Merge 66ee6aa2da2d71ababcbce6b7604c29dc83d2c82 into 2b4fbfce25902d557b86b003cf48f738129efce4

thx seproDev, bashonly: yt-dlp/yt-dlp#12748

test/test_youtube_signature.py

@@ -84,6 +84,21 @@ _SIG_TESTS = [
         '2aq0aqSyOoJXtK73m-uME_jv7-pT15gOFC02RFkGMqWpzEICs69VdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
         '0QJ8wRAIgXmPlOPSBkkUs1bYFYlJCfe29xxAj7v1pDL0QwbdV96sCIEzpWqMGkFR20CFOg51Tp-7vj_EMu-m37KtXJ2OySqa0q',
     ),
+    (
+        'https://www.youtube.com/s/player/643afba4/tv-player-ias.vflset/tv-player-ias.js',
+        '2aq0aqSyOoJXtK73m-uME_jv7-pT15gOFC02RFkGMqWpzEICs69VdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
+        'AAOAOq0QJ8wRAIgXmPlOPSBkkUs1bYFYlJCfe29xx8j7vgpDL0QwbdV06sCIEzpWqMGkFR20CFOS21Tp-7vj_EMu-m37KtXJoOy1',
+    ),
+    (
+        'https://www.youtube.com/s/player/363db69b/player_ias.vflset/en_US/base.js',
+        '2aq0aqSyOoJXtK73m-uME_jv7-pT15gOFC02RFkGMqWpzEICs69VdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
+        '0aqSyOoJXtK73m-uME_jv7-pT15gOFC02RFkGMqWpz2ICs6EVdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
+    ),
+    (
+        'https://www.youtube.com/s/player/4fcd6e4a/player_ias.vflset/en_US/base.js',
+        '2aq0aqSyOoJXtK73m-uME_jv7-pT15gOFC02RFkGMqWpzEICs69VdbwQ0LDp1v7j8xx92efCJlYFYb1sUkkBSPOlPmXgIARw8JQ0qOAOAA',
+        'wAOAOq0QJ8ARAIgXmPlOPSBkkUs1bYFYlJCfe29xx8q7v1pDL0QwbdV96sCIEzpWqMGkFR20CFOg51Tp-7vj_EMu-m37KtXJoOySqa0',
+    ),
 ]
 
 _NSIG_TESTS = [
@@ -153,7 +168,7 @@ _NSIG_TESTS = [
     ),
     (
         'https://www.youtube.com/s/player/c57c113c/player_ias.vflset/en_US/base.js',
-        '-Txvy6bT5R6LqgnQNx', 'dcklJCnRUHbgSg',
+        'M92UUMHa8PdvPd3wyM', '3hPqLJsiNZx7yA',
     ),
     (
         'https://www.youtube.com/s/player/5a3b6271/player_ias.vflset/en_US/base.js',
@@ -173,7 +188,7 @@ _NSIG_TESTS = [
     ),
     (
         'https://www.youtube.com/s/player/cfa9e7cb/player_ias.vflset/en_US/base.js',
-        'qO0NiMtYQ7TeJnfFG2', 'k9cuJDHNS5O7kQ',
+        'aCi3iElgd2kq0bxVbQ', 'QX1y8jGb2IbZ0w',
     ),
     (
         'https://www.youtube.com/s/player/8c7583ff/player_ias.vflset/en_US/base.js',
@@ -231,10 +246,6 @@ _NSIG_TESTS = [
         'https://www.youtube.com/s/player/f6e09c70/player_ias_tce.vflset/en_US/base.js',
         'W9HJZKktxuYoDTqW', 'jHbbkcaxm54',
     ),
-    (
-        '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',
@@ -259,6 +270,22 @@ _NSIG_TESTS = [
         'https://www.youtube.com/s/player/643afba4/tv-player-ias.vflset/tv-player-ias.js',
         'ir9-V6cdbCiyKxhr', '2PL7ZDYAALMfmA',
     ),
+    (
+        'https://www.youtube.com/s/player/643afba4/player_ias.vflset/en_US/base.js',
+        'W9HJZKktxuYoDTqW', 'larxUlagTRAcSw',
+    ),
+    (
+        'https://www.youtube.com/s/player/363db69b/player_ias.vflset/en_US/base.js',
+        'eWYu5d5YeY_4LyEDc', 'XJQqf-N7Xra3gg',
+    ),
+    (
+        'https://www.youtube.com/s/player/4fcd6e4a/player_ias.vflset/en_US/base.js',
+        'o_L251jm8yhZkWtBW', 'lXoxI3XvToqn6A',
+    ),
+    (
+        'https://www.youtube.com/s/player/4fcd6e4a/tv-player-ias.vflset/tv-player-ias.js',
+        'o_L251jm8yhZkWtBW', 'lXoxI3XvToqn6A',
+    ),
 ]
 
 
@@ -271,6 +298,8 @@ class TestPlayerInfo(unittest.TestCase):
             ('https://www.youtube.com/s/player/64dddad9/player-plasma-ias-phone-en_US.vflset/base.js', '64dddad9'),
             ('https://www.youtube.com/s/player/64dddad9/player-plasma-ias-phone-de_DE.vflset/base.js', '64dddad9'),
             ('https://www.youtube.com/s/player/64dddad9/player-plasma-ias-tablet-en_US.vflset/base.js', '64dddad9'),
+            ('https://www.youtube.com/s/player/e7567ecf/player_ias_tce.vflset/en_US/base.js', 'e7567ecf'),
+            ('https://www.youtube.com/s/player/643afba4/tv-player-ias.vflset/tv-player-ias.js', '643afba4'),
             # obsolete
             ('https://www.youtube.com/yts/jsbin/player_ias-vfle4-e03/en_US/base.js', 'vfle4-e03'),
             ('https://www.youtube.com/yts/jsbin/player_ias-vfl49f_g4/en_US/base.js', 'vfl49f_g4'),
@@ -280,8 +309,9 @@ class TestPlayerInfo(unittest.TestCase):
             ('https://s.ytimg.com/yts/jsbin/html5player-en_US-vflXGBaUN.js', 'vflXGBaUN'),
             ('https://s.ytimg.com/yts/jsbin/html5player-en_US-vflKjOTVq/html5player.js', 'vflKjOTVq'),
         )
+        ie = YoutubeIE(FakeYDL({'cachedir': False}))
         for player_url, expected_player_id in PLAYER_URLS:
-            player_id = YoutubeIE._extract_player_info(player_url)
+            player_id = ie._extract_player_info(player_url)
             self.assertEqual(player_id, expected_player_id)
 
 
@@ -301,8 +331,8 @@ class TestSignature(unittest.TestCase):
 def t_factory(name, sig_func, url_pattern):
     def make_tfunc(url, sig_input, expected_sig):
         m = url_pattern.match(url)
-        assert m, '%r should follow URL format' % url
-        test_id = m.group('id')
+        assert m, '{0!r} should follow URL format'.format(url)
+        test_id = re.sub(r'[/.-]', '_', m.group('id') or m.group('compat_id'))
 
         def test_func(self):
             basename = 'player-{0}-{1}.js'.format(name, test_id)
@@ -335,12 +365,16 @@ def n_sig(jscode, sig_input):
 
 
 make_sig_test = t_factory(
-    'signature', signature, re.compile(r'.*(?:-|/player/)(?P<id>[a-zA-Z0-9_-]+)(?:/.+\.js|(?:/watch_as3|/html5player)?\.[a-z]+)$'))
+    'signature', signature,
+    re.compile(r'''(?x)
+        .+/(?P<h5>html5)?player(?(h5)(?:-en_US)?-|/)(?P<id>[a-zA-Z0-9/._-]+)
+        (?(h5)/(?:watch_as3|html5player))?\.js$
+    '''))
 for test_spec in _SIG_TESTS:
     make_sig_test(*test_spec)
 
 make_nsig_test = t_factory(
-    'nsig', n_sig, re.compile(r'.+/player/(?P<id>[a-zA-Z0-9_-]+)/.+.js$'))
+    'nsig', n_sig, re.compile(r'.+/player/(?P<id>[a-zA-Z0-9_/.-]+)\.js$'))
 for test_spec in _NSIG_TESTS:
     make_nsig_test(*test_spec)
 

youtube_dl/extractor/common.py

@@ -1370,6 +1370,10 @@ class InfoExtractor(object):
         if isinstance(json_ld, dict):
             json_ld = [json_ld]
 
+        def valued_dict(items):
+            """Return dict from dict or iterable of pairs omitting None values"""
+            return dict((k, v) for k, v in (items.items() if isinstance(items, dict) else items) if v is not None)
+
         INTERACTION_TYPE_MAP = {
             'CommentAction': 'comment',
             'AgreeAction': 'like',
@@ -1461,19 +1465,25 @@ class InfoExtractor(object):
                     part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries')
                     if isinstance(part_of_series, dict) and part_of_series.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'):
                         info['series'] = unescapeHTML(part_of_series.get('name'))
-                elif item_type == 'Movie':
+                elif item_type in ('TVSeries', 'Series', 'CreativeWorkSeries'):
+                    series_name = unescapeHTML(e.get('name'))
                     info.update({
+                        'series': series_name,
+                    })
+                elif item_type == 'Movie':
+                    # here and in the next, don't erase existing value with None
+                    info.update(valued_dict({
                         'title': unescapeHTML(e.get('name')),
                         'description': unescapeHTML(e.get('description')),
                         'duration': parse_duration(e.get('duration')),
                         'timestamp': unified_timestamp(e.get('dateCreated')),
-                    })
+                    }))
                 elif item_type in ('Article', 'NewsArticle'):
-                    info.update({
+                    info.update(valued_dict({
                         'timestamp': parse_iso8601(e.get('datePublished')),
                         'title': unescapeHTML(e.get('headline')),
                         'description': unescapeHTML(e.get('articleBody')),
-                    })
+                    }))
                 elif item_type == 'VideoObject':
                     extract_video_object(e)
                     if expected_type is None:
@@ -1487,7 +1497,7 @@ class InfoExtractor(object):
                     continue
                 else:
                     break
-        return dict((k, v) for k, v in info.items() if v is not None)
+        return valued_dict(info)
 
     def _search_nextjs_data(self, webpage, video_id, **kw):
         # ..., *, transform_source=None, fatal=True, default=NO_DEFAULT

youtube_dl/extractor/extractors.py

@@ -1325,6 +1325,10 @@ from .trovo import (
     TrovoIE,
     TrovoVodIE,
 )
+from .trt import (
+    TRTIE,
+    TRTLiveIE,
+)
 from .trunews import TruNewsIE
 from .trutv import TruTVIE
 from .tube8 import Tube8IE

youtube_dl/extractor/trt.py

@@ -0,0 +1,325 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import json
+import re
+import time
+
+from .common import InfoExtractor
+
+from ..compat import (
+    compat_urlparse,
+)
+from ..utils import (
+    determine_ext,
+    dict_get,
+    merge_dicts,
+    parse_iso8601,
+    strip_or_none,
+    try_get,
+    url_or_none,
+    urljoin,
+)
+
+
+class TRTIE(InfoExtractor):
+    IE_DESC = 'TRT (Turkish State TV) programs and series'
+    _VALID_URL = r'''(?x)
+                     (?P<list>https?://(?:www\.)trtizle\.com/
+                         (?:diziler|programlar|belgesel|filmler|cocuk|trtarsiv|engelsiz)/
+                             (?P<show>[\w-]+))(?:/(?P<id>[\w-]+))?'''
+    _TESTS = [{
+        'url': 'https://www.trtizle.com/belgesel/dunya-tarihinin-donum-noktalari/dunya-tarihinin-donum-noktalari-1-bolum-125583',
+        'md5': 'c46dc0b9b53ad372c4ac6b3982805f05',
+        'info_dict': {
+            'id': 'dunya-tarihinin-donum-noktalari-1-bolum-125583',
+            'ext': 'mp4',
+            'title': 'Dünya Tarihinin Dönüm Noktaları 1.Bölüm',
+            'description': 'Bedelini insanların ödeyeceği bir imparatorluk çekişmesinde Persler, Yunanlara karşı...',
+            'timestamp': 1617148800,
+            'upload_date': '20210331',
+            'thumbnail': r're:https?://.+\.jpe?g',
+            'duration': float,
+            'series': 'Dünya Tarihinin Dönüm Noktaları',
+        },
+        'params': {
+            # adaptive download
+            'skip_download': True,
+        }
+    }, {
+        'url': 'https://www.trtizle.com/belgesel/dunya-tarihinin-donum-noktalari',
+        'info_dict': {
+            'id': 'dunya-tarihinin-donum-noktalari',
+            'title': 'Dünya Tarihinin Dönüm Noktaları',
+        },
+        'playlist_mincount': 22,
+    }, {
+        'url': 'https://www.trtizle.com/diziler/yol-ayrimi/yol-ayrimi-1-bolum-5774583',
+        'md5': '67ada6b2020b5dd0d3e24646b2725676',
+        'info_dict': {
+            'id': 'yol-ayrimi-1-bolum-5774583',
+            'ext': 'mp4',
+            'title': 'Yol Ayrımı 1.Bölüm',
+            'description': 'Seyrisefain balosunda, herkes bir haberin akıbetini beklemektedir…',
+            'timestamp': 1623888000,
+            'upload_date': '20210617',
+            'thumbnail': r're:https?://.+\.jpe?g',
+            'duration': float,
+            'series': 'Yol Ayrımı',
+        },
+        'params': {
+            # adaptive download
+            'skip_download': True,
+        },
+    }, {
+        'url': 'https://www.trtizle.com/diziler/yol-ayrimi/',
+        'info_dict': {
+            'id': 'yol-ayrimi',
+            'title': 'Yol Ayrımı',
+        },
+        'playlist_mincount': 5,
+    }, {
+        'url': 'https://www.trtizle.com/programlar/sade-saz/sade-saz-1-bolum-7646201',
+        'md5': '8f416e64379ea4d1d3ea0a65dc922f5c',
+        'info_dict': {
+            'id': 'sade-saz-1-bolum-7646201',
+            'ext': 'mp4',
+            'title': 'Sade Saz 1.Bölüm',
+            'description': 'Sade Saz’ın ilk bölümünün konuğu, tanbur icracısı K. Alper Uzkur.',
+            'timestamp': 1641772800,
+            'upload_date': '20220110',
+            'thumbnail': r're:https?://.+\.jpe?g',
+            'duration': float,
+            'series': 'Sade Saz',
+        },
+        'params': {
+            # adaptive download
+            'skip_download': True,
+        },
+    }, {
+        'url': 'https://www.trtizle.com/programlar/sade-saz',
+        'info_dict': {
+            'id': 'sade-saz',
+            'title': 'Sade Saz',
+        },
+        'playlist_mincount': 6,
+    }, {
+        'url': 'https://www.trtizle.com/filmler/looking-for-eric/looking-for-eric-8414201',
+        'md5': '833d61e4a10606d71b3903295cfa3c63',
+        'info_dict': {
+            'id': 'looking-for-eric-8414201',
+            'ext': 'mp4',
+            'title': 'Looking for Eric',
+            'description': 'Postacı Eric\'in hayatı krize sürüklenirken gerçek ve hayal birbirine karışır...',
+            'upload_date': '20220401',
+            'timestamp': 1648771200,
+            'thumbnail': r're:https?://.+\.jpe?g',
+            'duration': float,
+        },
+        'params': {
+            # adaptive download
+            'skip_download': True,
+        },
+    }, {
+        'url': 'https://www.trtizle.com/cocuk/kaptan-pengu-ve-arkadaslari/kaptan-pengu-ve-arkadaslari-okul-aciliyor-6034815',
+        'md5': '551c479d1a6bc7c538356907d4ea5d19',
+        'info_dict': {
+            'id': 'kaptan-pengu-ve-arkadaslari-okul-aciliyor-6034815',
+            'ext': 'mp4',
+            'title': 'Kaptan Pengu ve Arkadaşları 1.Bölüm',
+            'description': 'Hayvanlar Konseyi\'nden Kaptan Pengu\'ya bir mektup vardır...',
+            'timestamp': 1626134400,
+            'upload_date': '20210713',
+            'thumbnail': r're:https?://.+\.jpe?g',
+            'duration': float,
+            'series': 'Kaptan Pengu ve Arkadaşları',
+        },
+        'params': {
+            # adaptive download
+            'skip_download': True,
+        },
+    }, {
+        'url': 'https://www.trtizle.com/cocuk/kaptan-pengu-ve-arkadaslari',
+        'info_dict': {
+            'id': 'kaptan-pengu-ve-arkadaslari',
+            'title': 'Kaptan Pengu ve Arkadaşları',
+        },
+        'playlist_mincount': 41,
+    },
+    ]
+
+    def _extract_formats(self, fmt_url, video_id):
+        formats = []
+        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', fatal=False))
+        elif ext == 'mpd':
+            formats.extend(self._extract_mpd_formats(
+                fmt_url, video_id, mpd_id='dash', fatal=False))
+        else:
+            formats.append({
+                'url': fmt_url,
+            })
+        return formats
+
+    def _extract_list(self, playlist_id, url):
+        webpage = self._download_webpage(url, playlist_id)
+        LIST_RE = (
+            r'''<a\s[^>]*?\b%s\s*=\s*['"](%s(?:(?<=/)|/)[\w-]+)'''
+            % ('data-path' if 'data-path' in webpage else 'href',
+               re.escape(compat_urlparse.urlparse(url).path), ))
+
+        def entries():
+            for item_url in re.finditer(LIST_RE, webpage):
+                item_url = urljoin(url, item_url.group(1))
+                yield self._extract_video(self._match_id(item_url), item_url)
+
+        series = self._search_json_ld(webpage, playlist_id, default={}, expected_type='TVSeries')
+        return self.playlist_result(entries(), playlist_id, series.get('series'))
+
+    def _extract_video(self, video_id, url):
+        webpage = self._download_webpage(url, video_id)
+        result = self._search_json_ld(webpage, video_id, default={})
+        result['id'] = video_id
+        if 'title' not in result:
+            result['title'] = (
+                self._html_search_meta(('title', 'og:title', 'twitter:title'), webpage)
+                or self._html_search_regex(r'<title\b[^>]*>([^<]+)</title\b', webpage, 'title'))
+        fmt_url = result.get('url')
+        formats = []
+        if fmt_url:
+            del result['url']
+            formats = self._extract_formats(fmt_url, video_id)
+        self._sort_formats(formats)
+        result['formats'] = formats
+
+        return merge_dicts(
+            result, {
+                'description': self._html_search_meta(('description', 'og:description'), webpage, 'description'),
+                'thumbnail': url_or_none(self._og_search_thumbnail(webpage)),
+            })
+
+    def _real_extract(self, url):
+        show_id, video_id, playlist_url = re.match(self._VALID_URL, url).group('show', 'id', 'list')
+        # TODO: adapt --yes/no-playlist to make this work properly
+        # if not video_id or self._downloader.params.get('noplaylist') is False:
+        if not video_id:
+            return self._extract_list(show_id, playlist_url)
+
+        return self._extract_video(video_id, url)
+
+
+class TRTLiveIE(TRTIE):
+    IE_DESC = 'TRT (Turkish State TV and radio) live channels'
+    _VALID_URL = r'https?://(?:www\.)?trtizle\.com/canli/(?:tv/trt-|radyo/(?:radyo-)?)(?P<id>[\w-]+)'
+    _TESTS = [{
+        'url': 'https://www.trtizle.com/canli/tv/trt-world',
+        'info_dict': {
+            'id': 'trtworld',
+            'ext': 'mp4',
+            'title': r're:TRT WORLD .+',
+            'description': 'TRT World',
+            'is_live': True,
+        },
+        'params': {
+            # adaptive download
+            'skip_download': True,
+        }
+    },
+    ]
+
+    def _real_extract(self, url):
+        chan_id = self._match_id(url)
+        webpage = self._download_webpage(url, chan_id)
+        chan_id = self._search_regex(
+            r'\blivePlayer\s*\.\s*openPlayer\s*\([^)]*?\btrt\.com\.tr/trtportal/(?:[^/]+/)+thumbnails/([\w-]+)\.(?:jp|png)',
+            webpage, 'slug', fatal=False) or chan_id
+        chan_url = self._search_regex(
+            r'''\blivePlayerConfig\s*\.\s*baseEpgUrl\s*=\s*(?P<q>'|")(?P<url>https?://(?:(?!(?P=q)).)+)(?P=q)''',
+            webpage, 'player config', group='url')
+        chan_url = '%s%s.json' % (chan_url, chan_id)
+
+        def maybe_xml2json(src):
+            """Turn unexpected XML returned from an API URL into JSON"""
+            m = re.match(r'''^\s*<\?xml\b(?:[^/>]*?\bencoding\s*=\s*['"](?P<enc>[\w-]+))?[^/>]*\?>\s*(?P<xml><.+>)$''', src)
+            if m:
+
+                # Thanks https://stackoverflow.com/a/63556250 for inspiration
+                ATTR_RE = (
+                    r"""(?s)(?P<avr>\S ?)(?:\s*=\s*(?P<q>['"])(?P<avl>.*?)(?<!\\)(?P=q))?"""
+                )
+
+                def elt_value(attr_str, val_dict):
+                    v = {}
+                    attrs = dict((j.group("avr"), j.groupdict(True).get("avl"))
+                                 for j in re.finditer(ATTR_RE, attr_str.strip()))
+                    if attrs:
+                        v['@attributes'] = attrs
+                    v['@values'] = val_dict
+                    return v
+
+                def xml2dict(xml_str):
+                    elts = re.findall(
+                        r"(?s)<(?P<var>\S )(?P<attr>[^/>]*)(?:(?:>(?P<val>.*?)</(?P=var)>)|(?:/>))",
+                        xml_str,
+                    )
+
+                    if elts:
+                        elts = [{i[0]: elt_value(i[1], xml2dict(i[2]))} for i in elts]
+                        if len(elts) == 1:
+                            return elts[0]
+                        return elts
+                    return xml_str
+
+                try:
+                    return json.dumps(xml2dict(m.group('xml').encode(m.group('enc') or 'utf-8')))
+                except Exception:
+                    pass
+            return src
+
+        chan_info = self._download_json(
+            chan_url, chan_id, fatal=False,
+            note='Downloading player EPG JSON',
+            query={'_': int(time.time() * 1000)},
+            expected_status=403,
+            # errors are returned as XML
+            transform_source=maybe_xml2json)
+        if not isinstance(chan_info, dict) or 'Error' in chan_info:
+            chan_info = self._download_json(
+                'https://trtizle-api.cdn.wp.trt.com.tr/trttv/v3/livestream',
+                chan_id, fatal=False,
+                note='Downloading livestream API JSON',
+                query={'path': compat_urlparse.urlparse(url).path}) or {}
+
+        title = chan_info['channel']['title']
+
+        current = try_get(chan_info, lambda x: x['current'], dict) or {}
+        if current.get('geo_block'):
+            self._downloader.report_warning(
+                '[%s] %s' % (self.IE_NAME, 'Stream is geo-blocked'))
+
+        chan_info = chan_info['channel']
+        fmt_url = dict_get(chan_info, ('url', 'noneDvrUrl'))
+        formats = []
+        if fmt_url:
+            formats = self._extract_formats(fmt_url, chan_id)
+        self._sort_formats(formats)
+
+        start_end = [parse_iso8601(current.get(x)) for x in ('starttime', 'endtime')]
+        if None in start_end:
+            start_end = None
+
+        return {
+            'id': chan_id,
+            'title': self._live_title(current.get('title') or title),
+            'is_live': True,
+            'formats': formats,
+            'description': strip_or_none(chan_info.get('description')),
+            'thumbnail': next((url_or_none(chan_info.get(x))
+                              for x in ('thumbnail', 'thumbnailYoutubeUrl', 'square_logo', 'livestreamLogoUrl')),
+                              None),
+            'timestamp': start_end and start_end[0],
+            'duration': start_end and (start_end[1] - time.time()),
+        }

youtube_dl/extractor/youtube.py

@@ -692,9 +692,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         'invidious': '|'.join(_INVIDIOUS_SITES),
     }
     _PLAYER_INFO_RE = (
-        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$',
-        r'\b(?P<id>vfl[a-zA-Z0-9_-]+)\b.*?\.js$',
+        r'/s/player/(?P<id>[a-zA-Z0-9_-]{8,})/(?:tv-)?player',
+        r'/(?P<id>[a-zA-Z0-9_-]{8,})/player(?:_ias(?:_tce)?\.vflset(?:/[a-zA-Z]{2,3}_[a-zA-Z]{2,3})?|-plasma-ias-(?:phone|tablet)-[a-z]{2}_[A-Z]{2}\.vflset)/base\.js$',
+        r'\b(?P<id>vfl[a-zA-Z0-9_-]{6,})\b.*?\.js$',
     )
     _SUBTITLE_FORMATS = ('json3', 'srv1', 'srv2', 'srv3', 'ttml', 'vtt')
 
@@ -1626,15 +1626,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         """ Return a string representation of a signature """
         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
 
-    @classmethod
-    def _extract_player_info(cls, player_url):
-        for player_re in cls._PLAYER_INFO_RE:
-            id_m = re.search(player_re, player_url)
-            if id_m:
-                break
-        else:
-            raise ExtractorError('Cannot identify player %r' % player_url)
-        return id_m.group('id')
+    def _extract_player_info(self, player_url):
+        try:
+            return self._search_regex(
+                self._PLAYER_INFO_RE, player_url, 'player info', group='id')
+        except ExtractorError as e:
+            raise ExtractorError(
+                'Cannot identify player %r' % (player_url,), cause=e)
 
     def _load_player(self, video_id, player_url, fatal=True, player_id=None):
         if not player_id:
@@ -1711,6 +1709,23 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 '    return %s\n') % (signature_id_tuple, expr_code)
         self.to_screen('Extracted signature function:\n' + code)
 
+    def _extract_sig_fn(self, jsi, funcname):
+        var_ay = self._search_regex(
+            r'''(?x)
+                (?:\*/|\{|\n|^)\s*(?:'[^']+'\s*;\s*)
+                    (var\s*[\w$]+\s*=\s*(?:
+                        ('|")(?:\\\2|(?!\2).)+\2\s*\.\s*split\(\s*('|")\W+\3\s*\)|
+                        \[\s*(?:('|")(?:\\\4|(?!\4).)*\4\s*(?:(?=\])|,\s*))+\]
+                    ))(?=\s*[,;])
+            ''', jsi.code, 'useful values', default='')
+
+        sig_fn = jsi.extract_function_code(funcname)
+
+        if var_ay:
+            sig_fn = (sig_fn[0], ';\n'.join((var_ay, sig_fn[1])))
+
+        return sig_fn
+
     def _parse_sig_js(self, jscode):
         # Examples where `sig` is funcname:
         # sig=function(a){a=a.split(""); ... ;return a.join("")};
@@ -1736,8 +1751,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             jscode, 'Initial JS player signature function name', group='sig')
 
         jsi = JSInterpreter(jscode)
-        initial_function = jsi.extract_function(funcname)
-        return lambda s: initial_function([s])
+
+        initial_function = self._extract_sig_fn(jsi, funcname)
+
+        func = jsi.extract_function_from_code(*initial_function)
+
+        return lambda s: func([s])
 
     def _cached(self, func, *cache_id):
         def inner(*args, **kwargs):
@@ -1856,15 +1875,9 @@ 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\("\W+"\))(?=\s*[,;])',
-            jsi.code, 'useful values', default='')
-
         func_name = self._extract_n_function_name(jsi.code)
 
-        func_code = jsi.extract_function_code(func_name)
-        if var_ay:
-            func_code = (func_code[0], ';\n'.join((var_ay, func_code[1])))
+        func_code = self._extract_sig_fn(jsi, func_name)
 
         if player_id:
             self.cache.store('youtube-nsig', player_id, func_code)
@@ -2136,7 +2149,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                     video_details = merge_dicts(*traverse_obj(
                         (player_response, api_player_response),
                         (Ellipsis, 'videoDetails', T(dict))))
-                    player_response.update(api_player_response or {})
+                    player_response.update(filter_dict(
+                        api_player_response or {}, cndn=lambda k, _: k != 'captions'))
                     player_response['videoDetails'] = video_details
 
         def is_agegated(playability):
@@ -2566,8 +2580,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         }
 
         pctr = traverse_obj(
-            player_response,
-            ('captions', 'playerCaptionsTracklistRenderer', T(dict)))
+            (player_response, api_player_response),
+            (Ellipsis, 'captions', 'playerCaptionsTracklistRenderer', T(dict)))
         if pctr:
             def process_language(container, base_url, lang_code, query):
                 lang_subs = []
@@ -2584,20 +2598,21 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             def process_subtitles():
                 subtitles = {}
                 for caption_track in traverse_obj(pctr, (
-                        'captionTracks', lambda _, v: v.get('baseUrl'))):
+                        Ellipsis, 'captionTracks', lambda _, v: (
+                            v.get('baseUrl') and v.get('languageCode')))):
                     base_url = self._yt_urljoin(caption_track['baseUrl'])
                     if not base_url:
                         continue
+                    lang_code = caption_track['languageCode']
                     if caption_track.get('kind') != 'asr':
-                        lang_code = caption_track.get('languageCode')
-                        if not lang_code:
-                            continue
                         process_language(
                             subtitles, base_url, lang_code, {})
                         continue
                     automatic_captions = {}
+                    process_language(
+                        automatic_captions, base_url, lang_code, {})
                     for translation_language in traverse_obj(pctr, (
-                            'translationLanguages', lambda _, v: v.get('languageCode'))):
+                            Ellipsis, 'translationLanguages', lambda _, v: v.get('languageCode'))):
                         translation_language_code = translation_language['languageCode']
                         process_language(
                             automatic_captions, base_url, translation_language_code,

youtube_dl/jsinterp.py

@@ -678,7 +678,7 @@ class JSInterpreter(object):
             return len(obj)
         try:
             return obj[int(idx)] if isinstance(obj, list) else obj[compat_str(idx)]
-        except (TypeError, KeyError, IndexError) as e:
+        except (TypeError, KeyError, IndexError, ValueError) as e:
             # allow_undefined is None gives correct behaviour
             if allow_undefined or (
                     allow_undefined is None and not isinstance(e, TypeError)):
@@ -1038,6 +1038,10 @@ class JSInterpreter(object):
                     left_val = self._index(left_val, idx)
             if isinstance(idx, float):
                 idx = int(idx)
+            if isinstance(left_val, list) and len(left_val) <= int_or_none(idx, default=-1):
+                # JS Array is a sparsely assignable list
+                # TODO: handle extreme sparsity without memory bloat, eg using auxiliary dict
+                left_val.extend((idx - len(left_val) + 1) * [JS_Undefined])
             left_val[idx] = self._operator(
                 m.group('op'), self._index(left_val, idx) if m.group('op') else None,
                 m.group('expr'), expr, local_vars, allow_recursion)
@@ -1204,9 +1208,10 @@ class JSInterpreter(object):
                 elif member == 'join':
                     assertion(isinstance(obj, list), 'must be applied on a list')
                     assertion(len(argvals) <= 1, 'takes at most one argument')
-                    return (',' if len(argvals) == 0 else argvals[0]).join(
-                        ('' if x in (None, JS_Undefined) else _js_toString(x))
-                        for x in obj)
+                    return (',' if len(argvals) == 0 or argvals[0] in (None, JS_Undefined)
+                            else argvals[0]).join(
+                                ('' if x in (None, JS_Undefined) else _js_toString(x))
+                                for x in obj)
                 elif member == 'reverse':
                     assertion(not argvals, 'does not take any arguments')
                     obj.reverse()
@@ -1364,19 +1369,21 @@ class JSInterpreter(object):
         code, _ = self._separate_at_paren(func_m.group('code'))  # refine the match
         return self.build_arglist(func_m.group('args')), code
 
-    def extract_function(self, funcname):
+    def extract_function(self, funcname, *global_stack):
         return function_with_repr(
-            self.extract_function_from_code(*self.extract_function_code(funcname)),
+            self.extract_function_from_code(*itertools.chain(
+                self.extract_function_code(funcname), global_stack)),
             'F<%s>' % (funcname,))
 
     def extract_function_from_code(self, argnames, code, *global_stack):
         local_vars = {}
 
+        start = None
         while True:
-            mobj = re.search(r'function\((?P<args>[^)]*)\)\s*{', code)
+            mobj = re.search(r'function\((?P<args>[^)]*)\)\s*{', code[start:])
             if mobj is None:
                 break
-            start, body_start = mobj.span()
+            start, body_start = ((start or 0) + x for x in mobj.span())
             body, remaining = self._separate_at_paren(code[body_start - 1:])
             name = self._named_object(local_vars, self.extract_function_from_code(
                 [x.strip() for x in mobj.group('args').split(',')],