Merge b6bfdb25f200ba63ba1b011a54431cb424be46e9 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>

docs/supportedsites.md

@@ -910,6 +910,9 @@
  - **SVTPage**
  - **SVTPlay**: SVT Play and Öppet arkiv
  - **SVTSeries**
+ - **switchtube**
+ - **switchtube:channel**
+ - **switchtube:profile**
  - **SWRMediathek**
  - **Syfy**
  - **SztvHu**

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

@@ -1235,6 +1235,11 @@ from .svt import (
     SVTPlayIE,
     SVTSeriesIE,
 )
+from .switchtube import (
+  SwitchTubeIE,
+  SwitchTubeProfileIE,
+  SwitchTubeChannelIE,
+)
 from .swrmediathek import SWRMediathekIE
 from .syfy import SyfyIE
 from .sztvhu import SztvHuIE

youtube_dl/extractor/switchtube.py

@@ -0,0 +1,219 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import itertools
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    clean_html,
+    extract_attributes,
+    float_or_none,
+    get_element_by_attribute,
+    int_or_none,
+    parse_iso8601,
+    strip_or_none,
+    urljoin,
+)
+
+
+class SwitchTubeIE(InfoExtractor):
+    _VALID_URL = r'https?://tube\.switch\.ch/videos/(?P<id>[\da-zA-Z]+)'
+    IE_NAME = 'switchtube'
+    _TESTS = [{
+        'url': 'https://tube.switch.ch/videos/0T1XfaIFSX',
+        'info_dict': {
+            'id': '0T1XfaIFSX',
+            'title': '2016_ASE_sqC03-Entretien',
+            'channel': 'ASE Assistant-e socio-éducatif-ve CFC',
+            'channel_url': 'https://tube.switch.ch/channels/bsaer76yoL',
+            'channel_id': 'bsaer76yoL',
+            'ext': 'mp4',
+            'description': None,
+            'thumbnail': r're:^https?://tube.switch.ch/image/representations/[\w-]+$',
+            'license': 'All rights reserved',
+            'creator': 'Jean-Marc Pouly from Eidgenössische Hochschule für Berufsbildung',
+            'uploader': 'Jean-Marc Pouly from Eidgenössische Hochschule für Berufsbildung',
+            'uploader_url': 'https://tube.switch.ch/profiles/42481',
+            'uploader_id': '42481',
+            'upload_date': '20220309',
+            'timestamp': 1646839068,
+        }
+    }, {
+        'url': 'https://tube.switch.ch/videos/0cf3886d',
+        'info_dict': {
+            'id': '0cf3886d',
+            'ext': 'mp4',
+            'title': 'Introduction: Mini-Batches in On- and Off-Policy Deep Reinforcement Learning',
+            'license': 'All rights reserved',
+            'description': 'One of the challenges in Deep Reinforcement Learning is to decorrelate the data. How this is possible with replay buffers is explained here.',
+            'thumbnail': r're:^https?://tube.switch.ch/image/representations/[\w-]+$',
+            'channel': 'CS-456 Artificial Neural Networks',
+            'channel_url': 'https://tube.switch.ch/channels/1deb03e0',
+            'channel_id': '1deb03e0',
+            'timestamp': 1590733406,
+            'upload_date': '20200529',
+            'creator': 'Wulfram Gerstner from École polytechnique fédérale de Lausanne (EPFL)',
+            'uploader': 'Wulfram Gerstner from École polytechnique fédérale de Lausanne (EPFL)',
+            'uploader_url': 'https://tube.switch.ch/profiles/94248',
+            'uploader_id': '94248',
+        }
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        title = self._html_search_regex(r'(?s)<title\b[^>]*>(.+?)</title>', webpage, 'title')
+
+        info = {
+            'id': video_id,
+            'title': title,
+            'is_live': False,
+        }
+
+        info['view_count'] = int_or_none(self._html_search_regex(r'(?s)<p\b[^>]*>\s*(\d+)\s+views?\s*</p>', webpage,
+                                                                 'view_count', default=None))
+        info['license'] = self._html_search_regex(r'''(?s)<span\b[^>]*?\bproperty\s*=\s*["'](?:\b[cd]c:license\b\s*){2,}[^>]+>(.+)</span>''',
+                                                  webpage, 'license', default=None)
+        info['description'] = strip_or_none(clean_html(
+            get_element_by_attribute('property', 'dc:description', webpage)))
+        info['duration'] = float_or_none(
+            self._search_regex(r'''\bdata-duration\s*=\s*["']([\d.]+)''',
+                               webpage, 'duration', default=None))
+
+        channel_groups = self._search_regex(r'''(?s)<a\b[^>]+?\bhref\s*=\s*["']/channels/(.{4,}?)</a>''',
+                                            webpage, 'channel groups', default='')
+        channel_groups = re.split(r'''(?s)["'][^>]*>\s*''', channel_groups, 1)
+        if len(channel_groups) == 2:
+            for i, ch in enumerate(('channel_id', 'channel')):
+                info[ch] = strip_or_none(channel_groups[i])
+            if info['channel_id']:
+                info['channel_url'] = 'https://tube.switch.ch/channels/' + info['channel_id']
+
+        def outer_elements_by_attribute(attr, value, html, tag=None, escape_value=True, include_inner=False):
+            """generate matching HTML element strings
+               if include_inner, tuples of (element, content)"""
+            pattern = r'''(?s)(?P<element><(%s)\b[^>]+?\b%s\s*=\s*("|')%s\b[^>]+>)(?P<inner>.*?)</\2>''' % \
+                (re.escape(tag) if tag is not None else r'\w+', attr, re.escape(value) if escape_value else value)
+            matches = re.finditer(pattern, html)
+            for m in matches:
+                yield m.group('element', 'inner') if include_inner else m.group('element')
+
+        for dt in outer_elements_by_attribute('property', 'dc:date', webpage, tag='span'):
+            dt = extract_attributes(dt)
+            if dt.get('class') == 'dt-published':
+                info['timestamp'] = parse_iso8601(dt.get('content'))
+                break
+
+        creator_groups = self._search_regex(
+            r'''(?s)<span\b[^>]+?\bclass\s*=\s*("|')(?:(?!\1).)*\bp-author\b(?:(?!\1).)*\1\s*property\s*=\s*["']dc:creator\b[^>]+>\s*(.*?<span\b[^>]+?\bclass\s*=\s*["']p-name\b.*</span>).*?</span>''',
+            webpage, 'creator groups', default='', group=2)
+        creator_groups = re.match(r'''(?s)<a\b[^>]+?\bhref\s*=\s*("|')/profiles/(?P<profile_id>.+?)\1[^>]*>\s*<span\b[^>]+>\s*(?P<creator_name>.+?)\s*</span>[,\s]*<span\b[^>]+\bclass\s*=\s*["']p-organization-name\b[^>]+>\s*(?P<organization_name>.+?)\s*</span>''', creator_groups)
+        if creator_groups:
+            creator_groups = creator_groups.groupdict()
+            info['uploader'] = info['creator'] = ' from '.join((creator_groups['creator_name'], creator_groups['organization_name']))
+            info['uploader_id'] = creator_groups['profile_id']
+            info['uploader_url'] = 'https://tube.switch.ch/profiles/' + info['uploader_id']
+
+        parsed_media_entries = self._parse_html5_media_entries(url, webpage, video_id)[0]
+        info['thumbnail'] = parsed_media_entries['thumbnail']
+        info['formats'] = parsed_media_entries['formats']
+        self._sort_formats(info['formats'])
+
+        return info
+
+
+class SwitchTubeProfileIE(InfoExtractor):
+    _VALID_URL = r'https?://tube\.switch\.ch/profiles/(?P<id>[\da-zA-Z]+)'
+    IE_NAME = 'switchtube:profile'
+    _TESTS = [{
+        'url': 'https://tube.switch.ch/profiles/94248',
+        'info_dict': {
+            'id': '94248',
+            'title': 'Wulfram Gerstner',
+            'description': None,
+        },
+        'playlist_mincount': 94,
+    }]
+
+    @classmethod
+    def suitable(cls, url):
+        return False if SwitchTubeIE.suitable(url) else super(SwitchTubeProfileIE, cls).suitable(url)
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+        webpage = self._download_webpage(url, channel_id)
+        channel_title = self._html_search_regex(r'(?s)<title\b[^>]*>(.+?)</title>', webpage, 'title', default=None)
+
+        description = self._html_search_regex(r'''(?s)<div\b[^>]+class\s*=\s*("|')p-summary formatted\1[^>]+property\s*=\s*("|')dc:description\2[^>]*>\s*(.+?)\s*</div>''',
+                                              webpage, 'description', default=None, group=3)
+
+        entries = []
+        next_page = None
+        for current_page_number in itertools.count(1):
+            if next_page:
+                webpage = self._download_webpage(next_page, channel_id, note='Downloading page %d' % (current_page_number, ))
+
+            for _, video_path, video_id, title in re.findall(
+                    r'''(?s)<a\b[^>]+\bhref\s*=\s*("|')(/videos/((?:(?!\1).)+?))\1[^>]*>\s*<div\b[^>]+\bclass\s*=\s*["']title\b[^>]+>(.+?)</div>''',
+                    webpage):
+                video_url = urljoin(url, video_path)
+                if video_url:
+                    entries.append(self.url_result(video_url, ie=SwitchTubeIE.ie_key(), video_id=video_id))
+
+            next_page = self._search_regex(
+                r'''<a\b[^>]+?\bhref\s*=\s*("|')(?P<path>/profiles/%s\?(?:(?!\1).)+)\1[^>]*>\s*Next\s*</a>''' % (channel_id,),
+                webpage, 'next page', group='path', default=None)
+            if next_page:
+                next_page = urljoin(url, next_page)
+            if not next_page:
+                break
+
+        return self.playlist_result(entries, channel_id, channel_title,
+                                    description)
+
+
+class SwitchTubeChannelIE(InfoExtractor):
+    _VALID_URL = r'https?://tube\.switch\.ch/channels/(?P<id>[\da-zA-Z]+)'
+    IE_NAME = 'switchtube:channel'
+    _TESTS = [{
+        'url': 'https://tube.switch.ch/channels/1deb03e0',
+        'info_dict': {
+            'id': '1deb03e0',
+            'title': 'CS-456 Artificial Neural Networks',
+            'description': 'Class on Artificial Neural Networks and Reinforcement Learning designed for EPFL master students in CS and related disciplines.'
+        },
+        'playlist_mincount': 94,
+    }]
+
+    @classmethod
+    def suitable(cls, url):
+        return False if SwitchTubeIE.suitable(url) else super(SwitchTubeChannelIE, cls).suitable(url)
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+
+        entries = []
+        for current_page_number in itertools.count(0):
+            page_url = urljoin(url, '/channels/%s?order=episodes&page=%d' % (channel_id, current_page_number))
+            webpage = self._download_webpage(page_url, channel_id, note='Downloading page %d' % (current_page_number + 1, ))
+
+            if current_page_number == 0:
+                channel_title = self._html_search_regex(r'(?s)<title\b[^>]*>(.+?)</title>', webpage, 'title', default=None)
+                description = self._html_search_regex(r'''(?s)<div\b[^>]+class\s*=\s*("|')description formatted\1[^>]*>\s*<p>\s*(.+?)\s*</p>\s*</div>''',
+                                                      webpage, 'description', default=None, group=2)
+
+            this_page_still_has_something = False
+            for _, video_path, video_id in re.findall(
+                    r'''(?s)<a\b[^>]+\bhref\s*=\s*("|')(/videos/((?:(?!\1).)+?))\1[^>]*>''',
+                    webpage):
+                video_url = urljoin(url, video_path)
+                if video_url:
+                    this_page_still_has_something = True
+                    entries.append(self.url_result(video_url, ie=SwitchTubeIE.ie_key(), video_id=video_id))
+
+            if not this_page_still_has_something:
+                break
+
+        return self.playlist_result(entries, channel_id, channel_title,
+                                    description)

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)