = [{
'url': 'https://www.megacartoons.net/help-wanted/',
'md5': '4ba9be574f9a17abe0c074e2f955fded',
'info_dict': {
'id': 'help-wanted',
'ext': 'mp4',
'title': 'Help Wanted - SpongeBob SquarePants',
'upload_date': '20200223',
'timestamp': 1582416000,
'thumbnail': r're:^https?://.*\.jpg$',
'description': 'md5:2c909daa6c6cb16b2d4d791dd1a31632'
}
}, {
'url': 'https://www.megacartoons.net/1000-years-of-courage/',
'only_matching': True,
}, {
'url': 'https://www.megacartoons.net/911-2/',
'only_matching': True,
}]
# adapted from common.py pending yt-dlp back-port
def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
json_ld_list = list(re.finditer(JSON_LD_RE, html))
default = kwargs.get('default', NO_DEFAULT)
fatal = kwargs.get('fatal', True) if default is NO_DEFAULT else False
json_ld = []
for mobj in json_ld_list:
json_ld_item = self._parse_json(
mobj.group('json_ld'), video_id, fatal=fatal)
if not json_ld_item:
continue
if isinstance(json_ld_item, dict):
json_ld.append(json_ld_item)
elif isinstance(json_ld_item, (list, tuple)):
json_ld.extend(json_ld_item)
if json_ld:
# handle initial '@graph' with one level of children
if len(json_ld) > 0 and '@graph' in json_ld[0] and '@context' in json_ld[0]:
# should always be hit here
context = json_ld[0]['@context']
json_ld_g = json_ld[0]['@graph'] or []
for item in json_ld_g:
item.setdefault('@context', context)
json_ld = json_ld_g + json_ld[1:]
json_ld = self._json_ld(json_ld, video_id, fatal=fatal, expected_type=expected_type)
if json_ld:
return json_ld
if default is not NO_DEFAULT:
return default
elif fatal:
raise RegexNotFoundError('Unable to extract JSON-LD')
else:
self.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
return {}
def _real_extract(self, url):
# ID is equal to the episode name
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
info = self._search_json_ld(webpage, video_id, fatal=False) or {}
info.update({
'id': video_id,
# Try to find a good title or fallback to the ID
'title': info.get('title') or self._og_search_title(webpage) or video_id.replace('-', ' ').capitalize(),
})
if 'url' not in info or 'thumbnail' not in info:
# Video data is stored in a json -> extract it from the raw html
url_json = self._parse_json(self._html_search_regex(r'''<div\b[^>]+\bdata-item\s*=\s*(["'])(?P<videourls>\{.*})\1''', webpage, 'videourls', group='videourls', default='{}'), video_id, fatal=False) or {}
video_url = url_or_none(try_get(url_json, lambda x: x['sources'][0]['src'], compat_str) or self._og_search_video_url(webpage)) # Get the video url
video_thumbnail = url_or_none(url_json.get('splash') or self._og_search_thumbnail(webpage)) # Get the thumbnail
info = merge_dicts(info, {
'url': video_url,
'thumbnail': video_thumbnail,
})
# Find the <article> class in the html
article = self._search_regex(
r'(?s)<article\b[^>]*?\bclass\s*=\s*[^>]*?\bpost\b[^>]*>(.+?)</article\b', webpage, 'post', default='')
# Extract the actual description from it
info['description'] = (
self._html_search_regex(r'(?s)<p>\s*([^<]+)\s*</p>', article, 'videodescription', fatal=False)
or self._og_search_description(webpage))
return info

View File

@ -1659,17 +1659,46 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
def _extract_n_function_name(self, jscode):
func_name, idx = self._search_regex(
# 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
# old: .get("n"))&&(b=nfunc[idx](b)
# older: .get("n"))&&(b=nfunc(b)
# or: (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)
# or: (PL(a),b=a.j.n||null)&&(b=nfunc[idx](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)
(?:\(\s*(?P<b>[a-z])\s*=\s*(?:
String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
"n+"\[\s*\+?s*[\w$.]+\s*]
)\s*,(?P<c>[a-z])\s*=\s*[a-z]\s*)?
\.\s*get\s*\(\s*(?(b)(?P=b)|"n{1,2}")(?:\s*\)){2}\s*&&\s*\(\s*(?(c)(?P=c)|b)\s*=\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'))
\((?:[\w$()\s]+,)*?\s* # (
(?P<b>[a-z])\s*=\s* # b=
(?:
(?: # expect ,c=a.get(b) (etc)
String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
"n+"\[\s*\+?s*[\w$.]+\s*]
)\s*(?:,[\w$()\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*\(\s*[\w$]+\s*\)
''', 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:
return func_name

View File

@ -925,9 +925,16 @@ class JSInterpreter(object):
obj.reverse()
return obj
elif member == 'slice':
assertion(isinstance(obj, list), 'must be applied on a list')
assertion(len(argvals) == 1, 'takes exactly one argument')
return obj[argvals[0]:]
assertion(isinstance(obj, (list, compat_str)), 'must be applied on a list or string')
# From [1]:
# .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':
assertion(isinstance(obj, list), 'must be applied on a list')
assertion(argvals, 'takes one or more arguments')