Merge fed9c8d17b52c722d6e1fc3321f9083ad66ba97d 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/extractors.py

@@ -1242,6 +1242,10 @@ from .tagesschau import (
     TagesschauPlayerIE,
     TagesschauIE,
 )
+from .talktv import (
+    TalkTVIE,
+    TalkTVSeriesIE,
+)
 from .tass import TassIE
 from .tbs import TBSIE
 from .tdslifeway import TDSLifewayIE

youtube_dl/extractor/talktv.py

@@ -0,0 +1,178 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+import calendar
+from datetime import datetime
+import time
+
+from .common import InfoExtractor
+
+from ..compat import compat_str
+from ..utils import (
+    clean_html,
+    determine_ext,
+    extract_attributes,
+    ExtractorError,
+    get_elements_by_class,
+    HEADRequest,
+    parse_duration,
+    try_get,
+    unified_timestamp,
+    url_or_none,
+    urljoin,
+)
+
+
+class TalkTVIE(InfoExtractor):
+    IE_NAME = 'talk.tv'
+    IE_DESC = 'TalkTV UK catch-up and live shows'
+    _VALID_URL = r'https?://watch\.talk\.tv/(?P<id>watch/(?:vod|replay)/\d+|live)'
+    _TESTS = [{
+        'url': 'https://watch.talk.tv/watch/replay/12216792',
+        'md5': 'dc9071f7d26f48ce4057a98425894eb3',
+        'info_dict': {
+            'id': '12216792',
+            'ext': 'mp4',
+            'title': 'Piers Morgan Uncensored',
+            'description': 'The host interviews former US President Donald Trump',
+            'timestamp': 1650917390,
+            'upload_date': '20220425',
+            'duration': float,
+        },
+        'params': {
+            'skip_download': True,  # adaptive download
+        },
+    }, {
+        'url': 'https://watch.talk.tv/live',
+        'info_dict': {
+            'id': 'live',
+            'ext': 'mp4',
+            'title': 'Piers Morgan Uncensored',
+            'description': compat_str,
+            'timestamp': int,
+            # needs core fix to force compat_str type
+            'upload_date': r're:\d{8}',
+            'duration': float,
+        },
+        'params': {
+            'skip_download': True,
+        },
+    },
+    ]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url).rsplit('/', 1)[-1]
+        is_live = (video_id == 'live')
+        webpage = self._download_webpage(url, video_id)
+        title = self._html_search_regex(r'(?is)<h1\b[^>]+>\s*(.+?)\s*</h1', webpage, 'title')
+        player = self._search_regex(r'''(<[dD][iI][vV]\b[^>]+?\bid\s*=\s*(?P<q>"|')player(?P=q)[^>]*>)''', webpage, video_id)
+        player = extract_attributes(player)
+        expiry = player.get('expiry')
+        if expiry is not None and expiry < time.time():
+            raise ExtractorError('Video has expired', expected=True)
+        api_info = self._download_json(
+            'https://mm-v2.simplestream.com/ssmp/api.php?id=%(data-id)s&env=%(data-env)s' % player,
+            video_id, note='Downloading API info', fatal=False)
+        player['api_url'] = (
+            url_or_none(try_get(api_info, lambda x: x['response']['api_hostname']))
+            or 'https://v2-streams-elb.simplestreamcdn.com')
+        headers = {'Referer': url, }
+        for item in ('uvid', 'token', ('expiry', 'Token-Expiry')):
+            if isinstance(item, compat_str):
+                name = item.capitalize()
+            else:
+                item, name = item
+            val = player.get('data-' + item)
+            if val is not None:
+                headers[name] = val
+
+        stream_info = self._download_json(
+            '%(api_url)s/api/%(data-type)s/stream/%(data-uvid)s?key=%(data-key)s&platform=firefox&cc=%(data-country)s' % player,
+            video_id, headers=headers)
+        error = try_get(stream_info, lambda x: x['response']['error'])
+        if error:
+            raise ExtractorError('Streaming API reported: ' + error, expected=True)
+        fmt_url = (stream_info['response'].get('drm') in (None, False)) and stream_info['response']['stream']
+        formats = []
+        duration = None
+        description = None
+        timestamp = None
+        if fmt_url:
+            ext = determine_ext(fmt_url)
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    fmt_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    m3u8_id='hls', live=is_live, fatal=False))
+            elif ext == 'mpd':
+                formats.extend(self._extract_mpd_formats(
+                    fmt_url, video_id, mpd_id='dash', live=is_live, fatal=False))
+            else:
+                formats.append({
+                    'url': fmt_url,
+                })
+            if not is_live:
+                res = self._request_webpage(HEADRequest(fmt_url), video_id, note='Checking date', fatal=False)
+                if res is not False:
+                    timestamp = unified_timestamp(res.info().getheader('last-modified'))
+
+        self._sort_formats(formats)
+
+        text_fields = get_elements_by_class('text-start', webpage)
+        for text in text_fields:
+            text = clean_html(text)
+            if text.startswith('EPISODE'):
+                duration = parse_duration(
+                    self._html_search_regex(r'^EPISODE\b\W*(\w[\w\s]*?)\s*$', text, 'duration', default=None))
+            elif text.startswith('Live'):
+                duration = self._html_search_regex(r'^Live\b(?:<[^>]+>|\W)*([0-2]?\d:\d{2}\s*-\s*[0-2]?\d:\d{2})\s*$', text, 'duration', default=None)
+                duration = list(map(lambda x: datetime.strptime(x, '%H:%M'), re.split(r'\s*-\s*', duration)))
+                if None not in duration and len(duration) == 2:
+                    timestamp = datetime.now().replace(hour=duration[0].hour, minute=duration[0].minute, second=0, microsecond=0)
+                    timestamp = calendar.timegm(timestamp.timetuple())
+                    duration = duration[1] - duration[0]
+                    try:
+                        duration = duration.total_seconds()
+                    except AttributeError:
+                        # Py 2.6
+                        duration = duration.td_seconds
+                    if duration is not None and duration < 0:
+                        duration += 24 * 3600
+            else:
+                description = text
+
+        return {
+            # ensure live has a fixed ID
+            'id': player['data-uvid'] if not is_live else video_id,
+            'title': title,
+            'display_id': video_id if not is_live else player['data-uvid'],
+            'formats': formats,
+            'thumbnail': player.get('data-poster'),
+            'duration': duration,
+            'timestamp': timestamp,
+            'description': description,
+            'is_live': is_live,
+        }
+
+
+class TalkTVSeriesIE(InfoExtractor):
+    IE_NAME = 'talk.tv:series'
+    IE_DESC = 'TalkTV UK series catch-up'
+    _VALID_URL = r'https?://(?:watch\.|www\.)?talk\.tv/shows/(?P<id>[\da-f]{8}-(?:[\da-f]{4}-){3}[\da-f]{12})'
+    _TESTS = [{
+        'url': 'https://watch.talk.tv/shows/86dadc3e-c4d2-11ec-b4c6-0af62ebc70d1',
+        'info_dict': {
+            'id': '86dadc3e-c4d2-11ec-b4c6-0af62ebc70d1',
+        },
+        'playlist_mincount': 4,
+    },
+    ]
+
+    def _real_extract(self, url):
+        playlist_id = self._match_id(url)
+        webpage = self._download_webpage(url, playlist_id)
+        episodes = re.finditer(
+            r'''(?i)<a\b[^>]+?\bhref\s*=\s*(?P<q>"|')(?P<href>/watch/(?:(?!(?P=q)).)+)(?P=q)''',
+            webpage)
+        return self.playlist_from_matches(
+            episodes, playlist_id, getter=lambda x: urljoin(url, x.group('href')), ie='TalkTV')

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)