]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/youtube.py
Prepare to upload.
[youtubedl] / youtube_dl / extractor / youtube.py
index eacaa5ecdb70d2a16748b4c2e58edc14d7d69484..ce2212a7caa13111e1d2835fa65be371a4937847 100644 (file)
@@ -29,7 +29,6 @@ from ..compat import (
 from ..utils import (
     bool_or_none,
     clean_html,
-    dict_get,
     error_to_compat_str,
     extract_attributes,
     ExtractorError,
@@ -71,6 +70,11 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
 
     _PLAYLIST_ID_RE = r'(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}'
 
+    _YOUTUBE_CLIENT_HEADERS = {
+        'x-youtube-client-name': '1',
+        'x-youtube-client-version': '1.20200609.04.02',
+    }
+
     def _set_language(self):
         self._set_cookie(
             '.youtube.com', 'PREF', 'f1=50000000&hl=en',
@@ -302,7 +306,8 @@ class YoutubeEntryListBaseInfoExtractor(YoutubeBaseInfoExtractor):
                         'https://youtube.com/%s' % mobj.group('more'), playlist_id,
                         'Downloading page #%s%s'
                         % (page_num, ' (retry #%d)' % count if count else ''),
-                        transform_source=uppercase_escape)
+                        transform_source=uppercase_escape,
+                        headers=self._YOUTUBE_CLIENT_HEADERS)
                     break
                 except ExtractorError as e:
                     if isinstance(e.cause, compat_HTTPError) and e.cause.code in (500, 503):
@@ -389,8 +394,15 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                             (?:www\.)?invidious\.drycat\.fr/|
                             (?:www\.)?tube\.poal\.co/|
                             (?:www\.)?vid\.wxzm\.sx/|
+                            (?:www\.)?yewtu\.be/|
                             (?:www\.)?yt\.elukerio\.org/|
                             (?:www\.)?yt\.lelux\.fi/|
+                            (?:www\.)?invidious\.ggc-project\.de/|
+                            (?:www\.)?yt\.maisputain\.ovh/|
+                            (?:www\.)?invidious\.13ad\.de/|
+                            (?:www\.)?invidious\.toot\.koeln/|
+                            (?:www\.)?invidious\.fdn\.fr/|
+                            (?:www\.)?watch\.nettohikari\.com/|
                             (?:www\.)?kgg2m7yk5aybusll\.onion/|
                             (?:www\.)?qklhadlycap4cnod\.onion/|
                             (?:www\.)?axqzx4s6s54s32yentfqojs3x5i7faxza6xo3ehd4bzzsg2ii4fv2iid\.onion/|
@@ -398,6 +410,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                             (?:www\.)?fz253lmuao3strwbfbmx46yu7acac2jz27iwtorgmbqlkurlclmancad\.onion/|
                             (?:www\.)?invidious\.l4qlywnpwqsluw65ts7md3khrivpirse744un3x7mlskqauz5pyuzgqd\.onion/|
                             (?:www\.)?owxfohz4kjyv25fvlqilyxast7inivgiktls3th44jhk3ej3i7ya\.b32\.i2p/|
+                            (?:www\.)?4l2dgddgsrkf2ous66i6seeyi6etzfgrue332grh2n7madpwopotugyd\.onion/|
                             youtube\.googleapis\.com/)                        # the various hostnames, with wildcard subdomains
                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
                          (?:                                                  # the various things that can precede the ID:
@@ -427,6 +440,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                      (?(1).+)?                                                # if we found the ID, everything can follow
                      $""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
+    _PLAYER_INFO_RE = (
+        r'/(?P<id>[a-zA-Z0-9_-]{8,})/player_ias\.vflset(?:/[a-zA-Z]{2,3}_[a-zA-Z]{2,3})?/base\.(?P<ext>[a-z]+)$',
+        r'\b(?P<id>vfl[a-zA-Z0-9_-]+)\b.*?\.(?P<ext>[a-z]+)$',
+    )
     _formats = {
         '5': {'ext': 'flv', 'width': 400, 'height': 240, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
         '6': {'ext': 'flv', 'width': 450, 'height': 270, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
@@ -570,7 +587,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'upload_date': '20120506',
                 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
                 'alt_title': 'I Love It (feat. Charli XCX)',
-                'description': 'md5:f3ceb5ef83a08d95b9d146f973157cc8',
+                'description': 'md5:19a2f98d9032b9311e686ed039564f63',
                 'tags': ['Icona Pop i love it', 'sweden', 'pop music', 'big beat records', 'big beat', 'charli',
                          'xcx', 'charli xcx', 'girls', 'hbo', 'i love it', "i don't care", 'icona', 'pop',
                          'iconic ep', 'iconic', 'love', 'it'],
@@ -685,12 +702,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'id': 'nfWlot6h_JM',
                 'ext': 'm4a',
                 'title': 'Taylor Swift - Shake It Off',
-                'description': 'md5:bec2185232c05479482cb5a9b82719bf',
+                'description': 'md5:307195cd21ff7fa352270fe884570ef0',
                 'duration': 242,
                 'uploader': 'TaylorSwiftVEVO',
                 'uploader_id': 'TaylorSwiftVEVO',
                 'upload_date': '20140818',
-                'creator': 'Taylor Swift',
             },
             'params': {
                 'youtube_include_dash_manifest': True,
@@ -755,11 +771,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'upload_date': '20100430',
                 'uploader_id': 'deadmau5',
                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/deadmau5',
-                'creator': 'deadmau5',
+                'creator': 'Dada Life, deadmau5',
                 'description': 'md5:12c56784b8032162bb936a5f76d55360',
                 'uploader': 'deadmau5',
                 'title': 'Deadmau5 - Some Chords (HD)',
-                'alt_title': 'Some Chords',
+                'alt_title': 'This Machine Kills Some Chords',
             },
             'expected_warnings': [
                 'DASH manifest missing',
@@ -1135,6 +1151,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'skip_download': True,
                 'youtube_include_dash_manifest': False,
             },
+            'skip': 'not actual anymore',
         },
         {
             # Youtube Music Auto-generated description
@@ -1145,8 +1162,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'title': 'Voyeur Girl',
                 'description': 'md5:7ae382a65843d6df2685993e90a8628f',
                 'upload_date': '20190312',
-                'uploader': 'Various Artists - Topic',
-                'uploader_id': 'UCVWKBi1ELZn0QX2CBLSkiyw',
+                'uploader': 'Stephen - Topic',
+                'uploader_id': 'UC-pWHpBjdGG69N9mM2auIAA',
                 'artist': 'Stephen',
                 'track': 'Voyeur Girl',
                 'album': 'it\'s too much love to know my dear',
@@ -1210,7 +1227,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'id': '-hcAI0g-f5M',
                 'ext': 'mp4',
                 'title': 'Put It On Me',
-                'description': 'md5:93c55acc682ae7b0c668f2e34e1c069e',
+                'description': 'md5:f6422397c07c4c907c6638e1fee380a5',
                 'upload_date': '20180426',
                 'uploader': 'Matt Maeson - Topic',
                 'uploader_id': 'UCnEkIGqtGcQMLk73Kp-Q5LQ',
@@ -1228,6 +1245,26 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             'url': 'https://www.youtubekids.com/watch?v=3b8nCWDgZ6Q',
             'only_matching': True,
         },
+        {
+            # invalid -> valid video id redirection
+            'url': 'DJztXj2GPfl',
+            'info_dict': {
+                'id': 'DJztXj2GPfk',
+                'ext': 'mp4',
+                'title': 'Panjabi MC - Mundian To Bach Ke (The Dictator Soundtrack)',
+                'description': 'md5:bf577a41da97918e94fa9798d9228825',
+                'upload_date': '20090125',
+                'uploader': 'Prochorowka',
+                'uploader_id': 'Prochorowka',
+                'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/Prochorowka',
+                'artist': 'Panjabi MC',
+                'track': 'Beware of the Boys (Mundian to Bach Ke) - Motivo Hi-Lectro Remix',
+                'album': 'Beware of the Boys (Mundian To Bach Ke)',
+            },
+            'params': {
+                'skip_download': True,
+            },
+        }
     ]
 
     def __init__(self, *args, **kwargs):
@@ -1254,14 +1291,18 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         """ Return a string representation of a signature """
         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
 
-    def _extract_signature_function(self, video_id, player_url, example_sig):
-        id_m = re.match(
-            r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player(?:-new)?|(?:/[a-z]{2,3}_[A-Z]{2})?/base)?\.(?P<ext>[a-z]+)$',
-            player_url)
-        if not id_m:
+    @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)
-        player_type = id_m.group('ext')
-        player_id = id_m.group('id')
+        return id_m.group('ext'), id_m.group('id')
+
+    def _extract_signature_function(self, video_id, player_url, example_sig):
+        player_type, player_id = self._extract_player_info(player_url)
 
         # Read from filesystem cache
         func_id = '%s_%s_%s' % (
@@ -1617,8 +1658,63 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         video_id = mobj.group(2)
         return video_id
 
+    def _extract_chapters_from_json(self, webpage, video_id, duration):
+        if not webpage:
+            return
+        player = self._parse_json(
+            self._search_regex(
+                r'RELATED_PLAYER_ARGS["\']\s*:\s*({.+})\s*,?\s*\n', webpage,
+                'player args', default='{}'),
+            video_id, fatal=False)
+        if not player or not isinstance(player, dict):
+            return
+        watch_next_response = player.get('watch_next_response')
+        if not isinstance(watch_next_response, compat_str):
+            return
+        response = self._parse_json(watch_next_response, video_id, fatal=False)
+        if not response or not isinstance(response, dict):
+            return
+        chapters_list = try_get(
+            response,
+            lambda x: x['playerOverlays']
+                       ['playerOverlayRenderer']
+                       ['decoratedPlayerBarRenderer']
+                       ['decoratedPlayerBarRenderer']
+                       ['playerBar']
+                       ['chapteredPlayerBarRenderer']
+                       ['chapters'],
+            list)
+        if not chapters_list:
+            return
+
+        def chapter_time(chapter):
+            return float_or_none(
+                try_get(
+                    chapter,
+                    lambda x: x['chapterRenderer']['timeRangeStartMillis'],
+                    int),
+                scale=1000)
+        chapters = []
+        for next_num, chapter in enumerate(chapters_list, start=1):
+            start_time = chapter_time(chapter)
+            if start_time is None:
+                continue
+            end_time = (chapter_time(chapters_list[next_num])
+                        if next_num < len(chapters_list) else duration)
+            if end_time is None:
+                continue
+            title = try_get(
+                chapter, lambda x: x['chapterRenderer']['title']['simpleText'],
+                compat_str)
+            chapters.append({
+                'start_time': start_time,
+                'end_time': end_time,
+                'title': title,
+            })
+        return chapters
+
     @staticmethod
-    def _extract_chapters(description, duration):
+    def _extract_chapters_from_description(description, duration):
         if not description:
             return None
         chapter_lines = re.findall(
@@ -1652,6 +1748,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             })
         return chapters
 
+    def _extract_chapters(self, webpage, description, video_id, duration):
+        return (self._extract_chapters_from_json(webpage, video_id, duration)
+                or self._extract_chapters_from_description(description, duration))
+
     def _real_extract(self, url):
         url, smuggled_data = unsmuggle_url(url, {})
 
@@ -1679,7 +1779,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
         # Get video webpage
         url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
-        video_webpage = self._download_webpage(url, video_id)
+        video_webpage, urlh = self._download_webpage_handle(url, video_id)
+
+        qs = compat_parse_qs(compat_urllib_parse_urlparse(urlh.geturl()).query)
+        video_id = qs.get('v', [None])[0] or video_id
 
         # Attempt to extract SWF player URL
         mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
@@ -1708,9 +1811,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         def extract_view_count(v_info):
             return int_or_none(try_get(v_info, lambda x: x['view_count'][0]))
 
-        def extract_token(v_info):
-            return dict_get(v_info, ('account_playback_token', 'accountPlaybackToken', 'token'))
-
         def extract_player_response(player_response, video_id):
             pl_response = str_or_none(player_response)
             if not pl_response:
@@ -1723,6 +1823,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         player_response = {}
 
         # Get video info
+        video_info = {}
         embed_webpage = None
         if re.search(r'player-age-gate-content">', video_webpage) is not None:
             age_gate = True
@@ -1737,19 +1838,21 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                     r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
             })
             video_info_url = proto + '://www.youtube.com/get_video_info?' + data
-            video_info_webpage = self._download_webpage(
-                video_info_url, video_id,
-                note='Refetching age-gated info webpage',
-                errnote='unable to download video info webpage')
-            video_info = compat_parse_qs(video_info_webpage)
-            pl_response = video_info.get('player_response', [None])[0]
-            player_response = extract_player_response(pl_response, video_id)
-            add_dash_mpd(video_info)
-            view_count = extract_view_count(video_info)
+            try:
+                video_info_webpage = self._download_webpage(
+                    video_info_url, video_id,
+                    note='Refetching age-gated info webpage',
+                    errnote='unable to download video info webpage')
+            except ExtractorError:
+                video_info_webpage = None
+            if video_info_webpage:
+                video_info = compat_parse_qs(video_info_webpage)
+                pl_response = video_info.get('player_response', [None])[0]
+                player_response = extract_player_response(pl_response, video_id)
+                add_dash_mpd(video_info)
+                view_count = extract_view_count(video_info)
         else:
             age_gate = False
-            video_info = None
-            sts = None
             # Try looking directly into the video webpage
             ytplayer_config = self._get_ytplayer_config(video_id, video_webpage)
             if ytplayer_config:
@@ -1766,61 +1869,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                         args['ypc_vid'], YoutubeIE.ie_key(), video_id=args['ypc_vid'])
                 if args.get('livestream') == '1' or args.get('live_playback') == 1:
                     is_live = True
-                sts = ytplayer_config.get('sts')
                 if not player_response:
                     player_response = extract_player_response(args.get('player_response'), video_id)
             if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
                 add_dash_mpd_pr(player_response)
-                # We also try looking in get_video_info since it may contain different dashmpd
-                # URL that points to a DASH manifest with possibly different itag set (some itags
-                # are missing from DASH manifest pointed by webpage's dashmpd, some - from DASH
-                # manifest pointed by get_video_info's dashmpd).
-                # The general idea is to take a union of itags of both DASH manifests (for example
-                # video with such 'manifest behavior' see https://github.com/ytdl-org/youtube-dl/issues/6093)
-                self.report_video_info_webpage_download(video_id)
-                for el in ('embedded', 'detailpage', 'vevo', ''):
-                    query = {
-                        'video_id': video_id,
-                        'ps': 'default',
-                        'eurl': '',
-                        'gl': 'US',
-                        'hl': 'en',
-                    }
-                    if el:
-                        query['el'] = el
-                    if sts:
-                        query['sts'] = sts
-                    video_info_webpage = self._download_webpage(
-                        '%s://www.youtube.com/get_video_info' % proto,
-                        video_id, note=False,
-                        errnote='unable to download video info webpage',
-                        fatal=False, query=query)
-                    if not video_info_webpage:
-                        continue
-                    get_video_info = compat_parse_qs(video_info_webpage)
-                    if not player_response:
-                        pl_response = get_video_info.get('player_response', [None])[0]
-                        player_response = extract_player_response(pl_response, video_id)
-                    add_dash_mpd(get_video_info)
-                    if view_count is None:
-                        view_count = extract_view_count(get_video_info)
-                    if not video_info:
-                        video_info = get_video_info
-                    get_token = extract_token(get_video_info)
-                    if get_token:
-                        # Different get_video_info requests may report different results, e.g.
-                        # some may report video unavailability, but some may serve it without
-                        # any complaint (see https://github.com/ytdl-org/youtube-dl/issues/7362,
-                        # the original webpage as well as el=info and el=embedded get_video_info
-                        # requests report video unavailability due to geo restriction while
-                        # el=detailpage succeeds and returns valid data). This is probably
-                        # due to YouTube measures against IP ranges of hosting providers.
-                        # Working around by preferring the first succeeded video_info containing
-                        # the token if no such video_info yet was found.
-                        token = extract_token(video_info)
-                        if not token:
-                            video_info = get_video_info
-                        break
 
         def extract_unavailable_message():
             messages = []
@@ -1833,16 +1885,22 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             if messages:
                 return '\n'.join(messages)
 
-        if not video_info:
+        if not video_info and not player_response:
             unavailable_message = extract_unavailable_message()
             if not unavailable_message:
                 unavailable_message = 'Unable to extract video data'
             raise ExtractorError(
                 'YouTube said: %s' % unavailable_message, expected=True, video_id=video_id)
 
+        if not isinstance(video_info, dict):
+            video_info = {}
+
         video_details = try_get(
             player_response, lambda x: x['videoDetails'], dict) or {}
 
+        microformat = try_get(
+            player_response, lambda x: x['microformat']['playerMicroformatRenderer'], dict) or {}
+
         video_title = video_info.get('title', [None])[0] or video_details.get('title')
         if not video_title:
             self._downloader.report_warning('Unable to extract video title')
@@ -1889,15 +1947,26 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                         # fields may contain comma as well (see
                         # https://github.com/ytdl-org/youtube-dl/issues/8536)
                         feed_data = compat_parse_qs(compat_urllib_parse_unquote_plus(feed))
+
+                        def feed_entry(name):
+                            return try_get(feed_data, lambda x: x[name][0], compat_str)
+
+                        feed_id = feed_entry('id')
+                        if not feed_id:
+                            continue
+                        feed_title = feed_entry('title')
+                        title = video_title
+                        if feed_title:
+                            title += ' (%s)' % feed_title
                         entries.append({
                             '_type': 'url_transparent',
                             'ie_key': 'Youtube',
                             'url': smuggle_url(
                                 '%s://www.youtube.com/watch?v=%s' % (proto, feed_data['id'][0]),
                                 {'force_singlefeed': True}),
-                            'title': '%s (%s)' % (video_title, feed_data['title'][0]),
+                            'title': title,
                         })
-                        feed_ids.append(feed_data['id'][0])
+                        feed_ids.append(feed_id)
                     self.to_screen(
                         'Downloading multifeed video (%s) - add --no-playlist to just download video %s'
                         % (', '.join(feed_ids), video_id))
@@ -1909,6 +1978,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             view_count = extract_view_count(video_info)
         if view_count is None and video_details:
             view_count = int_or_none(video_details.get('viewCount'))
+        if view_count is None and microformat:
+            view_count = int_or_none(microformat.get('viewCount'))
 
         if is_live is None:
             is_live = bool_or_none(video_details.get('isLive'))
@@ -1968,12 +2039,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 }
 
             for fmt in streaming_formats:
-                if fmt.get('drm_families'):
+                if fmt.get('drmFamilies') or fmt.get('drm_families'):
                     continue
                 url = url_or_none(fmt.get('url'))
 
                 if not url:
-                    cipher = fmt.get('cipher')
+                    cipher = fmt.get('cipher') or fmt.get('signatureCipher')
                     if not cipher:
                         continue
                     url_data = compat_parse_qs(cipher)
@@ -2024,22 +2095,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
                         if self._downloader.params.get('verbose'):
                             if player_url is None:
-                                player_version = 'unknown'
                                 player_desc = 'unknown'
                             else:
-                                if player_url.endswith('swf'):
-                                    player_version = self._search_regex(
-                                        r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
-                                        'flash player', fatal=False)
-                                    player_desc = 'flash player %s' % player_version
-                                else:
-                                    player_version = self._search_regex(
-                                        [r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js',
-                                         r'(?:www|player(?:_ias)?)-([^/]+)(?:/[a-z]{2,3}_[A-Z]{2})?/base\.js'],
-                                        player_url,
-                                        'html5 player', fatal=False)
-                                    player_desc = 'html5 player %s' % player_version
-
+                                player_type, player_version = self._extract_player_info(player_url)
+                                player_desc = '%s player %s' % ('flash' if player_type == 'swf' else 'html5', player_version)
                             parts_sizes = self._signature_cache_id(encrypted_sig)
                             self.to_screen('{%s} signature length %s, %s' %
                                            (format_id, parts_sizes, player_desc))
@@ -2172,7 +2231,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             video_uploader_id = mobj.group('uploader_id')
             video_uploader_url = mobj.group('uploader_url')
         else:
-            self._downloader.report_warning('unable to extract uploader nickname')
+            owner_profile_url = url_or_none(microformat.get('ownerProfileUrl'))
+            if owner_profile_url:
+                video_uploader_id = self._search_regex(
+                    r'(?:user|channel)/([^/]+)', owner_profile_url, 'uploader id',
+                    default=None)
+                video_uploader_url = owner_profile_url
 
         channel_id = (
             str_or_none(video_details.get('channelId'))
@@ -2183,17 +2247,33 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 video_webpage, 'channel id', default=None, group='id'))
         channel_url = 'http://www.youtube.com/channel/%s' % channel_id if channel_id else None
 
-        # thumbnail image
-        # We try first to get a high quality image:
-        m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
-                            video_webpage, re.DOTALL)
-        if m_thumb is not None:
-            video_thumbnail = m_thumb.group(1)
-        elif 'thumbnail_url' not in video_info:
-            self._downloader.report_warning('unable to extract video thumbnail')
+        thumbnails = []
+        thumbnails_list = try_get(
+            video_details, lambda x: x['thumbnail']['thumbnails'], list) or []
+        for t in thumbnails_list:
+            if not isinstance(t, dict):
+                continue
+            thumbnail_url = url_or_none(t.get('url'))
+            if not thumbnail_url:
+                continue
+            thumbnails.append({
+                'url': thumbnail_url,
+                'width': int_or_none(t.get('width')),
+                'height': int_or_none(t.get('height')),
+            })
+
+        if not thumbnails:
             video_thumbnail = None
-        else:   # don't panic if we can't find it
-            video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0])
+            # We try first to get a high quality image:
+            m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
+                                video_webpage, re.DOTALL)
+            if m_thumb is not None:
+                video_thumbnail = m_thumb.group(1)
+            thumbnail_url = try_get(video_info, lambda x: x['thumbnail_url'][0], compat_str)
+            if thumbnail_url:
+                video_thumbnail = compat_urllib_parse_unquote_plus(thumbnail_url)
+            if video_thumbnail:
+                thumbnails.append({'url': video_thumbnail})
 
         # upload date
         upload_date = self._html_search_meta(
@@ -2203,6 +2283,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 [r'(?s)id="eow-date.*?>(.*?)</span>',
                  r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'],
                 video_webpage, 'upload date', default=None)
+        if not upload_date:
+            upload_date = microformat.get('publishDate') or microformat.get('uploadDate')
         upload_date = unified_strdate(upload_date)
 
         video_license = self._html_search_regex(
@@ -2335,7 +2417,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                     errnote='Unable to download video annotations', fatal=False,
                     data=urlencode_postdata({xsrf_field_name: xsrf_token}))
 
-        chapters = self._extract_chapters(description_original, video_duration)
+        chapters = self._extract_chapters(video_webpage, description_original, video_id, video_duration)
 
         # Look for the DASH manifest
         if self._downloader.params.get('youtube_include_dash_manifest', True):
@@ -2392,30 +2474,23 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                         f['stretched_ratio'] = ratio
 
         if not formats:
-            token = extract_token(video_info)
-            if not token:
-                if 'reason' in video_info:
-                    if 'The uploader has not made this video available in your country.' in video_info['reason']:
-                        regions_allowed = self._html_search_meta(
-                            'regionsAllowed', video_webpage, default=None)
-                        countries = regions_allowed.split(',') if regions_allowed else None
-                        self.raise_geo_restricted(
-                            msg=video_info['reason'][0], countries=countries)
-                    reason = video_info['reason'][0]
-                    if 'Invalid parameters' in reason:
-                        unavailable_message = extract_unavailable_message()
-                        if unavailable_message:
-                            reason = unavailable_message
-                    raise ExtractorError(
-                        'YouTube said: %s' % reason,
-                        expected=True, video_id=video_id)
-                else:
-                    raise ExtractorError(
-                        '"token" parameter not in video info for unknown reason',
-                        video_id=video_id)
-
-        if not formats and (video_info.get('license_info') or try_get(player_response, lambda x: x['streamingData']['licenseInfos'])):
-            raise ExtractorError('This video is DRM protected.', expected=True)
+            if 'reason' in video_info:
+                if 'The uploader has not made this video available in your country.' in video_info['reason']:
+                    regions_allowed = self._html_search_meta(
+                        'regionsAllowed', video_webpage, default=None)
+                    countries = regions_allowed.split(',') if regions_allowed else None
+                    self.raise_geo_restricted(
+                        msg=video_info['reason'][0], countries=countries)
+                reason = video_info['reason'][0]
+                if 'Invalid parameters' in reason:
+                    unavailable_message = extract_unavailable_message()
+                    if unavailable_message:
+                        reason = unavailable_message
+                raise ExtractorError(
+                    'YouTube said: %s' % reason,
+                    expected=True, video_id=video_id)
+            if video_info.get('license_info') or try_get(player_response, lambda x: x['streamingData']['licenseInfos']):
+                raise ExtractorError('This video is DRM protected.', expected=True)
 
         self._sort_formats(formats)
 
@@ -2433,7 +2508,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             'creator': video_creator or artist,
             'title': video_title,
             'alt_title': video_alt_title or track,
-            'thumbnail': video_thumbnail,
+            'thumbnails': thumbnails,
             'description': video_description,
             'categories': video_categories,
             'tags': video_tags,
@@ -2495,20 +2570,23 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
     _VIDEO_RE = _VIDEO_RE_TPL % r'(?P<id>[0-9A-Za-z_-]{11})'
     IE_NAME = 'youtube:playlist'
     _TESTS = [{
-        'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
+        'url': 'https://www.youtube.com/playlist?list=PL4lCao7KL_QFVb7Iudeipvc2BCavECqzc',
         'info_dict': {
-            'title': 'ytdl test PL',
-            'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
+            'uploader_id': 'UCmlqkdCBesrv2Lak1mF_MxA',
+            'uploader': 'Sergey M.',
+            'id': 'PL4lCao7KL_QFVb7Iudeipvc2BCavECqzc',
+            'title': 'youtube-dl public playlist',
         },
-        'playlist_count': 3,
+        'playlist_count': 1,
     }, {
-        'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
+        'url': 'https://www.youtube.com/playlist?list=PL4lCao7KL_QFodcLWhDpGCYnngnHtQ-Xf',
         'info_dict': {
-            'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
-            'title': 'YDL_Empty_List',
+            'uploader_id': 'UCmlqkdCBesrv2Lak1mF_MxA',
+            'uploader': 'Sergey M.',
+            'id': 'PL4lCao7KL_QFodcLWhDpGCYnngnHtQ-Xf',
+            'title': 'youtube-dl empty playlist',
         },
         'playlist_count': 0,
-        'skip': 'This playlist is private',
     }, {
         'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
         'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
@@ -2518,7 +2596,7 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
             'uploader': 'Christiaan008',
             'uploader_id': 'ChRiStIaAn008',
         },
-        'playlist_count': 95,
+        'playlist_count': 96,
     }, {
         'note': 'issue #673',
         'url': 'PLBB231211A4F62143',
@@ -3206,7 +3284,8 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
             more = self._download_json(
                 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE,
                 'Downloading page #%s' % page_num,
-                transform_source=uppercase_escape)
+                transform_source=uppercase_escape,
+                headers=self._YOUTUBE_CLIENT_HEADERS)
             content_html = more['content_html']
             more_widget_html = more['load_more_widget_html']