Merge a1ef0a66f3 into c5098961b0

[Youtube] Rework n function extraction pattern
Now also succeeds with player b12cc44b
2025-01-25 04:40:09 +09:00 · 2024-08-21 22:21:01 -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 · 2021-02-06 11:31:44 +01:00
6 changed files with 170 additions and 19 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_subtitles.py
+++ b/test/test_subtitles.py
@ -27,6 +27,7 @@ from youtube_dl.extractor import (
    ThePlatformFeedIE,
    RTVEALaCartaIE,
    DemocracynowIE,
+    RoosterTeethIE,
 )


@ -395,5 +396,21 @@ class TestDemocracynowSubtitles(BaseTestSubtitles):
        self.assertEqual(md5(subtitles['en']), 'a3cc4c0b5eadd74d9974f1c1f5101045')


+class TestRoosterTeethSubtitles(BaseTestSubtitles):
+    url = 'https://www.roosterteeth.com/watch/rwby-season-1-episode-1'
+    IE = RoosterTeethIE
+
+    def test_allsubtitles(self):
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        self.assertEqual(set(subtitles.keys()), set(['pt', 'de', 'fr', 'es', 'en']))
+        self.assertEqual(md5(subtitles['pt']), '96490f6884378403b4304b4355ddd028')
+        self.assertEqual(md5(subtitles['de']), 'a30fbfbc2574530457d12fcaf68b515c')
+        self.assertEqual(md5(subtitles['fr']), '64ff89d6a4dd8aa079f680d1cb799fde')
+        self.assertEqual(md5(subtitles['es']), '565a9b49173539ce5a3de9756bd3e3a2')
+        self.assertEqual(md5(subtitles['en']), '404252b16a423c3b89d3c8774445df65')
+
+
 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/roosterteeth.py
+++ b/youtube_dl/extractor/roosterteeth.py
@ -11,6 +11,10 @@ from ..utils import (
    int_or_none,
    str_or_none,
    urlencode_postdata,
+    parse_m3u8_attributes,
+    try_get,
+    url_or_none,
+    urljoin,
 )


@ -86,9 +90,13 @@ class RoosterTeethIE(InfoExtractor):
        api_episode_url = self._EPISODE_BASE_URL + display_id

        try:
-            m3u8_url = self._download_json(
-                api_episode_url + '/videos', display_id,
-                'Downloading video JSON metadata')['data'][0]['attributes']['url']
+            video_json = self._download_json(
+                api_episode_url + '/videos', display_id)['data'][0]
+            m3u8_url = url_or_none(try_get(
+                video_json, [
+                    lambda j: j['attributes']['url'],
+                    lambda j: j['links']['master']],
+                compat_str))
        except ExtractorError as e:
            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
                if self._parse_json(e.cause.read().decode(), display_id).get('access') is False:
@ -96,9 +104,15 @@ class RoosterTeethIE(InfoExtractor):
                        '%s is only available for FIRST members' % display_id)
            raise

-        formats = self._extract_m3u8_formats(
-            m3u8_url, display_id, 'mp4', 'm3u8_native', m3u8_id='hls')
-        self._sort_formats(formats)
+        if m3u8_url:
+            formats = self._extract_m3u8_formats(
+                m3u8_url, display_id, 'mp4', 'm3u8_native', m3u8_id='hls')
+            self._sort_formats(formats)
+
+            subtitles = self._extract_m3u8_subtitles(m3u8_url, display_id)
+        else:
+            formats = []
+            subtitles = None

        episode = self._download_json(
            api_episode_url, display_id,
@ -133,5 +147,53 @@ class RoosterTeethIE(InfoExtractor):
            'episode_id': str_or_none(episode.get('uuid')),
            'formats': formats,
            'channel_id': attributes.get('channel_id'),
+            'subtitles': subtitles,
            'duration': int_or_none(attributes.get('length')),
        }
+
+    def _extract_m3u8_subtitles(self, m3u8_url, video_id):
+        res = self._download_webpage_handle(
+            m3u8_url, video_id,
+            note='Downloading subtitle information',
+            errnote='Failed to download subtitle information',
+            fatal=False, data=None, headers={}, query={})
+        if res is False:
+            return None
+
+        m3u8_doc, urlh = res
+        m3u8_url = urlh.geturl()
+
+        subtitles = {}
+        for line in m3u8_doc.splitlines():
+            if not line.startswith("#EXT-X-MEDIA:"):
+                continue
+            media = parse_m3u8_attributes(line)
+
+            media_type, media_url_raw, media_lang = (
+                media.get('TYPE'), media.get('URI'), media.get('LANGUAGE'),)
+            if not (media_type in ('SUBTITLES',) and media_url_raw and media_lang):
+                continue
+
+            media_url = urljoin(m3u8_url, media_url_raw)
+            if not media_url:
+                continue
+
+            res = self._download_webpage_handle(
+                media_url, video_id,
+                note='Downloading subtitle information ({})'.format(media_lang),
+                errnote='Failed to download subtitle information ({})'.format(media_lang),
+                fatal=False, data=None, headers={}, query={})
+            if res is False:
+                continue
+
+            m3u8_subtitle_doc, _ = res
+            subtitle_url = None
+            for subtitle_line in m3u8_subtitle_doc.splitlines():
+                if subtitle_line.startswith("#"):
+                    continue
+                subtitle_url = urljoin(media_url, subtitle_line)
+                break
+
+            if subtitle_url:
+                subtitles[compat_str(media_lang)] = [{'url': subtitle_url, }, ]
+        return subtitles if len(subtitles) > 0 else None
--- 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
wtpckl	59c3286ac9	Merge `a1ef0a66f3` into `c5098961b0`	2024-08-21 22:21:01 -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
wtpckl	a1ef0a66f3	[RoosterTeeth] Handle missing media URL at higher level	2021-02-06 11:31:44 +01:00
wtpckl	1ed79ce6a1	[RoosterTeeth] flake8	2021-01-28 17:44:37 +01:00
wtpckl	0373c0fa89	[RoosterTeeth] Add unittest for subtitle extractor	2021-01-28 17:22:53 +01:00
wtpckl	b4bc4527e4	[RoosterTeeth] Clean up code after adding subtitle support	2021-01-28 17:09:54 +01:00
wtpckl	759e8ce15b	[RoosterTeeth] Add subtitle support to extractor	2021-01-28 16:20:28 +01:00