Merge e84b4d1c9ea50e11b687df93f3407bfabba9b729 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/porn91.py

@@ -1,11 +1,21 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
+from ..compat import (
+    compat_urllib_parse_unquote,
+)
 from ..utils import (
+    clean_html,
+    extract_attributes,
+    get_elements_by_class,
     parse_duration,
-    int_or_none,
     ExtractorError,
+    str_to_int,
+    strip_or_none,
+    unified_strdate,
 )
 
 
@@ -15,18 +25,21 @@ class Porn91IE(InfoExtractor):
 
     _TEST = {
         'url': 'http://91porn.com/view_video.php?viewkey=7e42283b4f5ab36da134',
-        'md5': '7fcdb5349354f40d41689bd0fa8db05a',
+        'md5': 'd869db281402e0ef4ddef3c38b866f86',
         'info_dict': {
             'id': '7e42283b4f5ab36da134',
             'title': '18岁大一漂亮学妹，水嫩性感，再爽一次！',
             'ext': 'mp4',
             'duration': 431,
             'age_limit': 18,
+            'upload_date': '20150520',
+            'uploader': '千岁九王爷',
         }
     }
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
+        # set language for page to be extracted
         self._set_cookie('91porn.com', 'language', 'cn_CN')
 
         webpage = self._download_webpage(
@@ -35,29 +48,53 @@ class Porn91IE(InfoExtractor):
         if '作为游客，你每天只可观看10个视频' in webpage:
             raise ExtractorError('91 Porn says: Daily limit 10 videos exceeded', expected=True)
 
-        title = self._search_regex(
+        title = self._html_search_regex(
-            r'<div id="viewvideo-title">([^<]+)</div>', webpage, 'title')
+            r'(?s)<title\b[^>]*>\s*(.+?)\s*(?:Chinese\s+homemade\s+video\s*)?</title', webpage, 'title')
-        title = title.replace('\n', '')
 
         video_link_url = self._search_regex(
-            r'<textarea[^>]+id=["\']fm-video_link[^>]+>([^<]+)</textarea>',
+            r'''document\s*\.\s*write\s*\(\s*strencode2\s*\((?P<q>"|')(?P<enc_str>[%\da-fA-F]+)(?P=q)\s*\)''',
-            webpage, 'video link')
+            webpage, 'video link', group='enc_str')
-        videopage = self._download_webpage(video_link_url, video_id)
+        video_link_url = compat_urllib_parse_unquote(video_link_url)
 
-        info_dict = self._parse_html5_media_entries(url, videopage, video_id)[0]
+        info_dict = self._parse_html5_media_entries(url, '<video>%s</video>' % (video_link_url, ), video_id)[0]
 
-        duration = parse_duration(self._search_regex(
+        # extract various fields in <tag class=info>Name: value</tag>
-            r'时长:\s*</span>\s*(\d+:\d+)', webpage, 'duration', fatal=False))
+        FIELD_MAP = {
+            # cn_CN name: (yt-dl key, value parser, en name)
+            '时长': ('duration', parse_duration, 'Runtime', ),
+            '查看': ('view_count', str_to_int, 'Views', ),
+            '留言': ('comment_count', str_to_int, 'Comments', ),
+            '收藏': ('like_count', str_to_int, 'Favorites', ),
+            '添加时间': ('upload_date', unified_strdate, 'Added', ),
+            # same as title for en, not description for cn_CN
+            '__ignore__': ('description', strip_or_none, 'Description', ),
+            '作者': ('uploader', strip_or_none, 'From', ),
+        }
+        # yt-dl's original implementation of get_elements_by_class() uses regex
+        # yt-dlp uses an actual HTML parser, and can be confused by bad HTML fragments
+        for elt in get_elements_by_class(
+                'info',
+                # concatenate <span>s ...
+                re.sub(r'(?i)</span>\s*<span\b[^>]*?>', '',
+                       # ... and strip out possibly unbalanced <font> for yt-dlp
+                       re.sub(r'(?i)(?:<font\b[^>]*?>|</font\s*>)', '', webpage))) or []:
+            elt = re.split(r':\s*', clean_html(elt), 1)
+            if len(elt) != 2 or elt[1] == '':
+                continue
+            parm = FIELD_MAP.get(elt[0].strip())
+            if parm and elt[1] is not None:
+                info_dict[parm[0]] = parm[1](elt[1]) if parm[1] else elt[1]
 
-        comment_count = int_or_none(self._search_regex(
+        thumbnail = extract_attributes(
-            r'留言:\s*</span>\s*(\d+)', webpage, 'comment count', fatal=False))
+            self._search_regex(
+                r'''(<video\b[^>]*\bid\s*=\s*(?P<q>"|')player_one(?P=q)[^>]*>)''',
+                webpage, 'poster', default='')).get('poster')
 
         info_dict.update({
             'id': video_id,
             'title': title,
-            'duration': duration,
-            'comment_count': comment_count,
             'age_limit': self._rta_search(webpage),
+            'thumbnail': thumbnail,
         })
 
         return info_dict

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)