Merge 29e60290cfc508f7823b9191d2a2a124128e9876 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

@@ -585,6 +585,7 @@ from .kuwo import (
     KuwoCategoryIE,
     KuwoMvIE,
 )
+from .la1ere import La1ereExtractorPageIE, La1ereExtractorShowPageIE
 from .la7 import LA7IE
 from .laola1tv import (
     Laola1TvEmbedIE,

youtube_dl/extractor/la1ere.py

@@ -0,0 +1,90 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from urllib.parse import quote
+
+from .common import InfoExtractor
+
+
+class La1ereExrtractorBaseIE(InfoExtractor):
+
+    def _extract_given_title(self, webpage, title):
+
+        # get the ID
+        video_id = self._html_search_regex(r'data-id="([^"]+)"', webpage, 'id')
+
+        # request the mpd-request url
+        json_request_url = 'https://k7.ftven.fr/videos/%s?country_code=FR&w=937&h=527&screen_w=1920&screen_h=1200&player_version=5.116.1&domain=la1ere.francetvinfo.fr&device_type=desktop&browser=chrome&browser_version=126&os=linux&diffusion_mode=tunnel&gmt=-0400&capabilities=drm' % video_id
+        json_request = self._download_json(json_request_url, video_id)
+        mpd_request_url = json_request['video']['url']
+
+        # using the mpd-request url, auth to get the full mpd URL with auth params
+        auth_request_url = 'https://hdfauth.ftven.fr/esi/TA?format=json&url=%s' % quote(mpd_request_url)
+        auth_request = self._download_json(auth_request_url, video_id)
+
+        # this is the full playlist URL with auth params
+        playlist_url = auth_request['url']
+        return video_id, playlist_url
+
+
+class La1ereExtractorPageIE(La1ereExrtractorBaseIE):
+    _VALID_URL = r'https://la1ere.francetvinfo.fr/(?P<region>[^/]+)/programme-video/diffusion/(?P<page>[^\.]+).html'
+    _TEST = {
+        'skip': 'Only available in FR',
+        'url': 'https://la1ere.francetvinfo.fr/martinique/programme-video/diffusion/4774522-origine-kongo.html',
+        'info_dict': {
+            'ext': 'mp4',
+            'title': 'Origine Kongo',
+        }
+    }
+
+    def _real_extract(self, url):
+        webpage = self._download_webpage(url, 'la1ere')
+
+        title = self._html_search_regex(r'<h1 .*title.*>(.+?)</h1>', webpage, 'title')
+
+        video_id, playlist_url = self._extract_given_title(webpage, title)
+
+        # get the mpd playlist
+        formats = self._extract_mpd_formats(playlist_url, video_id)
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'formats': formats,
+        }
+
+
+class La1ereExtractorShowPageIE(La1ereExrtractorBaseIE):
+    _VALID_URL = r'https://la1ere.francetvinfo.fr/(?P<region>[^/]+)/programme-video/(?P<show>[^/]+)/diffusion/(?P<page>[^\.]+).html'
+    _TEST = {
+        'skip': 'Only available in FR',
+        'url': 'https://la1ere.francetvinfo.fr/guadeloupe/programme-video/la1ere_guadeloupe_le-13h-en-guadeloupe/diffusion/5643549-emission-du-lundi-29-janvier-2024.html',
+        'info_dict': {
+            'ext': 'mp4',
+            'title': '13H en Guadeloupe - Émission du lundi 29 janvier 2024',
+        }
+    }
+
+    def _real_extract(self, url):
+        webpage = self._download_webpage(url, 'la1ere')
+
+        series_name = self._html_search_regex(r'<span class="m-dtv-player-title__subtitle">(.+?)</span>', webpage, 'series_name')
+        episode_name = self._html_search_regex(r'<h1 .*title.*>(.+?)</h1>', webpage, 'episode_name')
+        title = f'{series_name} - {episode_name}'
+
+        video_id, playlist_url = self._extract_given_title(webpage, title)
+
+        # get the m3u8 playlist
+        formats = self._extract_m3u8_formats(playlist_url, video_id)
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'formats': formats,
+        }
+
+
+

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)