Merge fad516e22e58763ce0feab1034cc0c7aa2da184c into da7223d4aa42ff9fc680b0951d043dd03cec2d30

* improve extraction of global "useful data" Array from player JS
* also handle tv-player and add tests: thx seproDev (yt-dlp/yt-dlp#12684)

Co-Authored-By: sepro <sepro@sepr0.com>

test/test_youtube_signature.py

@@ -232,8 +232,32 @@ _NSIG_TESTS = [
         'W9HJZKktxuYoDTqW', 'jHbbkcaxm54',
     ),
     (
-        'https://www.youtube.com/s/player/91201489/player_ias_tce.vflset/en_US/base.js',
+        'https://www.youtube.com/s/player/643afba4/player_ias.vflset/en_US/base.js',
-        'W9HJZKktxuYoDTqW', 'U48vOZHaeYS6vO',
+        'W9HJZKktxuYoDTqW', 'larxUlagTRAcSw',
+    ),
+    (
+        'https://www.youtube.com/s/player/e7567ecf/player_ias_tce.vflset/en_US/base.js',
+        'Sy4aDGc0VpYRR9ew_', '5UPOT1VhoZxNLQ',
+    ),
+    (
+        'https://www.youtube.com/s/player/d50f54ef/player_ias_tce.vflset/en_US/base.js',
+        'Ha7507LzRmH3Utygtj', 'XFTb2HoeOE5MHg',
+    ),
+    (
+        'https://www.youtube.com/s/player/074a8365/player_ias_tce.vflset/en_US/base.js',
+        'Ha7507LzRmH3Utygtj', 'ufTsrE0IVYrkl8v',
+    ),
+    (
+        'https://www.youtube.com/s/player/643afba4/player_ias.vflset/en_US/base.js',
+        'N5uAlLqm0eg1GyHO', 'dCBQOejdq5s-ww',
+    ),
+    (
+        'https://www.youtube.com/s/player/69f581a5/tv-player-ias.vflset/tv-player-ias.js',
+        '-qIP447rVlTTwaZjY', 'KNcGOksBAvwqQg',
+    ),
+    (
+        'https://www.youtube.com/s/player/643afba4/tv-player-ias.vflset/tv-player-ias.js',
+        'ir9-V6cdbCiyKxhr', '2PL7ZDYAALMfmA',
     ),
 ]
 

youtube_dl/extractor/tver.py

@@ -4,58 +4,229 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_str
+from ..compat import (
-from ..utils import (
+    compat_kwargs,
-    int_or_none,
+    compat_str,
-    remove_start,
-    smuggle_url,
-    try_get,
 )
+from ..utils import (
+    ExtractorError,
+    int_or_none,
+    smuggle_url,
+    str_or_none,
+)
+
+try:
+    from ..utils import traverse_obj
+except ImportError:
+    from ..compat import compat_collections_abc
+
+    def traverse_obj(obj, *path_list, **kw):
+        ''' Traverse nested list/dict/tuple'''
+
+        # parameter defaults
+        default = kw.get('default')
+        expected_type = kw.get('expected_type')
+        get_all = kw.get('get_all', True)
+        casesense = kw.get('casesense', True)
+        is_user_input = kw.get('is_user_input', False)
+        traverse_string = kw.get('traverse_string', False)
+
+        def variadic(x, allowed_types=(compat_str, bytes)):
+            return x if isinstance(x, compat_collections_abc.Iterable) and not isinstance(x, allowed_types) else (x,)
+
+        def listish(l):
+            return isinstance(l, (list, tuple))
+
+        def from_iterable(iterables):
+            # chain.from_iterable(['ABC', 'DEF']) --> A B C D E F
+            for it in iterables:
+                for element in it:
+                    yield element
+
+        class Nonlocal:
+            pass
+        nl = Nonlocal()
+
+        if not casesense:
+            _lower = lambda k: (k.lower() if isinstance(k, compat_str) else k)
+            path_list = (map(_lower, variadic(path)) for path in path_list)
+
+        def _traverse_obj(obj, path, _current_depth=0):
+            path = tuple(variadic(path))
+            for i, key in enumerate(path):
+                if obj is None:
+                    return None
+                if listish(key):
+                    obj = [_traverse_obj(obj, sub_key, _current_depth) for sub_key in key]
+                    key = Ellipsis
+                if key is Ellipsis:
+                    obj = (obj.values() if isinstance(obj, dict)
+                           else obj if listish(obj)
+                           else compat_str(obj) if traverse_string else [])
+                    _current_depth += 1
+                    nl.depth = max(nl.depth, _current_depth)
+                    return [_traverse_obj(inner_obj, path[i + 1:], _current_depth) for inner_obj in obj]
+                elif callable(key):
+                    if listish(obj):
+                        obj = enumerate(obj)
+                    elif isinstance(obj, dict):
+                        obj = obj.items()
+                    else:
+                        if not traverse_string:
+                            return None
+                        obj = str(obj)
+                    _current_depth += 1
+                    nl.depth = max(nl.depth, _current_depth)
+                    return [_traverse_obj(v, path[i + 1:], _current_depth) for k, v in obj if key(k)]
+                elif isinstance(obj, dict) and not (is_user_input and key == ':'):
+                    obj = (obj.get(key) if casesense or (key in obj)
+                           else next((v for k, v in obj.items() if _lower(k) == key), None))
+                else:
+                    if is_user_input:
+                        key = (int_or_none(key) if ':' not in key
+                               else slice(*map(int_or_none, key.split(':'))))
+                        if key == slice(None):
+                            return _traverse_obj(obj, tuple([Ellipsis] + list(path[i + 1:])), _current_depth)
+                    if not isinstance(key, (int, slice)):
+                        return None
+                    if not listish(obj):
+                        if not traverse_string:
+                            return None
+                        obj = compat_str(obj)
+                    try:
+                        obj = obj[key]
+                    except IndexError:
+                        return None
+            return obj
+
+        if isinstance(expected_type, type):
+            type_test = lambda val: val if isinstance(val, expected_type) else None
+        elif expected_type is not None:
+            type_test = expected_type
+        else:
+            type_test = lambda val: val
+
+        for path in path_list:
+            nl.depth = 0
+            val = _traverse_obj(obj, path)
+            if val is not None:
+                if nl.depth:
+                    for _ in range(nl.depth - 1):
+                        val = from_iterable(v for v in val if v is not None)
+                    val = [v for v in map(type_test, val) if v is not None]
+                    if val:
+                        return val if get_all else val[0]
+                else:
+                    val = type_test(val)
+                    if val is not None:
+                        return val
+        return default
 
 
 class TVerIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?tver\.jp/(?P<path>(?:corner|episode|feature)/(?P<id>f?\d+))'
+    _VALID_URL = r'https?://(?:www\.)?tver\.jp/(?:(?P<type>lp|corner|series|episodes?|feature|tokyo2020/video)/)+(?P<id>[a-zA-Z0-9]+)'
     # videos are only available for 7 days
     _TESTS = [{
-        'url': 'https://tver.jp/corner/f0062178',
+        'skip': 'videos are only available for 7 days',
+        'url': 'https://tver.jp/episodes/ephss8yveb',
+        'info_dict': {
+            'id': 'ref:af03de03-21ac-4b98-a53b-9ffd9b102e92',
+            'ext': 'mp4',
+            'title': '#44　料理と値段と店主にびっくり　オモてなしすぎウマい店　2時間SP',
+            'description': 'md5:66985373a66fed8ad3cd595a3cfebb13',
+            'upload_date': '20220329',
+            'uploader_id': '4394098882001',
+            'timestamp': 1648527032,
+        },
+        'add_ie': ['BrightcoveNew'],
+    }, {
+        'skip': 'videos are only available for 7 days',
+        'url': 'https://tver.jp/lp/episodes/ep6f16g26p',
+        'info_dict': {
+            'id': '6302378806001',
+            'ext': 'mp4',
+            # "April 11 (Mon) 23:06-Broadcast scheduled": sorry but this is "correct"
+            'title': '4月11日(月)23時06分 ~ 放送予定',
+            'description': 'md5:4029cc5f4b1e8090dfc5b7bd2bc5cd0b',
+            'upload_date': '20220331',
+            'uploader_id': '3971130137001',
+            'timestamp': 1648696456,
+        },
+        'add_ie': ['BrightcoveNew'],
+    }, {
+        'url': 'https://tver.jp/corner/f0103888',
         'only_matching': True,
     }, {
-        'url': 'https://tver.jp/feature/f0062413',
+        'url': 'https://tver.jp/lp/f0033031',
-        'only_matching': True,
-    }, {
-        'url': 'https://tver.jp/episode/79622438',
-        'only_matching': True,
-    }, {
-        # subtitle = ' '
-        'url': 'https://tver.jp/corner/f0068870',
         'only_matching': True,
     }]
-    _TOKEN = None
     BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s'
+    _PLATFORM_UID = None
+    _PLATFORM_TOKEN = None
+
+    def _download_json(self, url_or_request, video_id, **kwargs):
+
+        headers = {
+            'Origin': 'https://s.tver.jp',
+            'Referer': 'https://s.tver.jp/',
+        }
+        headers.update(kwargs.get('headers', {}))
+        kwargs.update(compat_kwargs({'headers': headers, }))
+
+        return super(TVerIE, self)._download_json(url_or_request, video_id, **kwargs)
 
     def _real_initialize(self):
-        self._TOKEN = self._download_json(
+        create_response = self._download_json(
-            'https://tver.jp/api/access_token.php', None)['token']
+            'https://platform-api.tver.jp/v2/api/platform_users/browser/create', None,
+            note='Creating session', data=b'device_type=pc', headers={
+                'Content-Type': 'application/x-www-form-urlencoded',
+            })
+        self._PLATFORM_UID = traverse_obj(create_response, ('result', 'platform_uid'), expected_type=compat_str)
+        self._PLATFORM_TOKEN = traverse_obj(create_response, ('result', 'platform_token'), expected_type=compat_str)
 
     def _real_extract(self, url):
-        path, video_id = re.match(self._VALID_URL, url).groups()
+        video_id = self._match_id(url)
-        main = self._download_json(
-            'https://api.tver.jp/v4/' + path, video_id,
-            query={'token': self._TOKEN})['main']
-        p_id = main['publisher_id']
-        service = remove_start(main['service'], 'ts_')
 
-        r_id = main['reference_id']
+        video_id, video_type = re.match(self._VALID_URL, url).group('id', 'type')
-        if service not in ('tx', 'russia2018', 'sebare2018live', 'gorin'):
+        if video_type not in ('episode', 'episodes', 'series'):
+            webpage = self._download_webpage(url, video_id, note='Resolving to new URL')
+            video_id = self._match_id(self._search_regex(
+                (r'''canonical['"]\s+href\s*=\s*(?P<q>'|")(?P<url>https?://tver\.jp/(?!(?P=q)).+?)(?P=q)''',
+                 r'&link=(?P<url>https?://tver\.jp/(?!(?P=q)).+?)[?&]'),
+                webpage, 'url regex', group='url'))
+        video_info = self._download_json(
+            'https://statics.tver.jp/content/episode/{0}.json'.format(video_id), video_id,
+            query={'v': '5'})
+        p_id = traverse_obj(video_info, ('video', 'accountID'), expected_type=compat_str)
+        r_id = traverse_obj(video_info, ('video', ('videoRefID', 'videoID')), get_all=False, expected_type=compat_str)
+        if None in (p_id, r_id):
+            raise ExtractorError(
+                'Failed to extract '
+                + ', '.join(
+                    (x[0] for x in (('accountID', p_id), ('videoRefID', r_id), )
+                     if x[1] is None)),
+                expected=False)
+        if not r_id.isdigit():
             r_id = 'ref:' + r_id
-        bc_url = smuggle_url(
+
-            self.BRIGHTCOVE_URL_TEMPLATE % (p_id, r_id),
+        additional_info = self._download_json(
-            {'geo_countries': ['JP']})
+            'https://platform-api.tver.jp/service/api/v1/callEpisode/{0}?require_data=mylist,later[epefy106ur],good[epefy106ur],resume[epefy106ur]'.format(video_id),
+            video_id, fatal=False,
+            query={
+                'platform_uid': self._PLATFORM_UID,
+                'platform_token': self._PLATFORM_TOKEN,
+            }, headers={
+                'x-tver-platform-type': 'web'
+            })
 
         return {
             '_type': 'url_transparent',
-            'description': try_get(main, lambda x: x['note'][0]['text'], compat_str),
+            'title': str_or_none(video_info.get('title')),
-            'episode_number': int_or_none(try_get(main, lambda x: x['ext']['episode_number'])),
+            'description': str_or_none(video_info.get('description')),
-            'url': bc_url,
+            'url': smuggle_url(
+                self.BRIGHTCOVE_URL_TEMPLATE % (p_id, r_id), {'geo_countries': ['JP']}),
+            'series': traverse_obj(
+                additional_info, ('result', ('episode', 'series'), 'content', ('seriesTitle', 'title')),
+                get_all=False, expected_type=compat_str),
             'ie_key': 'BrightcoveNew',
         }

youtube_dl/extractor/youtube.py

@@ -91,12 +91,12 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'IOS',
-                    'clientVersion': '19.45.4',
+                    'clientVersion': '20.10.4',
                     'deviceMake': 'Apple',
                     'deviceModel': 'iPhone16,2',
-                    'userAgent': 'com.google.ios.youtube/19.45.4 (iPhone16,2; U; CPU iOS 18_1_0 like Mac OS X;)',
+                    'userAgent': 'com.google.ios.youtube/20.10.4 (iPhone16,2; U; CPU iOS 18_3_2 like Mac OS X;)',
                     'osName': 'iPhone',
-                    'osVersion': '18.1.0.22B83',
+                    'osVersion': '18.3.2.22D82',
                 },
             },
             'INNERTUBE_CONTEXT_CLIENT_NAME': 5,
@@ -109,7 +109,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'MWEB',
-                    'clientVersion': '2.20241202.07.00',
+                    'clientVersion': '2.20250311.03.00',
                     # mweb previously did not require PO Token with this UA
                     'userAgent': 'Mozilla/5.0 (iPad; CPU OS 16_7_10 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/16.6 Mobile/15E148 Safari/604.1,gzip(gfe)',
                 },
@@ -122,7 +122,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'TVHTML5',
-                    'clientVersion': '7.20250120.19.00',
+                    'clientVersion': '7.20250312.16.00',
                     'userAgent': 'Mozilla/5.0 (ChromiumStylePlatform) Cobalt/Version',
                 },
             },
@@ -133,7 +133,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'WEB',
-                    'clientVersion': '2.20241126.01.00',
+                    'clientVersion': '2.20250312.04.00',
                 },
             },
             'INNERTUBE_CONTEXT_CLIENT_NAME': 1,
@@ -692,7 +692,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         'invidious': '|'.join(_INVIDIOUS_SITES),
     }
     _PLAYER_INFO_RE = (
-        r'/s/player/(?P<id>[a-zA-Z0-9_-]{8,})/player',
+        r'/s/player/(?P<id>[a-zA-Z0-9_-]{8,})//(?:tv-)?player',
         r'/(?P<id>[a-zA-Z0-9_-]{8,})/player(?:_ias\.vflset(?:/[a-zA-Z]{2,3}_[a-zA-Z]{2,3})?|-plasma-ias-(?:phone|tablet)-[a-z]{2}_[A-Z]{2}\.vflset)/base\.js$',
         r'\b(?P<id>vfl[a-zA-Z0-9_-]+)\b.*?\.js$',
     )
@@ -1857,7 +1857,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
     def _extract_n_function_code_jsi(self, video_id, jsi, player_id=None):
 
         var_ay = self._search_regex(
-            r'(?:[;\s]|^)\s*(var\s*[\w$]+\s*=\s*"[^"]+"\s*\.\s*split\("\{"\))(?=\s*[,;])',
+            r'(?:[;\s]|^)\s*(var\s*[\w$]+\s*=\s*"(?:\\"|[^"])+"\s*\.\s*split\("\W+"\))(?=\s*[,;])',
             jsi.code, 'useful values', default='')
 
         func_name = self._extract_n_function_name(jsi.code)