Merge 3872619ed5210e8a60f68cff7dbb64164bdc7b5c 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/ant1newsgr.py

@@ -0,0 +1,188 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..compat import (
+    compat_parse_qs,
+    compat_urllib_parse_urlparse,
+    compat_urlparse,
+)
+from ..utils import (
+    HEADRequest,
+    ExtractorError,
+    determine_ext,
+    smuggle_url,
+    unsmuggle_url,
+    unescapeHTML,
+)
+
+
+class Ant1NewsGrBaseIE(InfoExtractor):
+    @staticmethod
+    def _smuggle_parent_info(url, **info_dict):
+        return smuggle_url(url, {'parent_info': info_dict})
+
+    @staticmethod
+    def _unsmuggle_parent_info(url):
+        unsmuggled_url, data = unsmuggle_url(url, default={'parent_info': {}})
+        return unsmuggled_url, data['parent_info']
+
+    def _download_api_data(self, netloc, cid, scheme='https'):
+        url_parts = (scheme, netloc, self._API_PATH, None, None, None)
+        url = compat_urlparse.urlunparse(url_parts)
+        query = {'cid': cid}
+        return self._download_json(
+            url, cid,
+            'Downloading JSON',
+            'Unable to download JSON',
+            query=query)
+
+    def _download_and_extract_api_data(self, video_id, *args, **kwargs):
+        info = self._download_api_data(*args, **kwargs)
+        try:
+            source = info['url']
+        except KeyError:
+            raise ExtractorError('no source found for %s' % video_id)
+        formats = self._extract_m3u8_formats(source, video_id, 'mp4') \
+            if determine_ext(source) == 'm3u8' else [source]
+        self._sort_formats(formats)
+        return {
+            'id': video_id,
+            'title': info['title'],
+            'thumbnail': info['thumb'],
+            'formats': formats,
+        }
+
+
+class Ant1NewsGrWatchIE(Ant1NewsGrBaseIE):
+    IE_NAME = 'ant1newsgr:watch'
+    IE_DESC = 'ant1news.gr videos'
+    _VALID_URL = r'https?://(?:www\.)?ant1news\.gr/watch/(?P<id>\d+)/'
+    _API_PATH = '/templates/data/player'
+
+    _TEST = {
+        'url': 'https://www.ant1news.gr/watch/1506168/ant1-news-09112021-stis-18-45',
+        'md5': '60a984da5ffc98c9924e6d9dd46c6f04',
+        'info_dict': {
+            'id': '1506168',
+            'ext': 'mp4',
+            'title': 'md5:0ad00fa66ecf8aa233d26ab0dba7514a',
+            'description': 'md5:18665af715a6dcfeac1d6153a44f16b0',
+        },
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        scheme, netloc, _, _, _, _ = compat_urllib_parse_urlparse(url)
+        info = self._download_and_extract_api_data(
+            video_id, netloc, video_id, scheme=scheme)
+        info['description'] = self._og_search_description(webpage)
+        return info
+
+
+class Ant1NewsGrArticleIE(Ant1NewsGrBaseIE):
+    IE_NAME = 'ant1newsgr:article'
+    IE_DESC = 'ant1news.gr articles'
+    _VALID_URL = r'https?://(?:www\.)?ant1news\.gr/[^/]+/article/(?P<id>\d+)/'
+
+    _TESTS = [{
+        'url': 'https://www.ant1news.gr/afieromata/article/549468/o-tzeims-mpont-sta-meteora-oi-apeiles-kai-o-xesikomos-ton-kalogeron',
+        'md5': 'eb635a194c15272c2611a751766b0200',
+        'info_dict': {
+            'id': '_xvg/m_cmbatw=',
+            'ext': 'mp4',
+            'title': 'md5:a93e8ecf2e4073bfdffcb38f59945411',
+        },
+        'expected_warnings': [r'^[Uu]nable to extract JSON-LD'],
+    }, {
+        'url': 'https://ant1news.gr/Society/article/620286/symmoria-anilikon-dikigoros-thymaton-ithelan-na-toys-apoteleiosoyn',
+        'info_dict': {
+            'id': '620286',
+            'title': 'md5:91fe569e952e4d146485740ae927662b',
+        },
+        'expected_warnings': [r'^[Uu]nable to extract JSON-LD'],
+        'playlist_mincount': 2,
+        'params': {
+            'skip_download': True,
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        info = self._search_json_ld(webpage, video_id,
+                                    expected_type='NewsArticle',
+                                    fatal=False)
+        # workaround as _json_ld does not recognize @graph nesting
+        if not info:
+            info['title'] = self._og_search_title(webpage)
+        embed_urls = list(Ant1NewsGrEmbedIE._extract_urls(webpage, url, **info))
+        if not embed_urls:
+            raise ExtractorError('no videos found for %s' % video_id)
+        if len(embed_urls) == 1:
+            return self.url_result(embed_urls[0], ie=Ant1NewsGrEmbedIE.ie_key(),
+                                   video_title=info['title'])
+        return self.playlist_from_matches(
+            embed_urls, video_id, info['title'], ie=Ant1NewsGrEmbedIE.ie_key())
+
+
+class Ant1NewsGrEmbedIE(Ant1NewsGrBaseIE):
+    IE_NAME = 'ant1newsgr:embed'
+    IE_DESC = 'ant1news.gr embedded videos'
+    _VALID_URL = r'''(?x)https?://(?:[a-zA-Z0-9\-]+\.)?
+        (?:antenna|ant1news)\.gr/templates/pages/player
+        \?(?:(?:cid=(?P<id>[^&#]+)|[^&=#]+=[^&#]+)&?)+'''
+    _API_PATH = '/news/templates/data/jsonPlayer'
+
+    _TEST = {
+        'url': 'https://www.antenna.gr/templates/pages/player?cid=3f_li_c_az_jw_y_u=&w=670&h=377',
+        'md5': '12872b12af18b5dbf76528786728de8c',
+        'info_dict': {
+            'id': '3f_li_c_az_jw_y_u=',
+            'ext': 'mp4',
+            'title': 'md5:a30c93332455f53e1e84ae0724f0adf7',
+        },
+    }
+
+    @classmethod
+    def _extract_urls(cls, webpage, origin_url=None, **parent_info):
+        # make the scheme in _VALID_URL optional
+        _URL_RE = r'(?:https?:)?//' + cls._VALID_URL.split('://', 1)[1]
+        # simplify the query string part of _VALID_URL; after extracting iframe
+        # src, the URL will be matched again
+        _URL_RE = _URL_RE.split(r'\?', 1)[0] + r'\?(?:(?!(?P=_q1)).)+'
+        EMBED_RE = r'''(?x)
+            <iframe[^>]+?src=(?P<_q1>%(quot_re)s)(?P<url>%(url_re)s)(?P=_q1)
+        ''' % {'quot_re': r'["\']', 'url_re': _URL_RE}
+        for mobj in re.finditer(EMBED_RE, webpage):
+            url = unescapeHTML(mobj.group('url'))
+            if url.startswith('//'):
+                scheme = compat_urllib_parse_urlparse(origin_url).scheme \
+                    if origin_url else 'https'
+                url = '%s:%s' % (scheme, url)
+            if not cls.suitable(url):
+                continue
+            yield cls._smuggle_parent_info(url, **parent_info)
+
+    def _real_extract(self, url):
+        url, parent_info = type(self)._unsmuggle_parent_info(url)
+        video_id = self._match_id(url)
+
+        # resolve any redirects, to derive the proper base URL for the API query
+        canonical_url = self._request_webpage(
+            HEADRequest(url), video_id,
+            note='Resolve canonical player URL',
+            errnote='Could not resolve canonical player URL').geturl()
+        scheme, netloc, _, _, query, _ = compat_urllib_parse_urlparse(
+            canonical_url)
+        query = compat_parse_qs(query)
+        cid = query['cid'][0]
+
+        info = self._download_and_extract_api_data(
+            video_id, netloc, cid, scheme=scheme)
+        if 'timestamp' not in info and 'timestamp' in parent_info:
+            info['timestamp'] = parent_info['timestamp']
+        return info

youtube_dl/extractor/extractors.py

@@ -1078,6 +1078,11 @@ from .rutube import (
     RutubePersonIE,
     RutubePlaylistIE,
 )
+from .ant1newsgr import (
+    Ant1NewsGrWatchIE,
+    Ant1NewsGrArticleIE,
+    Ant1NewsGrEmbedIE,
+)
 from .rutv import RUTVIE
 from .ruutu import RuutuIE
 from .ruv import RuvIE

youtube_dl/extractor/generic.py

@@ -102,6 +102,7 @@ from .ustream import UstreamIE
 from .arte import ArteTVEmbedIE
 from .videopress import VideoPressIE
 from .rutube import RutubeIE
+from .ant1newsgr import Ant1NewsGrEmbedIE
 from .limelight import LimelightBaseIE
 from .anvato import AnvatoIE
 from .washingtonpost import WashingtonPostIE
@@ -3400,6 +3401,13 @@ class GenericIE(InfoExtractor):
             return self.playlist_from_matches(
                 rutube_urls, video_id, video_title, ie=RutubeIE.ie_key())
 
+        # Look for ant1news.gr embeds
+        ant1newsgr_urls = list(Ant1NewsGrEmbedIE._extract_urls(webpage, url,
+                                                               title=video_title))
+        if ant1newsgr_urls:
+            return self.playlist_from_matches(
+                ant1newsgr_urls, video_id, video_title, ie=Ant1NewsGrEmbedIE.ie_key())
+
         # Look for WashingtonPost embeds
         wapo_urls = WashingtonPostIE._extract_urls(webpage)
         if wapo_urls:

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)