Merge a784be7395cbb0426b8111ca126e73fe4cf653cc 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',
-        'W9HJZKktxuYoDTqW', 'U48vOZHaeYS6vO',
+        'https://www.youtube.com/s/player/643afba4/player_ias.vflset/en_US/base.js',
+        '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/extractors.py

@@ -1701,3 +1701,4 @@ from .zingmp3 import (
 )
 from .zoom import ZoomIE
 from .zype import ZypeIE
+from .invidious import InvidiousIE, InvidiousPlaylistIE

youtube_dl/extractor/invidious.py

@@ -0,0 +1,217 @@
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..compat import compat_urllib_parse
+
+INSTANCES = [
+    'y.com.sb',
+    'yt.artemislena.eu'
+]
+
+INSTANCES_HOST_REGEX = '(?:' + '|'.join([instance.replace('.', r'\.') for instance in INSTANCES]) + ')'
+
+
+class InvidiousIE(InfoExtractor):
+    _VALID_URL = r'(?:https?://)?(?:www\.)?' + INSTANCES_HOST_REGEX + r'/watch\?v=(?P<id>.+)'
+    _TESTS = [
+        {
+            'url': 'https://y.com.sb/watch?v=xKTygGa6hg0',
+            'info_dict': {
+                'id': 'xKTygGa6hg0',
+                'ext': 'mp4',
+                'title': 'Coding in C++ - Creating a Player Controller - CRYENGINE Summer Academy S1E5 - [Tutorial]',
+                'uploader': 'CRYENGINE',
+                'uploader_id': 'UCtaXcIVFp8HEpthm7qwtKCQ',
+                'description': 'md5:7aa75816d40ffccdbf3e15a90b05fca3',
+            }
+        },
+        {
+            'url': 'https://yt.artemislena.eu/watch?v=BaW_jenozKc',
+            'md5': '5515885fed58607bfae88f7d2090bc93',
+            'info_dict': {
+                'id': 'BaW_jenozKc',
+                'ext': 'mp4',
+                'title': 'youtube-dl test video "\'/\\ä↭𝕐',
+                'uploader': 'Philipp Hagemeister',
+                'uploader_id': 'UCLqxVugv74EIW3VWh2NOa3Q',
+                'channel_id': 'UCLqxVugv74EIW3VWh2NOa3Q',
+                'description': 'test chars:  "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
+                'tags': ['youtube-dl'],
+                'duration': 10,
+                'view_count': int,
+                'like_count': int,
+                'dislike_count': int,
+            }
+        },
+    ]
+
+    def __init__(self, downloader=None):
+        super().__init__(downloader)
+
+    # type is either 'video' or 'audio'
+    # ext is the file extension
+    @staticmethod
+    def _get_additional_format_data(format_type, bitrate, resolution, fps):
+        out = {}
+
+        try:
+            type_and_ext, codecs = format_type.split(';')
+        except Exception:
+            pass
+
+        try:
+            type_, ext = type_and_ext.split('/')
+            # codec = codecs.split('"')[1]
+            out['ext'] = ext
+            # if type_ == 'audio':
+            #     out['acodec'] = codec
+            # elif type_ == 'video':
+            #     out['vcodec'] = codec
+        except Exception:
+            pass
+
+        try:
+            bitrate = float(bitrate) / 1000
+            # if type_ == 'audio':
+            #     out['abr'] = bitrate
+            # elif type_ == 'video':
+            #     out['vbr'] = bitrate
+            # out['tbr'] = bitrate
+        except Exception:
+            pass
+
+        try:
+            if type_ == 'audio':
+                out['resolution'] = type_and_ext + ' @ ' + str(bitrate) + 'k - audio only'
+            elif type_ == 'video':
+                out['resolution'] = resolution + ' - ' + type_and_ext + ' @ ' + str(fps) + 'fps - video only'
+        except Exception:
+            pass
+
+        return out
+
+    def _patch_url(self, url):
+        return compat_urllib_parse.urlparse(url)._replace(netloc=self.url_netloc).geturl()
+
+    def _get_formats(self, api_response):
+        all_formats = []
+
+        # Video/audio only
+        for format_ in api_response.get('adaptiveFormats') or []:
+            all_formats.append({
+                'url': self._patch_url(format_['url']),
+                'format_id': format_.get('itag'),
+                # 'fps': format_.get('fps'),
+                # 'container': format_.get('container')
+            } | InvidiousIE._get_additional_format_data(format_.get('type'), format_.get('bitrate'), format_.get('resolution'), format_.get('fps')))
+
+        # Both video and audio
+        for format_ in api_response.get('formatStreams') or []:
+            all_formats.append({
+                'url': self._patch_url(format_['url']),
+                'format_id': format_.get('itag'),
+                # 'fps': format_.get('fps'),
+                # 'container': format_.get('container')
+            } | InvidiousIE._get_additional_format_data(format_.get('type'), format_.get('bitrate'), format_.get('resolution'), format_.get('fps')))
+
+        return all_formats
+
+    def _get_thumbnails(self, api_response):
+        thumbnails = []
+        video_thumbnails = api_response.get('videoThumbnails') or []
+
+        for inversed_quality, thumbnail in enumerate(video_thumbnails):
+            thumbnails.append({
+                'id': thumbnail.get('quality'),
+                'url': thumbnail.get('url'),
+                'quality': len(video_thumbnails) - inversed_quality,
+                'width': thumbnail.get('width'),
+                'height': thumbnail.get('height')
+            })
+
+        return thumbnails
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = None
+
+        # host_url will contain `http[s]://example.com` where `example.com` is the used invidious instance.
+        url_parsed = compat_urllib_parse.urlparse(url)
+        self.url_netloc = url_parsed.netloc
+        host_url = url_parsed.scheme + '://' + url_parsed.netloc
+
+        api_response = self._download_json(host_url + '/api/v1/videos/' + video_id, video_id)
+
+        def download_webpage_and(fn, fatal=True):
+            global webpage
+            if webpage is None:
+                webpage = self._download_webpage(url, video_id, fatal=fatal)
+            return fn()
+
+        out = {
+            'id': video_id,
+            'title': api_response.get('title') or download_webpage_and(lambda: self._og_search_title(webpage)),
+            'description': api_response.get('description') or download_webpage_and(lambda: self._og_search_description(webpage)),
+
+            'release_timestamp': api_response.get('published'),
+
+            'uploader': api_response.get('author'),
+            'uploader_id': api_response.get('authorId'),
+            'channel': api_response.get('author'),
+            'channel_id': api_response.get('authorId'),
+            'channel_url': host_url + api_response.get('authorUrl'),
+
+            'duration': api_response.get('lengthSeconds'),
+
+            'view_count': api_response.get('viewCount'),
+            'like_count': api_response.get('likeCount'),
+            'dislike_count': api_response.get('dislikeCount'),
+
+            # 'isFamilyFriendly': 18 if api_response.get('isFamilyFriendly') == False else None
+
+            'tags': api_response.get('keywords'),
+            'is_live': api_response.get('liveNow'),
+
+            'formats': self._get_formats(api_response),
+            'thumbnails': self._get_thumbnails(api_response)
+        }
+
+        if api_response.get('isFamilyFriendly') is False:
+            out['age_limit'] = 18
+
+        return out
+
+
+class InvidiousPlaylistIE(InfoExtractor):
+    _VALID_URL = r'(?:https?://)?(?:www\.)?' + INSTANCES_HOST_REGEX + r'/playlist\?list=(?P<id>.+)'
+    _TEST = {
+        'url': 'https://yt.artemislena.eu/playlist?list=PLowKtXNTBypGqImE405J2565dvjafglHU',
+        'md5': 'de4a9175071169961fe7cf2b6740da12',
+        'info_dict': {
+            'id': 'HyznrdDSSGM',
+            'ext': 'mp4',
+            'title': '8-bit computer update',
+            'uploader': 'Ben Eater',
+            'uploader_id': 'UCS0N5baNlQWJCUrhCEo8WlA',
+            'description': 'An update on my plans to build another 8-bit computer from scratch and make videos of the whole process! Buy a kit and build your own! https://eater.net/8bit/kits\n\nSupport me on Patreon: https://www.patreon.com/beneater',
+        }
+    }
+
+    def _get_entries(self, api_response):
+        return [InvidiousIE(self._downloader)._real_extract(self.host_url + '/watch?v=' + video['videoId'])
+                for video in api_response['videos']]
+
+    def _real_extract(self, url):
+        playlist_id = self._match_id(url)
+
+        # host_url will contain `http[s]://example.com` where `example.com` is the used invidious instance.
+        url_parsed = compat_urllib_parse.urlparse(url)
+        self.host_url = url_parsed.scheme + '://' + url_parsed.netloc
+
+        api_response = self._download_json(self.host_url + '/api/v1/playlists/' + playlist_id, playlist_id)
+        return InfoExtractor.playlist_result(self._get_entries(api_response), playlist_id, api_response.get('title'), api_response.get('description')) | {
+            'release_timestamp': api_response.get('updated'),
+
+            'uploader': api_response.get('author'),
+            'uploader_id': api_response.get('authorId'),
+        }

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)