Merge a571bbaab1 into c5098961b0

[Youtube] Rework n function extraction pattern
Now also succeeds with player b12cc44b
2025-01-10 21:40:11 +09:00 · 2024-08-21 22:33:25 -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 · 2021-02-02 09:58:49 +01:00
7 changed files with 260 additions and 13 deletions
--- a/docs/supportedsites.md
+++ b/docs/supportedsites.md
@ -613,6 +613,7 @@
 - **niconico**: ニコニコ動画
 - **NiconicoPlaylist**
 - **Nintendo**
+ - **Nitter**
 - **njoy**: N-JOY
 - **njoy:embed**
 - **NJPWWorld**: 新日本プロレスワールド
--- 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/extractors.py
+++ b/youtube_dl/extractor/extractors.py
@ -830,6 +830,7 @@ from .ninecninemedia import NineCNineMediaIE
 from .ninegag import NineGagIE
 from .ninenow import NineNowIE
 from .nintendo import NintendoIE
+from .nitter import NitterIE
 from .njpwworld import NJPWWorldIE
 from .nobelprize import NobelPrizeIE
 from .nonktube import NonkTubeIE
--- a/youtube_dl/extractor/nitter.py
+++ b/youtube_dl/extractor/nitter.py
@ -0,0 +1,173 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..compat import compat_urlparse
+from ..utils import (
+    parse_count,
+    unified_strdate,
+    unified_timestamp,
+    remove_end,
+    determine_ext,
+)
+import re
+
+
+class NitterIE(InfoExtractor):
+    # Taken from https://github.com/zedeus/nitter/wiki/Instances
+    INSTANCES = ('nitter.net',
+                 'nitter.snopyta.org',
+                 'nitter.42l.fr',
+                 'nitter.nixnet.services',
+                 'nitter.13ad.de',
+                 'nitter.pussthecat.org',
+                 'nitter.mastodont.cat',
+                 'nitter.dark.fail',
+                 'nitter.tedomum.net',
+                 'nitter.cattube.org',
+                 'nitter.fdn.fr',
+                 'nitter.1d4.us',
+                 'nitter.kavin.rocks',
+                 'tweet.lambda.dance',
+                 'nitter.cc',
+                 'nitter.weaponizedhumiliation.com',
+                 'nitter.vxempire.xyz',
+                 'nitter.unixfox.eu',
+                 'nitter.domain.glass',
+                 'nitter.himiko.cloud',
+                 'nitter.eu',
+                 'nitter.ethibox.fr',
+                 '3nzoldnxplag42gqjs23xvghtzf6t6yzssrtytnntc6ppc7xxuoneoad.onion',
+                 'nitter.l4qlywnpwqsluw65ts7md3khrivpirse744un3x7mlskqauz5pyuzgqd.onion',
+                 'npf37k3mtzwxreiw52ccs5ay4e6qt2fkcs2ndieurdyn2cuzzsfyfvid.onion')
+
+    _INSTANCES_RE = '(?:' + '|'.join([re.escape(instance) for instance in INSTANCES]) + ')'
+    _VALID_URL = r'https?://%(instance)s/(?P<uploader_id>.+)/status/(?P<id>[0-9]+)(#.)?' % {'instance': _INSTANCES_RE}
+    current_instance = INSTANCES[0]  # the test and official instance
+    _TESTS = [
+        {
+            # GIF (wrapped in mp4)
+            'url': 'https://' + current_instance + '/firefox/status/1314279897502629888#m',
+            'info_dict': {
+                'id': '1314279897502629888',
+                'ext': 'mp4',
+                'title': 'Firefox 🔥 - You know the old saying, if you see something say something. Now you actually can with the YouTube regrets extension.   Report harmful YouTube recommendations so others can avoid watching them. ➡️ https://mzl.la/3iFIiyg  #UnfckTheInternet',
+                'description': 'You know the old saying, if you see something say something. Now you actually can with the YouTube regrets extension.   Report harmful YouTube recommendations so others can avoid watching them. ➡️ https://mzl.la/3iFIiyg  #UnfckTheInternet',
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'uploader': 'Firefox 🔥',
+                'uploader_id': 'firefox',
+                'uploader_url': 'https://' + current_instance + '/firefox',
+                'upload_date': '20201008',
+                'timestamp': 1602183720,
+            },
+        }, {  # normal video
+            'url': 'https://' + current_instance + '/Le___Doc/status/1299715685392756737#m',
+            'info_dict': {
+                'id': '1299715685392756737',
+                'ext': 'mp4',
+                'title': 'Le Doc - "Je ne prédis jamais rien" D Raoult, Août 2020...',
+                'description': '"Je ne prédis jamais rien" D Raoult, Août 2020...',
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'uploader': 'Le Doc',
+                'uploader_id': 'Le___Doc',
+                'uploader_url': 'https://' + current_instance + '/Le___Doc',
+                'upload_date': '20200829',
+                'timestamp': 1598711341,
+                'view_count': int,
+                'like_count': int,
+                'repost_count': int,
+                'comment_count': int,
+            },
+        }, {  # video embed in a "Streaming Political Ads" box
+            'url': 'https://' + current_instance + '/mozilla/status/1321147074491092994#m',
+            'info_dict': {
+                'id': '1321147074491092994',
+                'ext': 'mp4',
+                'title': "Mozilla - Are you being targeted with weird, ominous or just plain annoying political ads while streaming your favorite shows?  This isn't a real political ad, but if you're watching streaming TV in the U.S., chances are you've seen quite a few.   Learn more ➡️ https://mzl.la/StreamingAds",
+                'description': "Are you being targeted with weird, ominous or just plain annoying political ads while streaming your favorite shows?  This isn't a real political ad, but if you're watching streaming TV in the U.S., chances are you've seen quite a few.   Learn more ➡️ https://mzl.la/StreamingAds",
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'uploader': 'Mozilla',
+                'uploader_id': 'mozilla',
+                'uploader_url': 'https://' + current_instance + '/mozilla',
+                'upload_date': '20201027',
+                'timestamp': 1603820982
+            },
+        },
+    ]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        parsed_url = compat_urlparse.urlparse(url)
+        base_url = parsed_url.scheme + '://' + parsed_url.netloc
+
+        self._set_cookie(parsed_url.netloc, 'hlsPlayback', 'on')
+        webpage = self._download_webpage(url, video_id)
+
+        video_url = base_url + self._html_search_regex(r'(?:<video[^>]+data-url|<source[^>]+src)="([^"]+)"', webpage, 'video url')
+        ext = determine_ext(video_url)
+
+        if ext == 'unknown_video':
+            formats = self._extract_m3u8_formats(video_url, video_id, ext='mp4')
+        else:
+            formats = [{
+                'url': video_url,
+                'ext': ext
+            }]
+
+        title = (
+            self._og_search_description(webpage).replace('\n', ' ')
+            or self._html_search_regex(r'<div class="tweet-content[^>]+>([^<]+)</div>', webpage, 'title'))
+        description = title
+
+        mobj = re.match(self._VALID_URL, url)
+        uploader_id = (
+            mobj.group('uploader_id')
+            or self._html_search_regex(r'<a class="fullname"[^>]+title="([^"]+)"', webpage, 'uploader name', fatal=False))
+
+        if uploader_id:
+            uploader_url = base_url + '/' + uploader_id
+
+        uploader = self._html_search_regex(r'<a class="fullname"[^>]+title="([^"]+)"', webpage, 'uploader name', fatal=False)
+
+        if uploader:
+            title = uploader + ' - ' + title
+
+        view_count = parse_count(self._html_search_regex(r'<span[^>]+class="icon-play[^>]*></span>\s([^<]+)</div>', webpage, 'view count', fatal=False))
+        like_count = parse_count(self._html_search_regex(r'<span[^>]+class="icon-heart[^>]*></span>\s([^<]+)</div>', webpage, 'like count', fatal=False))
+        repost_count = parse_count(self._html_search_regex(r'<span[^>]+class="icon-retweet[^>]*></span>\s([^<]+)</div>', webpage, 'repost count', fatal=False))
+        comment_count = parse_count(self._html_search_regex(r'<span[^>]+class="icon-comment[^>]*></span>\s([^<]+)</div>', webpage, 'repost count', fatal=False))
+
+        thumbnail = base_url + (self._html_search_meta('og:image', webpage, 'thumbnail url')
+                                or self._html_search_regex(r'<video[^>]+poster="([^"]+)"', webpage, 'thumbnail url', fatal=False))
+
+        thumbnail = remove_end(thumbnail, '%3Asmall')  # if parsed with regex, it should contain this
+
+        thumbnails = []
+        thumbnail_ids = ('thumb', 'small', 'large', 'medium', 'orig')
+        for id in thumbnail_ids:
+            thumbnails.append({
+                'id': id,
+                'url': thumbnail + '%3A' + id,
+            })
+
+        date = self._html_search_regex(r'<span[^>]+class="tweet-date"[^>]*><a[^>]+title="([^"]+)"', webpage, 'upload date', fatal=False)
+        upload_date = unified_strdate(date)
+        timestamp = unified_timestamp(date)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': description,
+            'uploader': uploader,
+            'timestamp': timestamp,
+            'uploader_id': uploader_id,
+            'uploader_url': uploader_url,
+            'view_count': view_count,
+            'like_count': like_count,
+            'repost_count': repost_count,
+            'comment_count': comment_count,
+            'formats': formats,
+            'thumbnails': thumbnails,
+            'thumbnail': thumbnail,
+            'upload_date': upload_date,
+        }
--- 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
bopol	567af82129	Merge `a571bbaab1` into `c5098961b0`	2024-08-21 22:33:25 -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
bopol	a571bbaab1	[nitter] Add new extractor	2021-02-02 09:58:49 +01:00