Merge 3091daa834 into c5098961b0

[Youtube] Rework n function extraction pattern
Now also succeeds with player b12cc44b
2024-12-19 07:02:06 +00:00 · 2024-08-21 22:32:46 -04:00 · 2024-08-06 20:59:09 +01:00 · 2024-08-06 20:51:38 +01:00 · 2024-08-01 19:18:34 +01:00 · 2022-07-21 21:18:56 +01:00
5 changed files with 134 additions and 55 deletions
--- a/test/test_jsinterp.py
+++ b/test/test_jsinterp.py
@ -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()
--- a/test/test_youtube_signature.py
+++ b/test/test_youtube_signature.py
@ -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',
+    ),
 ]


--- a/youtube_dl/extractor/newgrounds.py
+++ b/youtube_dl/extractor/newgrounds.py
@ -5,7 +5,6 @@ import re
 from .common import InfoExtractor
 from ..utils import (
    extract_attributes,
-    int_or_none,
    parse_duration,
    parse_filesize,
    unified_timestamp,
@ -27,15 +26,15 @@ class NewgroundsIE(InfoExtractor):
            'duration': 143,
        },
    }, {
-        'url': 'https://www.newgrounds.com/portal/view/673111',
-        'md5': '3394735822aab2478c31b1004fe5e5bc',
+        'url': 'https://www.newgrounds.com/portal/view/850292',
+        'md5': 'bb7cacf45e1b4d648e2dac2d79284d67',
        'info_dict': {
-            'id': '673111',
+            'id': '850292',
            'ext': 'mp4',
-            'title': 'Dancin',
-            'uploader': 'Squirrelman82',
-            'timestamp': 1460256780,
-            'upload_date': '20160410',
+            'title': 'Timeless (2021)',
+            'uploader': 'Kevuhn',
+            'timestamp': 1657896960,
+            'upload_date': '20220715',
        },
    }, {
        # source format unavailable, additional mp4 formats
@ -44,7 +43,7 @@ class NewgroundsIE(InfoExtractor):
            'id': '689400',
            'ext': 'mp4',
            'title': 'ZTV News Episode 8',
-            'uploader': 'BennettTheSage',
+            'uploader': 'ZONE-SAMA',
            'timestamp': 1487965140,
            'upload_date': '20170224',
        },
@ -58,40 +57,46 @@ class NewgroundsIE(InfoExtractor):

        webpage = self._download_webpage(url, media_id)

+        paths = url.split('/')
+        if paths[-3] == 'audio':
+            isAudio = True
+        else:
+            isAudio = False
+
+        if isAudio:
+            media_url = self._parse_json(self._search_regex(
+                r'"url"\s*:\s*("[^"]+"),', webpage, ''), media_id)
+
+            uploader = self._html_search_regex(
+                (r'(?s)<h4[^>]*>(.+?)</h4>.*?<em>\s*Author\s*</em>',
+                 r'(?:Author|Writer)\s*<a[^>]+>([^<]+)'), webpage, 'uploader',
+                fatal=False)
+
+            formats = [{
+                'url': media_url,
+                'format_id': 'source',
+                'quality': 1,
+            }]
+
+        else:
+            media_url = 'https://www.newgrounds.com/portal/video/' + media_id
+            media = self._download_json(media_url, media_id, headers={'X-Requested-With': 'XMLHttpRequest'})
+
+            uploader = media['author']
+
+            sources = media['sources']
+
+            formats = []
+            for source in sources:
+                for i in range(len(sources[source])):
+                    formats.append({
+                        'url': sources[source][i]['src'],
+                        'format_id': source,
+                        'height': int(source[:-2])  # 1080p -> 1080
+                    })
+
        title = self._html_search_regex(
            r'<title>([^>]+)</title>', webpage, 'title')
-
-        media_url = self._parse_json(self._search_regex(
-            r'"url"\s*:\s*("[^"]+"),', webpage, ''), media_id)
-
-        formats = [{
-            'url': media_url,
-            'format_id': 'source',
-            'quality': 1,
-        }]
-
-        max_resolution = int_or_none(self._search_regex(
-            r'max_resolution["\']\s*:\s*(\d+)', webpage, 'max resolution',
-            default=None))
-        if max_resolution:
-            url_base = media_url.rpartition('.')[0]
-            for resolution in (360, 720, 1080):
-                if resolution > max_resolution:
-                    break
-                formats.append({
-                    'url': '%s.%dp.mp4' % (url_base, resolution),
-                    'format_id': '%dp' % resolution,
-                    'height': resolution,
-                })
-
-        self._check_formats(formats, media_id)
-        self._sort_formats(formats)
-
-        uploader = self._html_search_regex(
-            (r'(?s)<h4[^>]*>(.+?)</h4>.*?<em>\s*Author\s*</em>',
-             r'(?:Author|Writer)\s*<a[^>]+>([^<]+)'), webpage, 'uploader',
-            fatal=False)
-
        timestamp = unified_timestamp(self._html_search_regex(
            (r'<dt>\s*Uploaded\s*</dt>\s*<dd>([^<]+</dd>\s*<dd>[^<]+)',
             r'<dt>\s*Uploaded\s*</dt>\s*<dd>([^<]+)'), webpage, 'timestamp',
@ -99,7 +104,6 @@ class NewgroundsIE(InfoExtractor):
        duration = parse_duration(self._search_regex(
            r'(?s)<dd>\s*Song\s*</dd>\s*<dd>.+?</dd>\s*<dd>([^<]+)', webpage,
            'duration', default=None))
-
        filesize_approx = parse_filesize(self._html_search_regex(
            r'(?s)<dd>\s*Song\s*</dd>\s*<dd>(.+?)</dd>', webpage, 'filesize',
            default=None))
@ -109,6 +113,9 @@ class NewgroundsIE(InfoExtractor):
        if '<dd>Song' in webpage:
            formats[0]['vcodec'] = 'none'

+        self._check_formats(formats, media_id)
+        self._sort_formats(formats)
+
        return {
            'id': media_id,
            'title': title,
--- a/youtube_dl/extractor/youtube.py
+++ b/youtube_dl/extractor/youtube.py
@ -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

--- a/youtube_dl/jsinterp.py
+++ b/youtube_dl/jsinterp.py
@ -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')
Author	SHA1	Message	Date
Niyazi Suleymanov	358aabf50b	Merge `3091daa834` 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	3091daa834	Update youtube_dl/extractor/newgrounds.py - remove f-string	2022-07-21 21:18:56 +01:00
Niyazi Suleymanov	c848538aff	follow coding conventions	2022-07-17 01:51:53 +04:00
Niyazi Suleymanov	f6c4b9da47	fix #30575	2022-07-17 01:40:44 +04:00
Niyazi Suleymanov	45ae186416	change incorrect uploader	2022-07-17 01:20:01 +04:00
Niyazi Suleymanov	9036cf5132	add new test	2022-07-17 01:19:12 +04:00
Niyazi Suleymanov	41f5fd818b	entry no longer exists	2022-07-16 22:52:19 +04:00