_ Git - youtube-dl/blob - youtube_dl/extractor/youtube.py

   1 # coding: utf-8
   2
   3 from __future__ import unicode_literals
   4
   5
   6 import itertools
   7 import json
   8 import os.path
   9 import random
  10 import re
  11 import time
  12 import traceback
  13
  14 from .common import InfoExtractor, SearchInfoExtractor
  15 from ..jsinterp import JSInterpreter
  16 from ..swfinterp import SWFInterpreter
  17 from ..compat import (
  18     compat_chr,
  19     compat_HTTPError,
  20     compat_kwargs,
  21     compat_parse_qs,
  22     compat_urllib_parse_unquote,
  23     compat_urllib_parse_unquote_plus,
  24     compat_urllib_parse_urlencode,
  25     compat_urllib_parse_urlparse,
  26     compat_urlparse,
  27     compat_str,
  28 )
  29 from ..utils import (
  30     bool_or_none,
  31     clean_html,
  32     dict_get,
  33     error_to_compat_str,
  34     ExtractorError,
  35     float_or_none,
  36     get_element_by_attribute,
  37     get_element_by_id,
  38     int_or_none,
  39     mimetype2ext,
  40     orderedSet,
  41     parse_codecs,
  42     parse_duration,
  43     qualities,
  44     remove_quotes,
  45     remove_start,
  46     smuggle_url,
  47     str_or_none,
  48     str_to_int,
  49     try_get,
  50     unescapeHTML,
  51     unified_strdate,
  52     unsmuggle_url,
  53     uppercase_escape,
  54     url_or_none,
  55     urlencode_postdata,
  56 )
  57
  58
  59 class YoutubeBaseInfoExtractor(InfoExtractor):
  60     """Provide base functions for Youtube extractors"""
  61     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  62     _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge'
  63
  64     _LOOKUP_URL = 'https://accounts.google.com/_/signin/sl/lookup'
  65     _CHALLENGE_URL = 'https://accounts.google.com/_/signin/sl/challenge'
  66     _TFA_URL = 'https://accounts.google.com/_/signin/challenge?hl=en&TL={0}'
  67
  68     _NETRC_MACHINE = 'youtube'
  69     # If True it will raise an error if no login info is provided
  70     _LOGIN_REQUIRED = False
  71
  72     _PLAYLIST_ID_RE = r'(?:PL|LL|EC|UU|FL|RD|UL|TL|OLAK5uy_)[0-9A-Za-z-_]{10,}'
  73
  74     def _set_language(self):
  75         self._set_cookie(
  76             '.youtube.com', 'PREF', 'f1=50000000&hl=en',
  77             # YouTube sets the expire time to about two months
  78             expire_time=time.time() + 2 * 30 * 24 * 3600)
  79
  80     def _ids_to_results(self, ids):
  81         return [
  82             self.url_result(vid_id, 'Youtube', video_id=vid_id)
  83             for vid_id in ids]
  84
  85     def _login(self):
  86         """
  87         Attempt to log in to YouTube.
  88         True is returned if successful or skipped.
  89         False is returned if login failed.
  90
  91         If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
  92         """
  93         username, password = self._get_login_info()
  94         # No authentication to be performed
  95         if username is None:
  96             if self._LOGIN_REQUIRED and self._downloader.params.get('cookiefile') is None:
  97                 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  98             return True
  99
 100         login_page = self._download_webpage(
 101             self._LOGIN_URL, None,
 102             note='Downloading login page',
 103             errnote='unable to fetch login page', fatal=False)
 104         if login_page is False:
 105             return
 106
 107         login_form = self._hidden_inputs(login_page)
 108
 109         def req(url, f_req, note, errnote):
 110             data = login_form.copy()
 111             data.update({
 112                 'pstMsg': 1,
 113                 'checkConnection': 'youtube',
 114                 'checkedDomains': 'youtube',
 115                 'hl': 'en',
 116                 'deviceinfo': '[null,null,null,[],null,"US",null,null,[],"GlifWebSignIn",null,[null,null,[]]]',
 117                 'f.req': json.dumps(f_req),
 118                 'flowName': 'GlifWebSignIn',
 119                 'flowEntry': 'ServiceLogin',
 120                 # TODO: reverse actual botguard identifier generation algo
 121                 'bgRequest': '["identifier",""]',
 122             })
 123             return self._download_json(
 124                 url, None, note=note, errnote=errnote,
 125                 transform_source=lambda s: re.sub(r'^[^[]*', '', s),
 126                 fatal=False,
 127                 data=urlencode_postdata(data), headers={
 128                     'Content-Type': 'application/x-www-form-urlencoded;charset=utf-8',
 129                     'Google-Accounts-XSRF': 1,
 130                 })
 131
 132         def warn(message):
 133             self._downloader.report_warning(message)
 134
 135         lookup_req = [
 136             username,
 137             None, [], None, 'US', None, None, 2, False, True,
 138             [
 139                 None, None,
 140                 [2, 1, None, 1,
 141                  'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn',
 142                  None, [], 4],
 143                 1, [None, None, []], None, None, None, True
 144             ],
 145             username,
 146         ]
 147
 148         lookup_results = req(
 149             self._LOOKUP_URL, lookup_req,
 150             'Looking up account info', 'Unable to look up account info')
 151
 152         if lookup_results is False:
 153             return False
 154
 155         user_hash = try_get(lookup_results, lambda x: x[0][2], compat_str)
 156         if not user_hash:
 157             warn('Unable to extract user hash')
 158             return False
 159
 160         challenge_req = [
 161             user_hash,
 162             None, 1, None, [1, None, None, None, [password, None, True]],
 163             [
 164                 None, None, [2, 1, None, 1, 'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn', None, [], 4],
 165                 1, [None, None, []], None, None, None, True
 166             ]]
 167
 168         challenge_results = req(
 169             self._CHALLENGE_URL, challenge_req,
 170             'Logging in', 'Unable to log in')
 171
 172         if challenge_results is False:
 173             return
 174
 175         login_res = try_get(challenge_results, lambda x: x[0][5], list)
 176         if login_res:
 177             login_msg = try_get(login_res, lambda x: x[5], compat_str)
 178             warn(
 179                 'Unable to login: %s' % 'Invalid password'
 180                 if login_msg == 'INCORRECT_ANSWER_ENTERED' else login_msg)
 181             return False
 182
 183         res = try_get(challenge_results, lambda x: x[0][-1], list)
 184         if not res:
 185             warn('Unable to extract result entry')
 186             return False
 187
 188         login_challenge = try_get(res, lambda x: x[0][0], list)
 189         if login_challenge:
 190             challenge_str = try_get(login_challenge, lambda x: x[2], compat_str)
 191             if challenge_str == 'TWO_STEP_VERIFICATION':
 192                 # SEND_SUCCESS - TFA code has been successfully sent to phone
 193                 # QUOTA_EXCEEDED - reached the limit of TFA codes
 194                 status = try_get(login_challenge, lambda x: x[5], compat_str)
 195                 if status == 'QUOTA_EXCEEDED':
 196                     warn('Exceeded the limit of TFA codes, try later')
 197                     return False
 198
 199                 tl = try_get(challenge_results, lambda x: x[1][2], compat_str)
 200                 if not tl:
 201                     warn('Unable to extract TL')
 202                     return False
 203
 204                 tfa_code = self._get_tfa_info('2-step verification code')
 205
 206                 if not tfa_code:
 207                     warn(
 208                         'Two-factor authentication required. Provide it either interactively or with --twofactor <code>'
 209                         '(Note that only TOTP (Google Authenticator App) codes work at this time.)')
 210                     return False
 211
 212                 tfa_code = remove_start(tfa_code, 'G-')
 213
 214                 tfa_req = [
 215                     user_hash, None, 2, None,
 216                     [
 217                         9, None, None, None, None, None, None, None,
 218                         [None, tfa_code, True, 2]
 219                     ]]
 220
 221                 tfa_results = req(
 222                     self._TFA_URL.format(tl), tfa_req,
 223                     'Submitting TFA code', 'Unable to submit TFA code')
 224
 225                 if tfa_results is False:
 226                     return False
 227
 228                 tfa_res = try_get(tfa_results, lambda x: x[0][5], list)
 229                 if tfa_res:
 230                     tfa_msg = try_get(tfa_res, lambda x: x[5], compat_str)
 231                     warn(
 232                         'Unable to finish TFA: %s' % 'Invalid TFA code'
 233                         if tfa_msg == 'INCORRECT_ANSWER_ENTERED' else tfa_msg)
 234                     return False
 235
 236                 check_cookie_url = try_get(
 237                     tfa_results, lambda x: x[0][-1][2], compat_str)
 238             else:
 239                 CHALLENGES = {
 240                     'LOGIN_CHALLENGE': "This device isn't recognized. For your security, Google wants to make sure it's really you.",
 241                     'USERNAME_RECOVERY': 'Please provide additional information to aid in the recovery process.',
 242                     'REAUTH': "There is something unusual about your activity. For your security, Google wants to make sure it's really you.",
 243                 }
 244                 challenge = CHALLENGES.get(
 245                     challenge_str,
 246                     '%s returned error %s.' % (self.IE_NAME, challenge_str))
 247                 warn('%s\nGo to https://accounts.google.com/, login and solve a challenge.' % challenge)
 248                 return False
 249         else:
 250             check_cookie_url = try_get(res, lambda x: x[2], compat_str)
 251
 252         if not check_cookie_url:
 253             warn('Unable to extract CheckCookie URL')
 254             return False
 255
 256         check_cookie_results = self._download_webpage(
 257             check_cookie_url, None, 'Checking cookie', fatal=False)
 258
 259         if check_cookie_results is False:
 260             return False
 261
 262         if 'https://myaccount.google.com/' not in check_cookie_results:
 263             warn('Unable to log in')
 264             return False
 265
 266         return True
 267
 268     def _download_webpage_handle(self, *args, **kwargs):
 269         query = kwargs.get('query', {}).copy()
 270         query['disable_polymer'] = 'true'
 271         kwargs['query'] = query
 272         return super(YoutubeBaseInfoExtractor, self)._download_webpage_handle(
 273             *args, **compat_kwargs(kwargs))
 274
 275     def _real_initialize(self):
 276         if self._downloader is None:
 277             return
 278         self._set_language()
 279         if not self._login():
 280             return
 281
 282
 283 class YoutubeEntryListBaseInfoExtractor(YoutubeBaseInfoExtractor):
 284     # Extract entries from page with "Load more" button
 285     def _entries(self, page, playlist_id):
 286         more_widget_html = content_html = page
 287         for page_num in itertools.count(1):
 288             for entry in self._process_page(content_html):
 289                 yield entry
 290
 291             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
 292             if not mobj:
 293                 break
 294
 295             count = 0
 296             retries = 3
 297             while count <= retries:
 298                 try:
 299                     # Downloading page may result in intermittent 5xx HTTP error
 300                     # that is usually worked around with a retry
 301                     more = self._download_json(
 302                         'https://youtube.com/%s' % mobj.group('more'), playlist_id,
 303                         'Downloading page #%s%s'
 304                         % (page_num, ' (retry #%d)' % count if count else ''),
 305                         transform_source=uppercase_escape)
 306                     break
 307                 except ExtractorError as e:
 308                     if isinstance(e.cause, compat_HTTPError) and e.cause.code in (500, 503):
 309                         count += 1
 310                         if count <= retries:
 311                             continue
 312                     raise
 313
 314             content_html = more['content_html']
 315             if not content_html.strip():
 316                 # Some webpages show a "Load more" button but they don't
 317                 # have more videos
 318                 break
 319             more_widget_html = more['load_more_widget_html']
 320
 321
 322 class YoutubePlaylistBaseInfoExtractor(YoutubeEntryListBaseInfoExtractor):
 323     def _process_page(self, content):
 324         for video_id, video_title in self.extract_videos_from_page(content):
 325             yield self.url_result(video_id, 'Youtube', video_id, video_title)
 326
 327     def extract_videos_from_page(self, page):
 328         ids_in_page = []
 329         titles_in_page = []
 330         for mobj in re.finditer(self._VIDEO_RE, page):
 331             # The link with index 0 is not the first video of the playlist (not sure if still actual)
 332             if 'index' in mobj.groupdict() and mobj.group('id') == '0':
 333                 continue
 334             video_id = mobj.group('id')
 335             video_title = unescapeHTML(mobj.group('title'))
 336             if video_title:
 337                 video_title = video_title.strip()
 338             try:
 339                 idx = ids_in_page.index(video_id)
 340                 if video_title and not titles_in_page[idx]:
 341                     titles_in_page[idx] = video_title
 342             except ValueError:
 343                 ids_in_page.append(video_id)
 344                 titles_in_page.append(video_title)
 345         return zip(ids_in_page, titles_in_page)
 346
 347
 348 class YoutubePlaylistsBaseInfoExtractor(YoutubeEntryListBaseInfoExtractor):
 349     def _process_page(self, content):
 350         for playlist_id in orderedSet(re.findall(
 351                 r'<h3[^>]+class="[^"]*yt-lockup-title[^"]*"[^>]*><a[^>]+href="/?playlist\?list=([0-9A-Za-z-_]{10,})"',
 352                 content)):
 353             yield self.url_result(
 354                 'https://www.youtube.com/playlist?list=%s' % playlist_id, 'YoutubePlaylist')
 355
 356     def _real_extract(self, url):
 357         playlist_id = self._match_id(url)
 358         webpage = self._download_webpage(url, playlist_id)
 359         title = self._og_search_title(webpage, fatal=False)
 360         return self.playlist_result(self._entries(webpage, playlist_id), playlist_id, title)
 361
 362
 363 class YoutubeIE(YoutubeBaseInfoExtractor):
 364     IE_DESC = 'YouTube.com'
 365     _VALID_URL = r"""(?x)^
 366                      (
 367                          (?:https?://|//)                                    # http(s):// or protocol-independent URL
 368                          (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
 369                             (?:www\.)?deturl\.com/www\.youtube\.com/|
 370                             (?:www\.)?pwnyoutube\.com/|
 371                             (?:www\.)?hooktube\.com/|
 372                             (?:www\.)?yourepeat\.com/|
 373                             tube\.majestyc\.net/|
 374                             (?:(?:www|dev)\.)?invidio\.us/|
 375                             (?:www\.)?invidiou\.sh/|
 376                             (?:www\.)?invidious\.snopyta\.org/|
 377                             (?:www\.)?invidious\.kabi\.tk/|
 378                             (?:www\.)?vid\.wxzm\.sx/|
 379                             youtube\.googleapis\.com/)                        # the various hostnames, with wildcard subdomains
 380                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
 381                          (?:                                                  # the various things that can precede the ID:
 382                              (?:(?:v|embed|e)/(?!videoseries))                # v/ or embed/ or e/
 383                              |(?:                                             # or the v= param in all its forms
 384                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)?  # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
 385                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
 386                                  (?:.*?[&;])??                                # any other preceding param (like /?s=tuff&v=xxxx or ?s=tuff&amp;v=V36LpHqtcDY)
 387                                  v=
 388                              )
 389                          ))
 390                          |(?:
 391                             youtu\.be|                                        # just youtu.be/xxxx
 392                             vid\.plus|                                        # or vid.plus/xxxx
 393                             zwearz\.com/watch|                                # or zwearz.com/watch/xxxx
 394                          )/
 395                          |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
 396                          )
 397                      )?                                                       # all until now is optional -> you can pass the naked ID
 398                      ([0-9A-Za-z_-]{11})                                      # here is it! the YouTube video ID
 399                      (?!.*?\blist=
 400                         (?:
 401                             %(playlist_id)s|                                  # combined list/video URLs are handled by the playlist IE
 402                             WL                                                # WL are handled by the watch later IE
 403                         )
 404                      )
 405                      (?(1).+)?                                                # if we found the ID, everything can follow
 406                      $""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
 407     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
 408     _formats = {
 409         '5': {'ext': 'flv', 'width': 400, 'height': 240, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
 410         '6': {'ext': 'flv', 'width': 450, 'height': 270, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
 411         '13': {'ext': '3gp', 'acodec': 'aac', 'vcodec': 'mp4v'},
 412         '17': {'ext': '3gp', 'width': 176, 'height': 144, 'acodec': 'aac', 'abr': 24, 'vcodec': 'mp4v'},
 413         '18': {'ext': 'mp4', 'width': 640, 'height': 360, 'acodec': 'aac', 'abr': 96, 'vcodec': 'h264'},
 414         '22': {'ext': 'mp4', 'width': 1280, 'height': 720, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 415         '34': {'ext': 'flv', 'width': 640, 'height': 360, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 416         '35': {'ext': 'flv', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 417         # itag 36 videos are either 320x180 (BaW_jenozKc) or 320x240 (__2ABJjxzNo), abr varies as well
 418         '36': {'ext': '3gp', 'width': 320, 'acodec': 'aac', 'vcodec': 'mp4v'},
 419         '37': {'ext': 'mp4', 'width': 1920, 'height': 1080, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 420         '38': {'ext': 'mp4', 'width': 4096, 'height': 3072, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 421         '43': {'ext': 'webm', 'width': 640, 'height': 360, 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8'},
 422         '44': {'ext': 'webm', 'width': 854, 'height': 480, 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8'},
 423         '45': {'ext': 'webm', 'width': 1280, 'height': 720, 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8'},
 424         '46': {'ext': 'webm', 'width': 1920, 'height': 1080, 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8'},
 425         '59': {'ext': 'mp4', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 426         '78': {'ext': 'mp4', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 427
 428
 429         # 3D videos
 430         '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -20},
 431         '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -20},
 432         '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264', 'preference': -20},
 433         '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264', 'preference': -20},
 434         '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8', 'preference': -20},
 435         '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8', 'preference': -20},
 436         '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8', 'preference': -20},
 437
 438         # Apple HTTP Live Streaming
 439         '91': {'ext': 'mp4', 'height': 144, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 440         '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 441         '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -10},
 442         '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -10},
 443         '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 256, 'vcodec': 'h264', 'preference': -10},
 444         '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 256, 'vcodec': 'h264', 'preference': -10},
 445         '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 446         '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 24, 'vcodec': 'h264', 'preference': -10},
 447
 448         # DASH mp4 video
 449         '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'h264'},
 450         '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'h264'},
 451         '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'h264'},
 452         '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'h264'},
 453         '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'h264'},
 454         '138': {'ext': 'mp4', 'format_note': 'DASH video', 'vcodec': 'h264'},  # Height can vary (https://github.com/ytdl-org/youtube-dl/issues/4559)
 455         '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'vcodec': 'h264'},
 456         '212': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'h264'},
 457         '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'h264'},
 458         '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'h264', 'fps': 60},
 459         '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'h264', 'fps': 60},
 460         '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'h264'},
 461
 462         # Dash mp4 audio
 463         '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 48, 'container': 'm4a_dash'},
 464         '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 128, 'container': 'm4a_dash'},
 465         '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 256, 'container': 'm4a_dash'},
 466         '256': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'container': 'm4a_dash'},
 467         '258': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'container': 'm4a_dash'},
 468         '325': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'dtse', 'container': 'm4a_dash'},
 469         '328': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'ec-3', 'container': 'm4a_dash'},
 470
 471         # Dash webm
 472         '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 473         '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 474         '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 475         '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 476         '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 477         '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 478         '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp9'},
 479         '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 480         '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 481         '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 482         '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 483         '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 484         '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 485         '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 486         '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 487         # itag 272 videos are either 3840x2160 (e.g. RtoitU2A-3E) or 7680x4320 (sLprVF6d7Ug)
 488         '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 489         '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 490         '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 491         '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 492         '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 493         '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 494
 495         # Dash webm audio
 496         '171': {'ext': 'webm', 'acodec': 'vorbis', 'format_note': 'DASH audio', 'abr': 128},
 497         '172': {'ext': 'webm', 'acodec': 'vorbis', 'format_note': 'DASH audio', 'abr': 256},
 498
 499         # Dash webm audio with opus inside
 500         '249': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50},
 501         '250': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70},
 502         '251': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160},
 503
 504         # RTMP (unnamed)
 505         '_rtmp': {'protocol': 'rtmp'},
 506
 507         # av01 video only formats sometimes served with "unknown" codecs
 508         '394': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 509         '395': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 510         '396': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 511         '397': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 512     }
 513     _SUBTITLE_FORMATS = ('srv1', 'srv2', 'srv3', 'ttml', 'vtt')
 514
 515     _GEO_BYPASS = False
 516
 517     IE_NAME = 'youtube'
 518     _TESTS = [
 519         {
 520             'url': 'https://www.youtube.com/watch?v=BaW_jenozKc&t=1s&end=9',
 521             'info_dict': {
 522                 'id': 'BaW_jenozKc',
 523                 'ext': 'mp4',
 524                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 525                 'uploader': 'Philipp Hagemeister',
 526                 'uploader_id': 'phihag',
 527                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/phihag',
 528                 'channel_id': 'UCLqxVugv74EIW3VWh2NOa3Q',
 529                 'channel_url': r're:https?://(?:www\.)?youtube\.com/channel/UCLqxVugv74EIW3VWh2NOa3Q',
 530                 'upload_date': '20121002',
 531                 'description': 'test chars:  "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
 532                 'categories': ['Science & Technology'],
 533                 'tags': ['youtube-dl'],
 534                 'duration': 10,
 535                 'view_count': int,
 536                 'like_count': int,
 537                 'dislike_count': int,
 538                 'start_time': 1,
 539                 'end_time': 9,
 540             }
 541         },
 542         {
 543             'url': 'https://www.youtube.com/watch?v=UxxajLWwzqY',
 544             'note': 'Test generic use_cipher_signature video (#897)',
 545             'info_dict': {
 546                 'id': 'UxxajLWwzqY',
 547                 'ext': 'mp4',
 548                 'upload_date': '20120506',
 549                 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
 550                 'alt_title': 'I Love It (feat. Charli XCX)',
 551                 'description': 'md5:f3ceb5ef83a08d95b9d146f973157cc8',
 552                 'tags': ['Icona Pop i love it', 'sweden', 'pop music', 'big beat records', 'big beat', 'charli',
 553                          'xcx', 'charli xcx', 'girls', 'hbo', 'i love it', "i don't care", 'icona', 'pop',
 554                          'iconic ep', 'iconic', 'love', 'it'],
 555                 'duration': 180,
 556                 'uploader': 'Icona Pop',
 557                 'uploader_id': 'IconaPop',
 558                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/IconaPop',
 559                 'creator': 'Icona Pop',
 560                 'track': 'I Love It (feat. Charli XCX)',
 561                 'artist': 'Icona Pop',
 562             }
 563         },
 564         {
 565             'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
 566             'note': 'Test VEVO video with age protection (#956)',
 567             'info_dict': {
 568                 'id': '07FYdnEawAQ',
 569                 'ext': 'mp4',
 570                 'upload_date': '20130703',
 571                 'title': 'Justin Timberlake - Tunnel Vision (Official Music Video) (Explicit)',
 572                 'alt_title': 'Tunnel Vision',
 573                 'description': 'md5:07dab3356cde4199048e4c7cd93471e1',
 574                 'duration': 419,
 575                 'uploader': 'justintimberlakeVEVO',
 576                 'uploader_id': 'justintimberlakeVEVO',
 577                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/justintimberlakeVEVO',
 578                 'creator': 'Justin Timberlake',
 579                 'track': 'Tunnel Vision',
 580                 'artist': 'Justin Timberlake',
 581                 'age_limit': 18,
 582             }
 583         },
 584         {
 585             'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
 586             'note': 'Embed-only video (#1746)',
 587             'info_dict': {
 588                 'id': 'yZIXLfi8CZQ',
 589                 'ext': 'mp4',
 590                 'upload_date': '20120608',
 591                 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
 592                 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
 593                 'uploader': 'SET India',
 594                 'uploader_id': 'setindia',
 595                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/setindia',
 596                 'age_limit': 18,
 597             }
 598         },
 599         {
 600             'url': 'https://www.youtube.com/watch?v=BaW_jenozKc&v=UxxajLWwzqY',
 601             'note': 'Use the first video ID in the URL',
 602             'info_dict': {
 603                 'id': 'BaW_jenozKc',
 604                 'ext': 'mp4',
 605                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 606                 'uploader': 'Philipp Hagemeister',
 607                 'uploader_id': 'phihag',
 608                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/phihag',
 609                 'upload_date': '20121002',
 610                 'description': 'test chars:  "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
 611                 'categories': ['Science & Technology'],
 612                 'tags': ['youtube-dl'],
 613                 'duration': 10,
 614                 'view_count': int,
 615                 'like_count': int,
 616                 'dislike_count': int,
 617             },
 618             'params': {
 619                 'skip_download': True,
 620             },
 621         },
 622         {
 623             'url': 'https://www.youtube.com/watch?v=a9LDPn-MO4I',
 624             'note': '256k DASH audio (format 141) via DASH manifest',
 625             'info_dict': {
 626                 'id': 'a9LDPn-MO4I',
 627                 'ext': 'm4a',
 628                 'upload_date': '20121002',
 629                 'uploader_id': '8KVIDEO',
 630                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/8KVIDEO',
 631                 'description': '',
 632                 'uploader': '8KVIDEO',
 633                 'title': 'UHDTV TEST 8K VIDEO.mp4'
 634             },
 635             'params': {
 636                 'youtube_include_dash_manifest': True,
 637                 'format': '141',
 638             },
 639             'skip': 'format 141 not served anymore',
 640         },
 641         # DASH manifest with encrypted signature
 642         {
 643             'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
 644             'info_dict': {
 645                 'id': 'IB3lcPjvWLA',
 646                 'ext': 'm4a',
 647                 'title': 'Afrojack, Spree Wilson - The Spark (Official Music Video) ft. Spree Wilson',
 648                 'description': 'md5:8f5e2b82460520b619ccac1f509d43bf',
 649                 'duration': 244,
 650                 'uploader': 'AfrojackVEVO',
 651                 'uploader_id': 'AfrojackVEVO',
 652                 'upload_date': '20131011',
 653             },
 654             'params': {
 655                 'youtube_include_dash_manifest': True,
 656                 'format': '141/bestaudio[ext=m4a]',
 657             },
 658         },
 659         # JS player signature function name containing $
 660         {
 661             'url': 'https://www.youtube.com/watch?v=nfWlot6h_JM',
 662             'info_dict': {
 663                 'id': 'nfWlot6h_JM',
 664                 'ext': 'm4a',
 665                 'title': 'Taylor Swift - Shake It Off',
 666                 'description': 'md5:bec2185232c05479482cb5a9b82719bf',
 667                 'duration': 242,
 668                 'uploader': 'TaylorSwiftVEVO',
 669                 'uploader_id': 'TaylorSwiftVEVO',
 670                 'upload_date': '20140818',
 671                 'creator': 'Taylor Swift',
 672             },
 673             'params': {
 674                 'youtube_include_dash_manifest': True,
 675                 'format': '141/bestaudio[ext=m4a]',
 676             },
 677         },
 678         # Controversy video
 679         {
 680             'url': 'https://www.youtube.com/watch?v=T4XJQO3qol8',
 681             'info_dict': {
 682                 'id': 'T4XJQO3qol8',
 683                 'ext': 'mp4',
 684                 'duration': 219,
 685                 'upload_date': '20100909',
 686                 'uploader': 'Amazing Atheist',
 687                 'uploader_id': 'TheAmazingAtheist',
 688                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/TheAmazingAtheist',
 689                 'title': 'Burning Everyone\'s Koran',
 690                 'description': 'SUBSCRIBE: http://www.youtube.com/saturninefilms\n\nEven Obama has taken a stand against freedom on this issue: http://www.huffingtonpost.com/2010/09/09/obama-gma-interview-quran_n_710282.html',
 691             }
 692         },
 693         # Normal age-gate video (No vevo, embed allowed)
 694         {
 695             'url': 'https://youtube.com/watch?v=HtVdAasjOgU',
 696             'info_dict': {
 697                 'id': 'HtVdAasjOgU',
 698                 'ext': 'mp4',
 699                 'title': 'The Witcher 3: Wild Hunt - The Sword Of Destiny Trailer',
 700                 'description': r're:(?s).{100,}About the Game\n.*?The Witcher 3: Wild Hunt.{100,}',
 701                 'duration': 142,
 702                 'uploader': 'The Witcher',
 703                 'uploader_id': 'WitcherGame',
 704                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/WitcherGame',
 705                 'upload_date': '20140605',
 706                 'age_limit': 18,
 707             },
 708         },
 709         # Age-gate video with encrypted signature
 710         {
 711             'url': 'https://www.youtube.com/watch?v=6kLq3WMV1nU',
 712             'info_dict': {
 713                 'id': '6kLq3WMV1nU',
 714                 'ext': 'mp4',
 715                 'title': 'Dedication To My Ex (Miss That) (Lyric Video)',
 716                 'description': 'md5:33765bb339e1b47e7e72b5490139bb41',
 717                 'duration': 246,
 718                 'uploader': 'LloydVEVO',
 719                 'uploader_id': 'LloydVEVO',
 720                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/LloydVEVO',
 721                 'upload_date': '20110629',
 722                 'age_limit': 18,
 723             },
 724         },
 725         # video_info is None (https://github.com/ytdl-org/youtube-dl/issues/4421)
 726         # YouTube Red ad is not captured for creator
 727         {
 728             'url': '__2ABJjxzNo',
 729             'info_dict': {
 730                 'id': '__2ABJjxzNo',
 731                 'ext': 'mp4',
 732                 'duration': 266,
 733                 'upload_date': '20100430',
 734                 'uploader_id': 'deadmau5',
 735                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/deadmau5',
 736                 'creator': 'deadmau5',
 737                 'description': 'md5:12c56784b8032162bb936a5f76d55360',
 738                 'uploader': 'deadmau5',
 739                 'title': 'Deadmau5 - Some Chords (HD)',
 740                 'alt_title': 'Some Chords',
 741             },
 742             'expected_warnings': [
 743                 'DASH manifest missing',
 744             ]
 745         },
 746         # Olympics (https://github.com/ytdl-org/youtube-dl/issues/4431)
 747         {
 748             'url': 'lqQg6PlCWgI',
 749             'info_dict': {
 750                 'id': 'lqQg6PlCWgI',
 751                 'ext': 'mp4',
 752                 'duration': 6085,
 753                 'upload_date': '20150827',
 754                 'uploader_id': 'olympic',
 755                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/olympic',
 756                 'description': 'HO09  - Women -  GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games',
 757                 'uploader': 'Olympic',
 758                 'title': 'Hockey - Women -  GER-AUS - London 2012 Olympic Games',
 759             },
 760             'params': {
 761                 'skip_download': 'requires avconv',
 762             }
 763         },
 764         # Non-square pixels
 765         {
 766             'url': 'https://www.youtube.com/watch?v=_b-2C3KPAM0',
 767             'info_dict': {
 768                 'id': '_b-2C3KPAM0',
 769                 'ext': 'mp4',
 770                 'stretched_ratio': 16 / 9.,
 771                 'duration': 85,
 772                 'upload_date': '20110310',
 773                 'uploader_id': 'AllenMeow',
 774                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/AllenMeow',
 775                 'description': 'made by Wacom from Korea | 字幕&加油添醋 by TY\'s Allen | 感謝heylisa00cavey1001同學熱情提供梗及翻譯',
 776                 'uploader': '孫ᄋᄅ',
 777                 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人',
 778             },
 779         },
 780         # url_encoded_fmt_stream_map is empty string
 781         {
 782             'url': 'qEJwOuvDf7I',
 783             'info_dict': {
 784                 'id': 'qEJwOuvDf7I',
 785                 'ext': 'webm',
 786                 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге',
 787                 'description': '',
 788                 'upload_date': '20150404',
 789                 'uploader_id': 'spbelect',
 790                 'uploader': 'Наблюдатели Петербурга',
 791             },
 792             'params': {
 793                 'skip_download': 'requires avconv',
 794             },
 795             'skip': 'This live event has ended.',
 796         },
 797         # Extraction from multiple DASH manifests (https://github.com/ytdl-org/youtube-dl/pull/6097)
 798         {
 799             'url': 'https://www.youtube.com/watch?v=FIl7x6_3R5Y',
 800             'info_dict': {
 801                 'id': 'FIl7x6_3R5Y',
 802                 'ext': 'webm',
 803                 'title': 'md5:7b81415841e02ecd4313668cde88737a',
 804                 'description': 'md5:116377fd2963b81ec4ce64b542173306',
 805                 'duration': 220,
 806                 'upload_date': '20150625',
 807                 'uploader_id': 'dorappi2000',
 808                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/dorappi2000',
 809                 'uploader': 'dorappi2000',
 810                 'formats': 'mincount:31',
 811             },
 812             'skip': 'not actual anymore',
 813         },
 814         # DASH manifest with segment_list
 815         {
 816             'url': 'https://www.youtube.com/embed/CsmdDsKjzN8',
 817             'md5': '8ce563a1d667b599d21064e982ab9e31',
 818             'info_dict': {
 819                 'id': 'CsmdDsKjzN8',
 820                 'ext': 'mp4',
 821                 'upload_date': '20150501',  # According to '<meta itemprop="datePublished"', but in other places it's 20150510
 822                 'uploader': 'Airtek',
 823                 'description': 'Retransmisión en directo de la XVIII media maratón de Zaragoza.',
 824                 'uploader_id': 'UCzTzUmjXxxacNnL8I3m4LnQ',
 825                 'title': 'Retransmisión XVIII Media maratón Zaragoza 2015',
 826             },
 827             'params': {
 828                 'youtube_include_dash_manifest': True,
 829                 'format': '135',  # bestvideo
 830             },
 831             'skip': 'This live event has ended.',
 832         },
 833         {
 834             # Multifeed videos (multiple cameras), URL is for Main Camera
 835             'url': 'https://www.youtube.com/watch?v=jqWvoWXjCVs',
 836             'info_dict': {
 837                 'id': 'jqWvoWXjCVs',
 838                 'title': 'teamPGP: Rocket League Noob Stream',
 839                 'description': 'md5:dc7872fb300e143831327f1bae3af010',
 840             },
 841             'playlist': [{
 842                 'info_dict': {
 843                     'id': 'jqWvoWXjCVs',
 844                     'ext': 'mp4',
 845                     'title': 'teamPGP: Rocket League Noob Stream (Main Camera)',
 846                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 847                     'duration': 7335,
 848                     'upload_date': '20150721',
 849                     'uploader': 'Beer Games Beer',
 850                     'uploader_id': 'beergamesbeer',
 851                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 852                     'license': 'Standard YouTube License',
 853                 },
 854             }, {
 855                 'info_dict': {
 856                     'id': '6h8e8xoXJzg',
 857                     'ext': 'mp4',
 858                     'title': 'teamPGP: Rocket League Noob Stream (kreestuh)',
 859                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 860                     'duration': 7337,
 861                     'upload_date': '20150721',
 862                     'uploader': 'Beer Games Beer',
 863                     'uploader_id': 'beergamesbeer',
 864                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 865                     'license': 'Standard YouTube License',
 866                 },
 867             }, {
 868                 'info_dict': {
 869                     'id': 'PUOgX5z9xZw',
 870                     'ext': 'mp4',
 871                     'title': 'teamPGP: Rocket League Noob Stream (grizzle)',
 872                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 873                     'duration': 7337,
 874                     'upload_date': '20150721',
 875                     'uploader': 'Beer Games Beer',
 876                     'uploader_id': 'beergamesbeer',
 877                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 878                     'license': 'Standard YouTube License',
 879                 },
 880             }, {
 881                 'info_dict': {
 882                     'id': 'teuwxikvS5k',
 883                     'ext': 'mp4',
 884                     'title': 'teamPGP: Rocket League Noob Stream (zim)',
 885                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 886                     'duration': 7334,
 887                     'upload_date': '20150721',
 888                     'uploader': 'Beer Games Beer',
 889                     'uploader_id': 'beergamesbeer',
 890                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 891                     'license': 'Standard YouTube License',
 892                 },
 893             }],
 894             'params': {
 895                 'skip_download': True,
 896             },
 897             'skip': 'This video is not available.',
 898         },
 899         {
 900             # Multifeed video with comma in title (see https://github.com/ytdl-org/youtube-dl/issues/8536)
 901             'url': 'https://www.youtube.com/watch?v=gVfLd0zydlo',
 902             'info_dict': {
 903                 'id': 'gVfLd0zydlo',
 904                 'title': 'DevConf.cz 2016 Day 2 Workshops 1 14:00 - 15:30',
 905             },
 906             'playlist_count': 2,
 907             'skip': 'Not multifeed anymore',
 908         },
 909         {
 910             'url': 'https://vid.plus/FlRa-iH7PGw',
 911             'only_matching': True,
 912         },
 913         {
 914             'url': 'https://zwearz.com/watch/9lWxNJF-ufM/electra-woman-dyna-girl-official-trailer-grace-helbig.html',
 915             'only_matching': True,
 916         },
 917         {
 918             # Title with JS-like syntax "};" (see https://github.com/ytdl-org/youtube-dl/issues/7468)
 919             # Also tests cut-off URL expansion in video description (see
 920             # https://github.com/ytdl-org/youtube-dl/issues/1892,
 921             # https://github.com/ytdl-org/youtube-dl/issues/8164)
 922             'url': 'https://www.youtube.com/watch?v=lsguqyKfVQg',
 923             'info_dict': {
 924                 'id': 'lsguqyKfVQg',
 925                 'ext': 'mp4',
 926                 'title': '{dark walk}; Loki/AC/Dishonored; collab w/Elflover21',
 927                 'alt_title': 'Dark Walk - Position Music',
 928                 'description': 'md5:8085699c11dc3f597ce0410b0dcbb34a',
 929                 'duration': 133,
 930                 'upload_date': '20151119',
 931                 'uploader_id': 'IronSoulElf',
 932                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/IronSoulElf',
 933                 'uploader': 'IronSoulElf',
 934                 'creator': 'Todd Haberman,  Daniel Law Heath and Aaron Kaplan',
 935                 'track': 'Dark Walk - Position Music',
 936                 'artist': 'Todd Haberman,  Daniel Law Heath and Aaron Kaplan',
 937                 'album': 'Position Music - Production Music Vol. 143 - Dark Walk',
 938             },
 939             'params': {
 940                 'skip_download': True,
 941             },
 942         },
 943         {
 944             # Tags with '};' (see https://github.com/ytdl-org/youtube-dl/issues/7468)
 945             'url': 'https://www.youtube.com/watch?v=Ms7iBXnlUO8',
 946             'only_matching': True,
 947         },
 948         {
 949             # Video with yt:stretch=17:0
 950             'url': 'https://www.youtube.com/watch?v=Q39EVAstoRM',
 951             'info_dict': {
 952                 'id': 'Q39EVAstoRM',
 953                 'ext': 'mp4',
 954                 'title': 'Clash Of Clans#14 Dicas De Ataque Para CV 4',
 955                 'description': 'md5:ee18a25c350637c8faff806845bddee9',
 956                 'upload_date': '20151107',
 957                 'uploader_id': 'UCCr7TALkRbo3EtFzETQF1LA',
 958                 'uploader': 'CH GAMER DROID',
 959             },
 960             'params': {
 961                 'skip_download': True,
 962             },
 963             'skip': 'This video does not exist.',
 964         },
 965         {
 966             # Video licensed under Creative Commons
 967             'url': 'https://www.youtube.com/watch?v=M4gD1WSo5mA',
 968             'info_dict': {
 969                 'id': 'M4gD1WSo5mA',
 970                 'ext': 'mp4',
 971                 'title': 'md5:e41008789470fc2533a3252216f1c1d1',
 972                 'description': 'md5:a677553cf0840649b731a3024aeff4cc',
 973                 'duration': 721,
 974                 'upload_date': '20150127',
 975                 'uploader_id': 'BerkmanCenter',
 976                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/BerkmanCenter',
 977                 'uploader': 'The Berkman Klein Center for Internet & Society',
 978                 'license': 'Creative Commons Attribution license (reuse allowed)',
 979             },
 980             'params': {
 981                 'skip_download': True,
 982             },
 983         },
 984         {
 985             # Channel-like uploader_url
 986             'url': 'https://www.youtube.com/watch?v=eQcmzGIKrzg',
 987             'info_dict': {
 988                 'id': 'eQcmzGIKrzg',
 989                 'ext': 'mp4',
 990                 'title': 'Democratic Socialism and Foreign Policy | Bernie Sanders',
 991                 'description': 'md5:dda0d780d5a6e120758d1711d062a867',
 992                 'duration': 4060,
 993                 'upload_date': '20151119',
 994                 'uploader': 'Bernie Sanders',
 995                 'uploader_id': 'UCH1dpzjCEiGAt8CXkryhkZg',
 996                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCH1dpzjCEiGAt8CXkryhkZg',
 997                 'license': 'Creative Commons Attribution license (reuse allowed)',
 998             },
 999             'params': {
1000                 'skip_download': True,
1001             },
1002         },
1003         {
1004             'url': 'https://www.youtube.com/watch?feature=player_embedded&amp;amp;v=V36LpHqtcDY',
1005             'only_matching': True,
1006         },
1007         {
1008             # YouTube Red paid video (https://github.com/ytdl-org/youtube-dl/issues/10059)
1009             'url': 'https://www.youtube.com/watch?v=i1Ko8UG-Tdo',
1010             'only_matching': True,
1011         },
1012         {
1013             # Rental video preview
1014             'url': 'https://www.youtube.com/watch?v=yYr8q0y5Jfg',
1015             'info_dict': {
1016                 'id': 'uGpuVWrhIzE',
1017                 'ext': 'mp4',
1018                 'title': 'Piku - Trailer',
1019                 'description': 'md5:c36bd60c3fd6f1954086c083c72092eb',
1020                 'upload_date': '20150811',
1021                 'uploader': 'FlixMatrix',
1022                 'uploader_id': 'FlixMatrixKaravan',
1023                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/FlixMatrixKaravan',
1024                 'license': 'Standard YouTube License',
1025             },
1026             'params': {
1027                 'skip_download': True,
1028             },
1029             'skip': 'This video is not available.',
1030         },
1031         {
1032             # YouTube Red video with episode data
1033             'url': 'https://www.youtube.com/watch?v=iqKdEhx-dD4',
1034             'info_dict': {
1035                 'id': 'iqKdEhx-dD4',
1036                 'ext': 'mp4',
1037                 'title': 'Isolation - Mind Field (Ep 1)',
1038                 'description': 'md5:46a29be4ceffa65b92d277b93f463c0f',
1039                 'duration': 2085,
1040                 'upload_date': '20170118',
1041                 'uploader': 'Vsauce',
1042                 'uploader_id': 'Vsauce',
1043                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/Vsauce',
1044                 'series': 'Mind Field',
1045                 'season_number': 1,
1046                 'episode_number': 1,
1047             },
1048             'params': {
1049                 'skip_download': True,
1050             },
1051             'expected_warnings': [
1052                 'Skipping DASH manifest',
1053             ],
1054         },
1055         {
1056             # The following content has been identified by the YouTube community
1057             # as inappropriate or offensive to some audiences.
1058             'url': 'https://www.youtube.com/watch?v=6SJNVb0GnPI',
1059             'info_dict': {
1060                 'id': '6SJNVb0GnPI',
1061                 'ext': 'mp4',
1062                 'title': 'Race Differences in Intelligence',
1063                 'description': 'md5:5d161533167390427a1f8ee89a1fc6f1',
1064                 'duration': 965,
1065                 'upload_date': '20140124',
1066                 'uploader': 'New Century Foundation',
1067                 'uploader_id': 'UCEJYpZGqgUob0zVVEaLhvVg',
1068                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCEJYpZGqgUob0zVVEaLhvVg',
1069             },
1070             'params': {
1071                 'skip_download': True,
1072             },
1073         },
1074         {
1075             # itag 212
1076             'url': '1t24XAntNCY',
1077             'only_matching': True,
1078         },
1079         {
1080             # geo restricted to JP
1081             'url': 'sJL6WA-aGkQ',
1082             'only_matching': True,
1083         },
1084         {
1085             'url': 'https://www.youtube.com/watch?v=MuAGGZNfUkU&list=RDMM',
1086             'only_matching': True,
1087         },
1088         {
1089             'url': 'https://invidio.us/watch?v=BaW_jenozKc',
1090             'only_matching': True,
1091         },
1092         {
1093             # DRM protected
1094             'url': 'https://www.youtube.com/watch?v=s7_qI6_mIXc',
1095             'only_matching': True,
1096         },
1097         {
1098             # Video with unsupported adaptive stream type formats
1099             'url': 'https://www.youtube.com/watch?v=Z4Vy8R84T1U',
1100             'info_dict': {
1101                 'id': 'Z4Vy8R84T1U',
1102                 'ext': 'mp4',
1103                 'title': 'saman SMAN 53 Jakarta(Sancety) opening COFFEE4th at SMAN 53 Jakarta',
1104                 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
1105                 'duration': 433,
1106                 'upload_date': '20130923',
1107                 'uploader': 'Amelia Putri Harwita',
1108                 'uploader_id': 'UCpOxM49HJxmC1qCalXyB3_Q',
1109                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCpOxM49HJxmC1qCalXyB3_Q',
1110                 'formats': 'maxcount:10',
1111             },
1112             'params': {
1113                 'skip_download': True,
1114                 'youtube_include_dash_manifest': False,
1115             },
1116         },
1117         {
1118             # Youtube Music Auto-generated description
1119             'url': 'https://music.youtube.com/watch?v=MgNrAu2pzNs',
1120             'info_dict': {
1121                 'id': 'MgNrAu2pzNs',
1122                 'ext': 'mp4',
1123                 'title': 'Voyeur Girl',
1124                 'description': 'md5:7ae382a65843d6df2685993e90a8628f',
1125                 'upload_date': '20190312',
1126                 'uploader': 'Various Artists - Topic',
1127                 'uploader_id': 'UCVWKBi1ELZn0QX2CBLSkiyw',
1128                 'artist': 'Stephen',
1129                 'track': 'Voyeur Girl',
1130                 'album': 'it\'s too much love to know my dear',
1131                 'release_date': '20190313',
1132                 'release_year': 2019,
1133             },
1134             'params': {
1135                 'skip_download': True,
1136             },
1137         },
1138         {
1139             # Youtube Music Auto-generated description
1140             # Retrieve 'artist' field from 'Artist:' in video description
1141             # when it is present on youtube music video
1142             'url': 'https://www.youtube.com/watch?v=k0jLE7tTwjY',
1143             'info_dict': {
1144                 'id': 'k0jLE7tTwjY',
1145                 'ext': 'mp4',
1146                 'title': 'Latch Feat. Sam Smith',
1147                 'description': 'md5:3cb1e8101a7c85fcba9b4fb41b951335',
1148                 'upload_date': '20150110',
1149                 'uploader': 'Various Artists - Topic',
1150                 'uploader_id': 'UCNkEcmYdjrH4RqtNgh7BZ9w',
1151                 'artist': 'Disclosure',
1152                 'track': 'Latch Feat. Sam Smith',
1153                 'album': 'Latch Featuring Sam Smith',
1154                 'release_date': '20121008',
1155                 'release_year': 2012,
1156             },
1157             'params': {
1158                 'skip_download': True,
1159             },
1160         },
1161         {
1162             # Youtube Music Auto-generated description
1163             # handle multiple artists on youtube music video
1164             'url': 'https://www.youtube.com/watch?v=74qn0eJSjpA',
1165             'info_dict': {
1166                 'id': '74qn0eJSjpA',
1167                 'ext': 'mp4',
1168                 'title': 'Eastside',
1169                 'description': 'md5:290516bb73dcbfab0dcc4efe6c3de5f2',
1170                 'upload_date': '20180710',
1171                 'uploader': 'Benny Blanco - Topic',
1172                 'uploader_id': 'UCzqz_ksRu_WkIzmivMdIS7A',
1173                 'artist': 'benny blanco, Halsey, Khalid',
1174                 'track': 'Eastside',
1175                 'album': 'Eastside',
1176                 'release_date': '20180713',
1177                 'release_year': 2018,
1178             },
1179             'params': {
1180                 'skip_download': True,
1181             },
1182         },
1183         {
1184             # Youtube Music Auto-generated description
1185             # handle youtube music video with release_year and no release_date
1186             'url': 'https://www.youtube.com/watch?v=-hcAI0g-f5M',
1187             'info_dict': {
1188                 'id': '-hcAI0g-f5M',
1189                 'ext': 'mp4',
1190                 'title': 'Put It On Me',
1191                 'description': 'md5:93c55acc682ae7b0c668f2e34e1c069e',
1192                 'upload_date': '20180426',
1193                 'uploader': 'Matt Maeson - Topic',
1194                 'uploader_id': 'UCnEkIGqtGcQMLk73Kp-Q5LQ',
1195                 'artist': 'Matt Maeson',
1196                 'track': 'Put It On Me',
1197                 'album': 'The Hearse',
1198                 'release_date': None,
1199                 'release_year': 2018,
1200             },
1201             'params': {
1202                 'skip_download': True,
1203             },
1204         },
1205     ]
1206
1207     def __init__(self, *args, **kwargs):
1208         super(YoutubeIE, self).__init__(*args, **kwargs)
1209         self._player_cache = {}
1210
1211     def report_video_info_webpage_download(self, video_id):
1212         """Report attempt to download video info webpage."""
1213         self.to_screen('%s: Downloading video info webpage' % video_id)
1214
1215     def report_information_extraction(self, video_id):
1216         """Report attempt to extract video information."""
1217         self.to_screen('%s: Extracting video information' % video_id)
1218
1219     def report_unavailable_format(self, video_id, format):
1220         """Report extracted video URL."""
1221         self.to_screen('%s: Format %s not available' % (video_id, format))
1222
1223     def report_rtmp_download(self):
1224         """Indicate the download will use the RTMP protocol."""
1225         self.to_screen('RTMP download detected')
1226
1227     def _signature_cache_id(self, example_sig):
1228         """ Return a string representation of a signature """
1229         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
1230
1231     def _extract_signature_function(self, video_id, player_url, example_sig):
1232         id_m = re.match(
1233             r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player(?:-new)?|(?:/[a-z]{2,3}_[A-Z]{2})?/base)?\.(?P<ext>[a-z]+)$',
1234             player_url)
1235         if not id_m:
1236             raise ExtractorError('Cannot identify player %r' % player_url)
1237         player_type = id_m.group('ext')
1238         player_id = id_m.group('id')
1239
1240         # Read from filesystem cache
1241         func_id = '%s_%s_%s' % (
1242             player_type, player_id, self._signature_cache_id(example_sig))
1243         assert os.path.basename(func_id) == func_id
1244
1245         cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
1246         if cache_spec is not None:
1247             return lambda s: ''.join(s[i] for i in cache_spec)
1248
1249         download_note = (
1250             'Downloading player %s' % player_url
1251             if self._downloader.params.get('verbose') else
1252             'Downloading %s player %s' % (player_type, player_id)
1253         )
1254         if player_type == 'js':
1255             code = self._download_webpage(
1256                 player_url, video_id,
1257                 note=download_note,
1258                 errnote='Download of %s failed' % player_url)
1259             res = self._parse_sig_js(code)
1260         elif player_type == 'swf':
1261             urlh = self._request_webpage(
1262                 player_url, video_id,
1263                 note=download_note,
1264                 errnote='Download of %s failed' % player_url)
1265             code = urlh.read()
1266             res = self._parse_sig_swf(code)
1267         else:
1268             assert False, 'Invalid player type %r' % player_type
1269
1270         test_string = ''.join(map(compat_chr, range(len(example_sig))))
1271         cache_res = res(test_string)
1272         cache_spec = [ord(c) for c in cache_res]
1273
1274         self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
1275         return res
1276
1277     def _print_sig_code(self, func, example_sig):
1278         def gen_sig_code(idxs):
1279             def _genslice(start, end, step):
1280                 starts = '' if start == 0 else str(start)
1281                 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
1282                 steps = '' if step == 1 else (':%d' % step)
1283                 return 's[%s%s%s]' % (starts, ends, steps)
1284
1285             step = None
1286             # Quelch pyflakes warnings - start will be set when step is set
1287             start = '(Never used)'
1288             for i, prev in zip(idxs[1:], idxs[:-1]):
1289                 if step is not None:
1290                     if i - prev == step:
1291                         continue
1292                     yield _genslice(start, prev, step)
1293                     step = None
1294                     continue
1295                 if i - prev in [-1, 1]:
1296                     step = i - prev
1297                     start = prev
1298                     continue
1299                 else:
1300                     yield 's[%d]' % prev
1301             if step is None:
1302                 yield 's[%d]' % i
1303             else:
1304                 yield _genslice(start, i, step)
1305
1306         test_string = ''.join(map(compat_chr, range(len(example_sig))))
1307         cache_res = func(test_string)
1308         cache_spec = [ord(c) for c in cache_res]
1309         expr_code = ' + '.join(gen_sig_code(cache_spec))
1310         signature_id_tuple = '(%s)' % (
1311             ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
1312         code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
1313                 '    return %s\n') % (signature_id_tuple, expr_code)
1314         self.to_screen('Extracted signature function:\n' + code)
1315
1316     def _parse_sig_js(self, jscode):
1317         funcname = self._search_regex(
1318             (r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1319              r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1320              r'(?P<sig>[a-zA-Z0-9$]+)\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)',
1321              # Obsolete patterns
1322              r'(["\'])signature\1\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1323              r'\.sig\|\|(?P<sig>[a-zA-Z0-9$]+)\(',
1324              r'yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?:encodeURIComponent\s*\()?\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1325              r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1326              r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1327              r'\bc\s*&&\s*a\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1328              r'\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1329              r'\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\('),
1330             jscode, 'Initial JS player signature function name', group='sig')
1331
1332         jsi = JSInterpreter(jscode)
1333         initial_function = jsi.extract_function(funcname)
1334         return lambda s: initial_function([s])
1335
1336     def _parse_sig_swf(self, file_contents):
1337         swfi = SWFInterpreter(file_contents)
1338         TARGET_CLASSNAME = 'SignatureDecipher'
1339         searched_class = swfi.extract_class(TARGET_CLASSNAME)
1340         initial_function = swfi.extract_function(searched_class, 'decipher')
1341         return lambda s: initial_function([s])
1342
1343     def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
1344         """Turn the encrypted s field into a working signature"""
1345
1346         if player_url is None:
1347             raise ExtractorError('Cannot decrypt signature without player_url')
1348
1349         if player_url.startswith('//'):
1350             player_url = 'https:' + player_url
1351         elif not re.match(r'https?://', player_url):
1352             player_url = compat_urlparse.urljoin(
1353                 'https://www.youtube.com', player_url)
1354         try:
1355             player_id = (player_url, self._signature_cache_id(s))
1356             if player_id not in self._player_cache:
1357                 func = self._extract_signature_function(
1358                     video_id, player_url, s
1359                 )
1360                 self._player_cache[player_id] = func
1361             func = self._player_cache[player_id]
1362             if self._downloader.params.get('youtube_print_sig_code'):
1363                 self._print_sig_code(func, s)
1364             return func(s)
1365         except Exception as e:
1366             tb = traceback.format_exc()
1367             raise ExtractorError(
1368                 'Signature extraction failed: ' + tb, cause=e)
1369
1370     def _get_subtitles(self, video_id, webpage):
1371         try:
1372             subs_doc = self._download_xml(
1373                 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
1374                 video_id, note=False)
1375         except ExtractorError as err:
1376             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
1377             return {}
1378
1379         sub_lang_list = {}
1380         for track in subs_doc.findall('track'):
1381             lang = track.attrib['lang_code']
1382             if lang in sub_lang_list:
1383                 continue
1384             sub_formats = []
1385             for ext in self._SUBTITLE_FORMATS:
1386                 params = compat_urllib_parse_urlencode({
1387                     'lang': lang,
1388                     'v': video_id,
1389                     'fmt': ext,
1390                     'name': track.attrib['name'].encode('utf-8'),
1391                 })
1392                 sub_formats.append({
1393                     'url': 'https://www.youtube.com/api/timedtext?' + params,
1394                     'ext': ext,
1395                 })
1396             sub_lang_list[lang] = sub_formats
1397         if not sub_lang_list:
1398             self._downloader.report_warning('video doesn\'t have subtitles')
1399             return {}
1400         return sub_lang_list
1401
1402     def _get_ytplayer_config(self, video_id, webpage):
1403         patterns = (
1404             # User data may contain arbitrary character sequences that may affect
1405             # JSON extraction with regex, e.g. when '};' is contained the second
1406             # regex won't capture the whole JSON. Yet working around by trying more
1407             # concrete regex first keeping in mind proper quoted string handling
1408             # to be implemented in future that will replace this workaround (see
1409             # https://github.com/ytdl-org/youtube-dl/issues/7468,
1410             # https://github.com/ytdl-org/youtube-dl/pull/7599)
1411             r';ytplayer\.config\s*=\s*({.+?});ytplayer',
1412             r';ytplayer\.config\s*=\s*({.+?});',
1413         )
1414         config = self._search_regex(
1415             patterns, webpage, 'ytplayer.config', default=None)
1416         if config:
1417             return self._parse_json(
1418                 uppercase_escape(config), video_id, fatal=False)
1419
1420     def _get_automatic_captions(self, video_id, webpage):
1421         """We need the webpage for getting the captions url, pass it as an
1422            argument to speed up the process."""
1423         self.to_screen('%s: Looking for automatic captions' % video_id)
1424         player_config = self._get_ytplayer_config(video_id, webpage)
1425         err_msg = 'Couldn\'t find automatic captions for %s' % video_id
1426         if not player_config:
1427             self._downloader.report_warning(err_msg)
1428             return {}
1429         try:
1430             args = player_config['args']
1431             caption_url = args.get('ttsurl')
1432             if caption_url:
1433                 timestamp = args['timestamp']
1434                 # We get the available subtitles
1435                 list_params = compat_urllib_parse_urlencode({
1436                     'type': 'list',
1437                     'tlangs': 1,
1438                     'asrs': 1,
1439                 })
1440                 list_url = caption_url + '&' + list_params
1441                 caption_list = self._download_xml(list_url, video_id)
1442                 original_lang_node = caption_list.find('track')
1443                 if original_lang_node is None:
1444                     self._downloader.report_warning('Video doesn\'t have automatic captions')
1445                     return {}
1446                 original_lang = original_lang_node.attrib['lang_code']
1447                 caption_kind = original_lang_node.attrib.get('kind', '')
1448
1449                 sub_lang_list = {}
1450                 for lang_node in caption_list.findall('target'):
1451                     sub_lang = lang_node.attrib['lang_code']
1452                     sub_formats = []
1453                     for ext in self._SUBTITLE_FORMATS:
1454                         params = compat_urllib_parse_urlencode({
1455                             'lang': original_lang,
1456                             'tlang': sub_lang,
1457                             'fmt': ext,
1458                             'ts': timestamp,
1459                             'kind': caption_kind,
1460                         })
1461                         sub_formats.append({
1462                             'url': caption_url + '&' + params,
1463                             'ext': ext,
1464                         })
1465                     sub_lang_list[sub_lang] = sub_formats
1466                 return sub_lang_list
1467
1468             def make_captions(sub_url, sub_langs):
1469                 parsed_sub_url = compat_urllib_parse_urlparse(sub_url)
1470                 caption_qs = compat_parse_qs(parsed_sub_url.query)
1471                 captions = {}
1472                 for sub_lang in sub_langs:
1473                     sub_formats = []
1474                     for ext in self._SUBTITLE_FORMATS:
1475                         caption_qs.update({
1476                             'tlang': [sub_lang],
1477                             'fmt': [ext],
1478                         })
1479                         sub_url = compat_urlparse.urlunparse(parsed_sub_url._replace(
1480                             query=compat_urllib_parse_urlencode(caption_qs, True)))
1481                         sub_formats.append({
1482                             'url': sub_url,
1483                             'ext': ext,
1484                         })
1485                     captions[sub_lang] = sub_formats
1486                 return captions
1487
1488             # New captions format as of 22.06.2017
1489             player_response = args.get('player_response')
1490             if player_response and isinstance(player_response, compat_str):
1491                 player_response = self._parse_json(
1492                     player_response, video_id, fatal=False)
1493                 if player_response:
1494                     renderer = player_response['captions']['playerCaptionsTracklistRenderer']
1495                     base_url = renderer['captionTracks'][0]['baseUrl']
1496                     sub_lang_list = []
1497                     for lang in renderer['translationLanguages']:
1498                         lang_code = lang.get('languageCode')
1499                         if lang_code:
1500                             sub_lang_list.append(lang_code)
1501                     return make_captions(base_url, sub_lang_list)
1502
1503             # Some videos don't provide ttsurl but rather caption_tracks and
1504             # caption_translation_languages (e.g. 20LmZk1hakA)
1505             # Does not used anymore as of 22.06.2017
1506             caption_tracks = args['caption_tracks']
1507             caption_translation_languages = args['caption_translation_languages']
1508             caption_url = compat_parse_qs(caption_tracks.split(',')[0])['u'][0]
1509             sub_lang_list = []
1510             for lang in caption_translation_languages.split(','):
1511                 lang_qs = compat_parse_qs(compat_urllib_parse_unquote_plus(lang))
1512                 sub_lang = lang_qs.get('lc', [None])[0]
1513                 if sub_lang:
1514                     sub_lang_list.append(sub_lang)
1515             return make_captions(caption_url, sub_lang_list)
1516         # An extractor error can be raise by the download process if there are
1517         # no automatic captions but there are subtitles
1518         except (KeyError, IndexError, ExtractorError):
1519             self._downloader.report_warning(err_msg)
1520             return {}
1521
1522     def _mark_watched(self, video_id, video_info, player_response):
1523         playback_url = url_or_none(try_get(
1524             player_response,
1525             lambda x: x['playbackTracking']['videostatsPlaybackUrl']['baseUrl']) or try_get(
1526             video_info, lambda x: x['videostats_playback_base_url'][0]))
1527         if not playback_url:
1528             return
1529         parsed_playback_url = compat_urlparse.urlparse(playback_url)
1530         qs = compat_urlparse.parse_qs(parsed_playback_url.query)
1531
1532         # cpn generation algorithm is reverse engineered from base.js.
1533         # In fact it works even with dummy cpn.
1534         CPN_ALPHABET = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_'
1535         cpn = ''.join((CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16)))
1536
1537         qs.update({
1538             'ver': ['2'],
1539             'cpn': [cpn],
1540         })
1541         playback_url = compat_urlparse.urlunparse(
1542             parsed_playback_url._replace(query=compat_urllib_parse_urlencode(qs, True)))
1543
1544         self._download_webpage(
1545             playback_url, video_id, 'Marking watched',
1546             'Unable to mark watched', fatal=False)
1547
1548     @staticmethod
1549     def _extract_urls(webpage):
1550         # Embedded YouTube player
1551         entries = [
1552             unescapeHTML(mobj.group('url'))
1553             for mobj in re.finditer(r'''(?x)
1554             (?:
1555                 <iframe[^>]+?src=|
1556                 data-video-url=|
1557                 <embed[^>]+?src=|
1558                 embedSWF\(?:\s*|
1559                 <object[^>]+data=|
1560                 new\s+SWFObject\(
1561             )
1562             (["\'])
1563                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
1564                 (?:embed|v|p)/[0-9A-Za-z_-]{11}.*?)
1565             \1''', webpage)]
1566
1567         # lazyYT YouTube embed
1568         entries.extend(list(map(
1569             unescapeHTML,
1570             re.findall(r'class="lazyYT" data-youtube-id="([^"]+)"', webpage))))
1571
1572         # Wordpress "YouTube Video Importer" plugin
1573         matches = re.findall(r'''(?x)<div[^>]+
1574             class=(?P<q1>[\'"])[^\'"]*\byvii_single_video_player\b[^\'"]*(?P=q1)[^>]+
1575             data-video_id=(?P<q2>[\'"])([^\'"]+)(?P=q2)''', webpage)
1576         entries.extend(m[-1] for m in matches)
1577
1578         return entries
1579
1580     @staticmethod
1581     def _extract_url(webpage):
1582         urls = YoutubeIE._extract_urls(webpage)
1583         return urls[0] if urls else None
1584
1585     @classmethod
1586     def extract_id(cls, url):
1587         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
1588         if mobj is None:
1589             raise ExtractorError('Invalid URL: %s' % url)
1590         video_id = mobj.group(2)
1591         return video_id
1592
1593     def _extract_annotations(self, video_id):
1594         return self._download_webpage(
1595             'https://www.youtube.com/annotations_invideo', video_id,
1596             note='Downloading annotations',
1597             errnote='Unable to download video annotations', fatal=False,
1598             query={
1599                 'features': 1,
1600                 'legacy': 1,
1601                 'video_id': video_id,
1602             })
1603
1604     @staticmethod
1605     def _extract_chapters(description, duration):
1606         if not description:
1607             return None
1608         chapter_lines = re.findall(
1609             r'(?:^|<br\s*/>)([^<]*<a[^>]+onclick=["\']yt\.www\.watch\.player\.seekTo[^>]+>(\d{1,2}:\d{1,2}(?::\d{1,2})?)</a>[^>]*)(?=$|<br\s*/>)',
1610             description)
1611         if not chapter_lines:
1612             return None
1613         chapters = []
1614         for next_num, (chapter_line, time_point) in enumerate(
1615                 chapter_lines, start=1):
1616             start_time = parse_duration(time_point)
1617             if start_time is None:
1618                 continue
1619             if start_time > duration:
1620                 break
1621             end_time = (duration if next_num == len(chapter_lines)
1622                         else parse_duration(chapter_lines[next_num][1]))
1623             if end_time is None:
1624                 continue
1625             if end_time > duration:
1626                 end_time = duration
1627             if start_time > end_time:
1628                 break
1629             chapter_title = re.sub(
1630                 r'<a[^>]+>[^<]+</a>', '', chapter_line).strip(' \t-')
1631             chapter_title = re.sub(r'\s+', ' ', chapter_title)
1632             chapters.append({
1633                 'start_time': start_time,
1634                 'end_time': end_time,
1635                 'title': chapter_title,
1636             })
1637         return chapters
1638
1639     def _real_extract(self, url):
1640         url, smuggled_data = unsmuggle_url(url, {})
1641
1642         proto = (
1643             'http' if self._downloader.params.get('prefer_insecure', False)
1644             else 'https')
1645
1646         start_time = None
1647         end_time = None
1648         parsed_url = compat_urllib_parse_urlparse(url)
1649         for component in [parsed_url.fragment, parsed_url.query]:
1650             query = compat_parse_qs(component)
1651             if start_time is None and 't' in query:
1652                 start_time = parse_duration(query['t'][0])
1653             if start_time is None and 'start' in query:
1654                 start_time = parse_duration(query['start'][0])
1655             if end_time is None and 'end' in query:
1656                 end_time = parse_duration(query['end'][0])
1657
1658         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1659         mobj = re.search(self._NEXT_URL_RE, url)
1660         if mobj:
1661             url = proto + '://www.youtube.com/' + compat_urllib_parse_unquote(mobj.group(1)).lstrip('/')
1662         video_id = self.extract_id(url)
1663
1664         # Get video webpage
1665         url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
1666         video_webpage = self._download_webpage(url, video_id)
1667
1668         # Attempt to extract SWF player URL
1669         mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1670         if mobj is not None:
1671             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1672         else:
1673             player_url = None
1674
1675         dash_mpds = []
1676
1677         def add_dash_mpd(video_info):
1678             dash_mpd = video_info.get('dashmpd')
1679             if dash_mpd and dash_mpd[0] not in dash_mpds:
1680                 dash_mpds.append(dash_mpd[0])
1681
1682         def add_dash_mpd_pr(pl_response):
1683             dash_mpd = url_or_none(try_get(
1684                 pl_response, lambda x: x['streamingData']['dashManifestUrl'],
1685                 compat_str))
1686             if dash_mpd and dash_mpd not in dash_mpds:
1687                 dash_mpds.append(dash_mpd)
1688
1689         is_live = None
1690         view_count = None
1691
1692         def extract_view_count(v_info):
1693             return int_or_none(try_get(v_info, lambda x: x['view_count'][0]))
1694
1695         def extract_token(v_info):
1696             return dict_get(v_info, ('account_playback_token', 'accountPlaybackToken', 'token'))
1697
1698         player_response = {}
1699
1700         # Get video info
1701         embed_webpage = None
1702         if re.search(r'player-age-gate-content">', video_webpage) is not None:
1703             age_gate = True
1704             # We simulate the access to the video from www.youtube.com/v/{video_id}
1705             # this can be viewed without login into Youtube
1706             url = proto + '://www.youtube.com/embed/%s' % video_id
1707             embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage')
1708             data = compat_urllib_parse_urlencode({
1709                 'video_id': video_id,
1710                 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1711                 'sts': self._search_regex(
1712                     r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
1713             })
1714             video_info_url = proto + '://www.youtube.com/get_video_info?' + data
1715             video_info_webpage = self._download_webpage(
1716                 video_info_url, video_id,
1717                 note='Refetching age-gated info webpage',
1718                 errnote='unable to download video info webpage')
1719             video_info = compat_parse_qs(video_info_webpage)
1720             add_dash_mpd(video_info)
1721         else:
1722             age_gate = False
1723             video_info = None
1724             sts = None
1725             # Try looking directly into the video webpage
1726             ytplayer_config = self._get_ytplayer_config(video_id, video_webpage)
1727             if ytplayer_config:
1728                 args = ytplayer_config['args']
1729                 if args.get('url_encoded_fmt_stream_map') or args.get('hlsvp'):
1730                     # Convert to the same format returned by compat_parse_qs
1731                     video_info = dict((k, [v]) for k, v in args.items())
1732                     add_dash_mpd(video_info)
1733                 # Rental video is not rented but preview is available (e.g.
1734                 # https://www.youtube.com/watch?v=yYr8q0y5Jfg,
1735                 # https://github.com/ytdl-org/youtube-dl/issues/10532)
1736                 if not video_info and args.get('ypc_vid'):
1737                     return self.url_result(
1738                         args['ypc_vid'], YoutubeIE.ie_key(), video_id=args['ypc_vid'])
1739                 if args.get('livestream') == '1' or args.get('live_playback') == 1:
1740                     is_live = True
1741                 sts = ytplayer_config.get('sts')
1742                 if not player_response:
1743                     pl_response = str_or_none(args.get('player_response'))
1744                     if pl_response:
1745                         pl_response = self._parse_json(pl_response, video_id, fatal=False)
1746                         if isinstance(pl_response, dict):
1747                             player_response = pl_response
1748             if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
1749                 add_dash_mpd_pr(player_response)
1750                 # We also try looking in get_video_info since it may contain different dashmpd
1751                 # URL that points to a DASH manifest with possibly different itag set (some itags
1752                 # are missing from DASH manifest pointed by webpage's dashmpd, some - from DASH
1753                 # manifest pointed by get_video_info's dashmpd).
1754                 # The general idea is to take a union of itags of both DASH manifests (for example
1755                 # video with such 'manifest behavior' see https://github.com/ytdl-org/youtube-dl/issues/6093)
1756                 self.report_video_info_webpage_download(video_id)
1757                 for el in ('embedded', 'detailpage', 'vevo', ''):
1758                     query = {
1759                         'video_id': video_id,
1760                         'ps': 'default',
1761                         'eurl': '',
1762                         'gl': 'US',
1763                         'hl': 'en',
1764                     }
1765                     if el:
1766                         query['el'] = el
1767                     if sts:
1768                         query['sts'] = sts
1769                     video_info_webpage = self._download_webpage(
1770                         '%s://www.youtube.com/get_video_info' % proto,
1771                         video_id, note=False,
1772                         errnote='unable to download video info webpage',
1773                         fatal=False, query=query)
1774                     if not video_info_webpage:
1775                         continue
1776                     get_video_info = compat_parse_qs(video_info_webpage)
1777                     if not player_response:
1778                         pl_response = get_video_info.get('player_response', [None])[0]
1779                         if isinstance(pl_response, dict):
1780                             player_response = pl_response
1781                             add_dash_mpd_pr(player_response)
1782                     add_dash_mpd(get_video_info)
1783                     if view_count is None:
1784                         view_count = extract_view_count(get_video_info)
1785                     if not video_info:
1786                         video_info = get_video_info
1787                     get_token = extract_token(get_video_info)
1788                     if get_token:
1789                         # Different get_video_info requests may report different results, e.g.
1790                         # some may report video unavailability, but some may serve it without
1791                         # any complaint (see https://github.com/ytdl-org/youtube-dl/issues/7362,
1792                         # the original webpage as well as el=info and el=embedded get_video_info
1793                         # requests report video unavailability due to geo restriction while
1794                         # el=detailpage succeeds and returns valid data). This is probably
1795                         # due to YouTube measures against IP ranges of hosting providers.
1796                         # Working around by preferring the first succeeded video_info containing
1797                         # the token if no such video_info yet was found.
1798                         token = extract_token(video_info)
1799                         if not token:
1800                             video_info = get_video_info
1801                         break
1802
1803         def extract_unavailable_message():
1804             return self._html_search_regex(
1805                 r'(?s)<h1[^>]+id="unavailable-message"[^>]*>(.+?)</h1>',
1806                 video_webpage, 'unavailable message', default=None)
1807
1808         if not video_info:
1809             unavailable_message = extract_unavailable_message()
1810             if not unavailable_message:
1811                 unavailable_message = 'Unable to extract video data'
1812             raise ExtractorError(
1813                 'YouTube said: %s' % unavailable_message, expected=True, video_id=video_id)
1814
1815         video_details = try_get(
1816             player_response, lambda x: x['videoDetails'], dict) or {}
1817
1818         # title
1819         if 'title' in video_info:
1820             video_title = video_info['title'][0]
1821         elif 'title' in player_response:
1822             video_title = video_details['title']
1823         else:
1824             self._downloader.report_warning('Unable to extract video title')
1825             video_title = '_'
1826
1827         # description
1828         description_original = video_description = get_element_by_id("eow-description", video_webpage)
1829         if video_description:
1830
1831             def replace_url(m):
1832                 redir_url = compat_urlparse.urljoin(url, m.group(1))
1833                 parsed_redir_url = compat_urllib_parse_urlparse(redir_url)
1834                 if re.search(r'^(?:www\.)?(?:youtube(?:-nocookie)?\.com|youtu\.be)$', parsed_redir_url.netloc) and parsed_redir_url.path == '/redirect':
1835                     qs = compat_parse_qs(parsed_redir_url.query)
1836                     q = qs.get('q')
1837                     if q and q[0]:
1838                         return q[0]
1839                 return redir_url
1840
1841             description_original = video_description = re.sub(r'''(?x)
1842                 <a\s+
1843                     (?:[a-zA-Z-]+="[^"]*"\s+)*?
1844                     (?:title|href)="([^"]+)"\s+
1845                     (?:[a-zA-Z-]+="[^"]*"\s+)*?
1846                     class="[^"]*"[^>]*>
1847                 [^<]+\.{3}\s*
1848                 </a>
1849             ''', replace_url, video_description)
1850             video_description = clean_html(video_description)
1851         else:
1852             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1853             if fd_mobj:
1854                 video_description = unescapeHTML(fd_mobj.group(1))
1855             else:
1856                 video_description = ''
1857
1858         if not smuggled_data.get('force_singlefeed', False):
1859             if not self._downloader.params.get('noplaylist'):
1860                 multifeed_metadata_list = try_get(
1861                     player_response,
1862                     lambda x: x['multicamera']['playerLegacyMulticameraRenderer']['metadataList'],
1863                     compat_str) or try_get(
1864                     video_info, lambda x: x['multifeed_metadata_list'][0], compat_str)
1865                 if multifeed_metadata_list:
1866                     entries = []
1867                     feed_ids = []
1868                     for feed in multifeed_metadata_list.split(','):
1869                         # Unquote should take place before split on comma (,) since textual
1870                         # fields may contain comma as well (see
1871                         # https://github.com/ytdl-org/youtube-dl/issues/8536)
1872                         feed_data = compat_parse_qs(compat_urllib_parse_unquote_plus(feed))
1873                         entries.append({
1874                             '_type': 'url_transparent',
1875                             'ie_key': 'Youtube',
1876                             'url': smuggle_url(
1877                                 '%s://www.youtube.com/watch?v=%s' % (proto, feed_data['id'][0]),
1878                                 {'force_singlefeed': True}),
1879                             'title': '%s (%s)' % (video_title, feed_data['title'][0]),
1880                         })
1881                         feed_ids.append(feed_data['id'][0])
1882                     self.to_screen(
1883                         'Downloading multifeed video (%s) - add --no-playlist to just download video %s'
1884                         % (', '.join(feed_ids), video_id))
1885                     return self.playlist_result(entries, video_id, video_title, video_description)
1886             else:
1887                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1888
1889         if view_count is None:
1890             view_count = extract_view_count(video_info)
1891         if view_count is None and video_details:
1892             view_count = int_or_none(video_details.get('viewCount'))
1893
1894         if is_live is None:
1895             is_live = bool_or_none(dict_get(
1896                 video_details, ('isLive', 'isLiveContent'),
1897                 skip_false_values=False))
1898
1899         # Check for "rental" videos
1900         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1901             raise ExtractorError('"rental" videos not supported. See https://github.com/ytdl-org/youtube-dl/issues/359 for more information.', expected=True)
1902
1903         def _extract_filesize(media_url):
1904             return int_or_none(self._search_regex(
1905                 r'\bclen[=/](\d+)', media_url, 'filesize', default=None))
1906
1907         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1908             self.report_rtmp_download()
1909             formats = [{
1910                 'format_id': '_rtmp',
1911                 'protocol': 'rtmp',
1912                 'url': video_info['conn'][0],
1913                 'player_url': player_url,
1914             }]
1915         elif not is_live and (len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1):
1916             encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
1917             if 'rtmpe%3Dyes' in encoded_url_map:
1918                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/ytdl-org/youtube-dl/issues/343 for more information.', expected=True)
1919             formats_spec = {}
1920             fmt_list = video_info.get('fmt_list', [''])[0]
1921             if fmt_list:
1922                 for fmt in fmt_list.split(','):
1923                     spec = fmt.split('/')
1924                     if len(spec) > 1:
1925                         width_height = spec[1].split('x')
1926                         if len(width_height) == 2:
1927                             formats_spec[spec[0]] = {
1928                                 'resolution': spec[1],
1929                                 'width': int_or_none(width_height[0]),
1930                                 'height': int_or_none(width_height[1]),
1931                             }
1932             q = qualities(['small', 'medium', 'hd720'])
1933             streaming_formats = try_get(player_response, lambda x: x['streamingData']['formats'], list)
1934             if streaming_formats:
1935                 for fmt in streaming_formats:
1936                     itag = str_or_none(fmt.get('itag'))
1937                     if not itag:
1938                         continue
1939                     quality = fmt.get('quality')
1940                     quality_label = fmt.get('qualityLabel') or quality
1941                     formats_spec[itag] = {
1942                         'asr': int_or_none(fmt.get('audioSampleRate')),
1943                         'filesize': int_or_none(fmt.get('contentLength')),
1944                         'format_note': quality_label,
1945                         'fps': int_or_none(fmt.get('fps')),
1946                         'height': int_or_none(fmt.get('height')),
1947                         'quality': q(quality),
1948                         # bitrate for itag 43 is always 2147483647
1949                         'tbr': float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000) if itag != '43' else None,
1950                         'width': int_or_none(fmt.get('width')),
1951                     }
1952             formats = []
1953             for url_data_str in encoded_url_map.split(','):
1954                 url_data = compat_parse_qs(url_data_str)
1955                 if 'itag' not in url_data or 'url' not in url_data or url_data.get('drm_families'):
1956                     continue
1957                 stream_type = int_or_none(try_get(url_data, lambda x: x['stream_type'][0]))
1958                 # Unsupported FORMAT_STREAM_TYPE_OTF
1959                 if stream_type == 3:
1960                     continue
1961                 format_id = url_data['itag'][0]
1962                 url = url_data['url'][0]
1963
1964                 if 's' in url_data or self._downloader.params.get('youtube_include_dash_manifest', True):
1965                     ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
1966                     jsplayer_url_json = self._search_regex(
1967                         ASSETS_RE,
1968                         embed_webpage if age_gate else video_webpage,
1969                         'JS player URL (1)', default=None)
1970                     if not jsplayer_url_json and not age_gate:
1971                         # We need the embed website after all
1972                         if embed_webpage is None:
1973                             embed_url = proto + '://www.youtube.com/embed/%s' % video_id
1974                             embed_webpage = self._download_webpage(
1975                                 embed_url, video_id, 'Downloading embed webpage')
1976                         jsplayer_url_json = self._search_regex(
1977                             ASSETS_RE, embed_webpage, 'JS player URL')
1978
1979                     player_url = json.loads(jsplayer_url_json)
1980                     if player_url is None:
1981                         player_url_json = self._search_regex(
1982                             r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
1983                             video_webpage, 'age gate player URL')
1984                         player_url = json.loads(player_url_json)
1985
1986                 if 'sig' in url_data:
1987                     url += '&signature=' + url_data['sig'][0]
1988                 elif 's' in url_data:
1989                     encrypted_sig = url_data['s'][0]
1990
1991                     if self._downloader.params.get('verbose'):
1992                         if player_url is None:
1993                             player_version = 'unknown'
1994                             player_desc = 'unknown'
1995                         else:
1996                             if player_url.endswith('swf'):
1997                                 player_version = self._search_regex(
1998                                     r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
1999                                     'flash player', fatal=False)
2000                                 player_desc = 'flash player %s' % player_version
2001                             else:
2002                                 player_version = self._search_regex(
2003                                     [r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js',
2004                                      r'(?:www|player(?:_ias)?)-([^/]+)(?:/[a-z]{2,3}_[A-Z]{2})?/base\.js'],
2005                                     player_url,
2006                                     'html5 player', fatal=False)
2007                                 player_desc = 'html5 player %s' % player_version
2008
2009                         parts_sizes = self._signature_cache_id(encrypted_sig)
2010                         self.to_screen('{%s} signature length %s, %s' %
2011                                        (format_id, parts_sizes, player_desc))
2012
2013                     signature = self._decrypt_signature(
2014                         encrypted_sig, video_id, player_url, age_gate)
2015                     sp = try_get(url_data, lambda x: x['sp'][0], compat_str) or 'signature'
2016                     url += '&%s=%s' % (sp, signature)
2017                 if 'ratebypass' not in url:
2018                     url += '&ratebypass=yes'
2019
2020                 dct = {
2021                     'format_id': format_id,
2022                     'url': url,
2023                     'player_url': player_url,
2024                 }
2025                 if format_id in self._formats:
2026                     dct.update(self._formats[format_id])
2027                 if format_id in formats_spec:
2028                     dct.update(formats_spec[format_id])
2029
2030                 # Some itags are not included in DASH manifest thus corresponding formats will
2031                 # lack metadata (see https://github.com/ytdl-org/youtube-dl/pull/5993).
2032                 # Trying to extract metadata from url_encoded_fmt_stream_map entry.
2033                 mobj = re.search(r'^(?P<width>\d+)[xX](?P<height>\d+)$', url_data.get('size', [''])[0])
2034                 width, height = (int(mobj.group('width')), int(mobj.group('height'))) if mobj else (None, None)
2035
2036                 filesize = int_or_none(url_data.get(
2037                     'clen', [None])[0]) or _extract_filesize(url)
2038
2039                 quality = url_data.get('quality', [None])[0]
2040
2041                 more_fields = {
2042                     'filesize': filesize,
2043                     'tbr': float_or_none(url_data.get('bitrate', [None])[0], 1000),
2044                     'width': width,
2045                     'height': height,
2046                     'fps': int_or_none(url_data.get('fps', [None])[0]),
2047                     'format_note': url_data.get('quality_label', [None])[0] or quality,
2048                     'quality': q(quality),
2049                 }
2050                 for key, value in more_fields.items():
2051                     if value:
2052                         dct[key] = value
2053                 type_ = url_data.get('type', [None])[0]
2054                 if type_:
2055                     type_split = type_.split(';')
2056                     kind_ext = type_split[0].split('/')
2057                     if len(kind_ext) == 2:
2058                         kind, _ = kind_ext
2059                         dct['ext'] = mimetype2ext(type_split[0])
2060                         if kind in ('audio', 'video'):
2061                             codecs = None
2062                             for mobj in re.finditer(
2063                                     r'(?P<key>[a-zA-Z_-]+)=(?P<quote>["\']?)(?P<val>.+?)(?P=quote)(?:;|$)', type_):
2064                                 if mobj.group('key') == 'codecs':
2065                                     codecs = mobj.group('val')
2066                                     break
2067                             if codecs:
2068                                 dct.update(parse_codecs(codecs))
2069                 if dct.get('acodec') == 'none' or dct.get('vcodec') == 'none':
2070                     dct['downloader_options'] = {
2071                         # Youtube throttles chunks >~10M
2072                         'http_chunk_size': 10485760,
2073                     }
2074                 formats.append(dct)
2075         else:
2076             manifest_url = (
2077                 url_or_none(try_get(
2078                     player_response,
2079                     lambda x: x['streamingData']['hlsManifestUrl'],
2080                     compat_str))
2081                 or url_or_none(try_get(
2082                     video_info, lambda x: x['hlsvp'][0], compat_str)))
2083             if manifest_url:
2084                 formats = []
2085                 m3u8_formats = self._extract_m3u8_formats(
2086                     manifest_url, video_id, 'mp4', fatal=False)
2087                 for a_format in m3u8_formats:
2088                     itag = self._search_regex(
2089                         r'/itag/(\d+)/', a_format['url'], 'itag', default=None)
2090                     if itag:
2091                         a_format['format_id'] = itag
2092                         if itag in self._formats:
2093                             dct = self._formats[itag].copy()
2094                             dct.update(a_format)
2095                             a_format = dct
2096                     a_format['player_url'] = player_url
2097                     # Accept-Encoding header causes failures in live streams on Youtube and Youtube Gaming
2098                     a_format.setdefault('http_headers', {})['Youtubedl-no-compression'] = 'True'
2099                     formats.append(a_format)
2100             else:
2101                 error_message = clean_html(video_info.get('reason', [None])[0])
2102                 if not error_message:
2103                     error_message = extract_unavailable_message()
2104                 if error_message:
2105                     raise ExtractorError(error_message, expected=True)
2106                 raise ExtractorError('no conn, hlsvp, hlsManifestUrl or url_encoded_fmt_stream_map information found in video info')
2107
2108         # uploader
2109         video_uploader = try_get(
2110             video_info, lambda x: x['author'][0],
2111             compat_str) or str_or_none(video_details.get('author'))
2112         if video_uploader:
2113             video_uploader = compat_urllib_parse_unquote_plus(video_uploader)
2114         else:
2115             self._downloader.report_warning('unable to extract uploader name')
2116
2117         # uploader_id
2118         video_uploader_id = None
2119         video_uploader_url = None
2120         mobj = re.search(
2121             r'<link itemprop="url" href="(?P<uploader_url>https?://www\.youtube\.com/(?:user|channel)/(?P<uploader_id>[^"]+))">',
2122             video_webpage)
2123         if mobj is not None:
2124             video_uploader_id = mobj.group('uploader_id')
2125             video_uploader_url = mobj.group('uploader_url')
2126         else:
2127             self._downloader.report_warning('unable to extract uploader nickname')
2128
2129         channel_id = (
2130             str_or_none(video_details.get('channelId'))
2131             or self._html_search_meta(
2132                 'channelId', video_webpage, 'channel id', default=None)
2133             or self._search_regex(
2134                 r'data-channel-external-id=(["\'])(?P<id>(?:(?!\1).)+)\1',
2135                 video_webpage, 'channel id', default=None, group='id'))
2136         channel_url = 'http://www.youtube.com/channel/%s' % channel_id if channel_id else None
2137
2138         # thumbnail image
2139         # We try first to get a high quality image:
2140         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
2141                             video_webpage, re.DOTALL)
2142         if m_thumb is not None:
2143             video_thumbnail = m_thumb.group(1)
2144         elif 'thumbnail_url' not in video_info:
2145             self._downloader.report_warning('unable to extract video thumbnail')
2146             video_thumbnail = None
2147         else:   # don't panic if we can't find it
2148             video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0])
2149
2150         # upload date
2151         upload_date = self._html_search_meta(
2152             'datePublished', video_webpage, 'upload date', default=None)
2153         if not upload_date:
2154             upload_date = self._search_regex(
2155                 [r'(?s)id="eow-date.*?>(.*?)</span>',
2156                  r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'],
2157                 video_webpage, 'upload date', default=None)
2158         upload_date = unified_strdate(upload_date)
2159
2160         video_license = self._html_search_regex(
2161             r'<h4[^>]+class="title"[^>]*>\s*License\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li',
2162             video_webpage, 'license', default=None)
2163
2164         m_music = re.search(
2165             r'''(?x)
2166                 <h4[^>]+class="title"[^>]*>\s*Music\s*</h4>\s*
2167                 <ul[^>]*>\s*
2168                 <li>(?P<title>.+?)
2169                 by (?P<creator>.+?)
2170                 (?:
2171                     \(.+?\)|
2172                     <a[^>]*
2173                         (?:
2174                             \bhref=["\']/red[^>]*>|             # drop possible
2175                             >\s*Listen ad-free with YouTube Red # YouTube Red ad
2176                         )
2177                     .*?
2178                 )?</li
2179             ''',
2180             video_webpage)
2181         if m_music:
2182             video_alt_title = remove_quotes(unescapeHTML(m_music.group('title')))
2183             video_creator = clean_html(m_music.group('creator'))
2184         else:
2185             video_alt_title = video_creator = None
2186
2187         def extract_meta(field):
2188             return self._html_search_regex(
2189                 r'<h4[^>]+class="title"[^>]*>\s*%s\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li>\s*' % field,
2190                 video_webpage, field, default=None)
2191
2192         track = extract_meta('Song')
2193         artist = extract_meta('Artist')
2194         album = extract_meta('Album')
2195
2196         # Youtube Music Auto-generated description
2197         release_date = release_year = None
2198         if video_description:
2199             mobj = re.search(r'(?s)Provided to YouTube by [^\n]+\n+(?P<track>[^·]+)·(?P<artist>[^\n]+)\n+(?P<album>[^\n]+)(?:.+?℗\s*(?P<release_year>\d{4})(?!\d))?(?:.+?Released on\s*:\s*(?P<release_date>\d{4}-\d{2}-\d{2}))?(.+?\nArtist\s*:\s*(?P<clean_artist>[^\n]+))?', video_description)
2200             if mobj:
2201                 if not track:
2202                     track = mobj.group('track').strip()
2203                 if not artist:
2204                     artist = mobj.group('clean_artist') or ', '.join(a.strip() for a in mobj.group('artist').split('·'))
2205                 if not album:
2206                     album = mobj.group('album'.strip())
2207                 release_year = mobj.group('release_year')
2208                 release_date = mobj.group('release_date')
2209                 if release_date:
2210                     release_date = release_date.replace('-', '')
2211                     if not release_year:
2212                         release_year = int(release_date[:4])
2213                 if release_year:
2214                     release_year = int(release_year)
2215
2216         m_episode = re.search(
2217             r'<div[^>]+id="watch7-headline"[^>]*>\s*<span[^>]*>.*?>(?P<series>[^<]+)</a></b>\s*S(?P<season>\d+)\s*•\s*E(?P<episode>\d+)</span>',
2218             video_webpage)
2219         if m_episode:
2220             series = unescapeHTML(m_episode.group('series'))
2221             season_number = int(m_episode.group('season'))
2222             episode_number = int(m_episode.group('episode'))
2223         else:
2224             series = season_number = episode_number = None
2225
2226         m_cat_container = self._search_regex(
2227             r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
2228             video_webpage, 'categories', default=None)
2229         if m_cat_container:
2230             category = self._html_search_regex(
2231                 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
2232                 default=None)
2233             video_categories = None if category is None else [category]
2234         else:
2235             video_categories = None
2236
2237         video_tags = [
2238             unescapeHTML(m.group('content'))
2239             for m in re.finditer(self._meta_regex('og:video:tag'), video_webpage)]
2240
2241         def _extract_count(count_name):
2242             return str_to_int(self._search_regex(
2243                 r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>'
2244                 % re.escape(count_name),
2245                 video_webpage, count_name, default=None))
2246
2247         like_count = _extract_count('like')
2248         dislike_count = _extract_count('dislike')
2249
2250         if view_count is None:
2251             view_count = str_to_int(self._search_regex(
2252                 r'<[^>]+class=["\']watch-view-count[^>]+>\s*([\d,\s]+)', video_webpage,
2253                 'view count', default=None))
2254
2255         average_rating = (
2256             float_or_none(video_details.get('averageRating'))
2257             or try_get(video_info, lambda x: float_or_none(x['avg_rating'][0])))
2258
2259         # subtitles
2260         video_subtitles = self.extract_subtitles(video_id, video_webpage)
2261         automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
2262
2263         video_duration = try_get(
2264             video_info, lambda x: int_or_none(x['length_seconds'][0]))
2265         if not video_duration:
2266             video_duration = int_or_none(video_details.get('lengthSeconds'))
2267         if not video_duration:
2268             video_duration = parse_duration(self._html_search_meta(
2269                 'duration', video_webpage, 'video duration'))
2270
2271         # annotations
2272         video_annotations = None
2273         if self._downloader.params.get('writeannotations', False):
2274             video_annotations = self._extract_annotations(video_id)
2275
2276         chapters = self._extract_chapters(description_original, video_duration)
2277
2278         # Look for the DASH manifest
2279         if self._downloader.params.get('youtube_include_dash_manifest', True):
2280             dash_mpd_fatal = True
2281             for mpd_url in dash_mpds:
2282                 dash_formats = {}
2283                 try:
2284                     def decrypt_sig(mobj):
2285                         s = mobj.group(1)
2286                         dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
2287                         return '/signature/%s' % dec_s
2288
2289                     mpd_url = re.sub(r'/s/([a-fA-F0-9\.]+)', decrypt_sig, mpd_url)
2290
2291                     for df in self._extract_mpd_formats(
2292                             mpd_url, video_id, fatal=dash_mpd_fatal,
2293                             formats_dict=self._formats):
2294                         if not df.get('filesize'):
2295                             df['filesize'] = _extract_filesize(df['url'])
2296                         # Do not overwrite DASH format found in some previous DASH manifest
2297                         if df['format_id'] not in dash_formats:
2298                             dash_formats[df['format_id']] = df
2299                         # Additional DASH manifests may end up in HTTP Error 403 therefore
2300                         # allow them to fail without bug report message if we already have
2301                         # some DASH manifest succeeded. This is temporary workaround to reduce
2302                         # burst of bug reports until we figure out the reason and whether it
2303                         # can be fixed at all.
2304                         dash_mpd_fatal = False
2305                 except (ExtractorError, KeyError) as e:
2306                     self.report_warning(
2307                         'Skipping DASH manifest: %r' % e, video_id)
2308                 if dash_formats:
2309                     # Remove the formats we found through non-DASH, they
2310                     # contain less info and it can be wrong, because we use
2311                     # fixed values (for example the resolution). See
2312                     # https://github.com/ytdl-org/youtube-dl/issues/5774 for an
2313                     # example.
2314                     formats = [f for f in formats if f['format_id'] not in dash_formats.keys()]
2315                     formats.extend(dash_formats.values())
2316
2317         # Check for malformed aspect ratio
2318         stretched_m = re.search(
2319             r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">',
2320             video_webpage)
2321         if stretched_m:
2322             w = float(stretched_m.group('w'))
2323             h = float(stretched_m.group('h'))
2324             # yt:stretch may hold invalid ratio data (e.g. for Q39EVAstoRM ratio is 17:0).
2325             # We will only process correct ratios.
2326             if w > 0 and h > 0:
2327                 ratio = w / h
2328                 for f in formats:
2329                     if f.get('vcodec') != 'none':
2330                         f['stretched_ratio'] = ratio
2331
2332         if not formats:
2333             token = extract_token(video_info)
2334             if not token:
2335                 if 'reason' in video_info:
2336                     if 'The uploader has not made this video available in your country.' in video_info['reason']:
2337                         regions_allowed = self._html_search_meta(
2338                             'regionsAllowed', video_webpage, default=None)
2339                         countries = regions_allowed.split(',') if regions_allowed else None
2340                         self.raise_geo_restricted(
2341                             msg=video_info['reason'][0], countries=countries)
2342                     reason = video_info['reason'][0]
2343                     if 'Invalid parameters' in reason:
2344                         unavailable_message = extract_unavailable_message()
2345                         if unavailable_message:
2346                             reason = unavailable_message
2347                     raise ExtractorError(
2348                         'YouTube said: %s' % reason,
2349                         expected=True, video_id=video_id)
2350                 else:
2351                     raise ExtractorError(
2352                         '"token" parameter not in video info for unknown reason',
2353                         video_id=video_id)
2354
2355         if not formats and (video_info.get('license_info') or try_get(player_response, lambda x: x['streamingData']['licenseInfos'])):
2356             raise ExtractorError('This video is DRM protected.', expected=True)
2357
2358         self._sort_formats(formats)
2359
2360         self.mark_watched(video_id, video_info, player_response)
2361
2362         return {
2363             'id': video_id,
2364             'uploader': video_uploader,
2365             'uploader_id': video_uploader_id,
2366             'uploader_url': video_uploader_url,
2367             'channel_id': channel_id,
2368             'channel_url': channel_url,
2369             'upload_date': upload_date,
2370             'license': video_license,
2371             'creator': video_creator or artist,
2372             'title': video_title,
2373             'alt_title': video_alt_title or track,
2374             'thumbnail': video_thumbnail,
2375             'description': video_description,
2376             'categories': video_categories,
2377             'tags': video_tags,
2378             'subtitles': video_subtitles,
2379             'automatic_captions': automatic_captions,
2380             'duration': video_duration,
2381             'age_limit': 18 if age_gate else 0,
2382             'annotations': video_annotations,
2383             'chapters': chapters,
2384             'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
2385             'view_count': view_count,
2386             'like_count': like_count,
2387             'dislike_count': dislike_count,
2388             'average_rating': average_rating,
2389             'formats': formats,
2390             'is_live': is_live,
2391             'start_time': start_time,
2392             'end_time': end_time,
2393             'series': series,
2394             'season_number': season_number,
2395             'episode_number': episode_number,
2396             'track': track,
2397             'artist': artist,
2398             'album': album,
2399             'release_date': release_date,
2400             'release_year': release_year,
2401         }
2402
2403
2404 class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
2405     IE_DESC = 'YouTube.com playlists'
2406     _VALID_URL = r"""(?x)(?:
2407                         (?:https?://)?
2408                         (?:\w+\.)?
2409                         (?:
2410                             (?:
2411                                 youtube\.com|
2412                                 invidio\.us
2413                             )
2414                             /
2415                             (?:
2416                                (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/(?:videoseries|[0-9A-Za-z_-]{11}))
2417                                \? (?:.*?[&;])*? (?:p|a|list)=
2418                             |  p/
2419                             )|
2420                             youtu\.be/[0-9A-Za-z_-]{11}\?.*?\blist=
2421                         )
2422                         (
2423                             (?:PL|LL|EC|UU|FL|RD|UL|TL|OLAK5uy_)?[0-9A-Za-z-_]{10,}
2424                             # Top tracks, they can also include dots
2425                             |(?:MC)[\w\.]*
2426                         )
2427                         .*
2428                      |
2429                         (%(playlist_id)s)
2430                      )""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
2431     _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
2432     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)(?:[^>]+>(?P<title>[^<]+))?'
2433     IE_NAME = 'youtube:playlist'
2434     _TESTS = [{
2435         'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
2436         'info_dict': {
2437             'title': 'ytdl test PL',
2438             'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
2439         },
2440         'playlist_count': 3,
2441     }, {
2442         'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
2443         'info_dict': {
2444             'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
2445             'title': 'YDL_Empty_List',
2446         },
2447         'playlist_count': 0,
2448         'skip': 'This playlist is private',
2449     }, {
2450         'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
2451         'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
2452         'info_dict': {
2453             'title': '29C3: Not my department',
2454             'id': 'PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
2455         },
2456         'playlist_count': 95,
2457     }, {
2458         'note': 'issue #673',
2459         'url': 'PLBB231211A4F62143',
2460         'info_dict': {
2461             'title': '[OLD]Team Fortress 2 (Class-based LP)',
2462             'id': 'PLBB231211A4F62143',
2463         },
2464         'playlist_mincount': 26,
2465     }, {
2466         'note': 'Large playlist',
2467         'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
2468         'info_dict': {
2469             'title': 'Uploads from Cauchemar',
2470             'id': 'UUBABnxM4Ar9ten8Mdjj1j0Q',
2471         },
2472         'playlist_mincount': 799,
2473     }, {
2474         'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
2475         'info_dict': {
2476             'title': 'YDL_safe_search',
2477             'id': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
2478         },
2479         'playlist_count': 2,
2480         'skip': 'This playlist is private',
2481     }, {
2482         'note': 'embedded',
2483         'url': 'https://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
2484         'playlist_count': 4,
2485         'info_dict': {
2486             'title': 'JODA15',
2487             'id': 'PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
2488         }
2489     }, {
2490         'url': 'http://www.youtube.com/embed/_xDOZElKyNU?list=PLsyOSbh5bs16vubvKePAQ1x3PhKavfBIl',
2491         'playlist_mincount': 485,
2492         'info_dict': {
2493             'title': '2017 華語最新單曲 (2/24更新)',
2494             'id': 'PLsyOSbh5bs16vubvKePAQ1x3PhKavfBIl',
2495         }
2496     }, {
2497         'note': 'Embedded SWF player',
2498         'url': 'https://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
2499         'playlist_count': 4,
2500         'info_dict': {
2501             'title': 'JODA7',
2502             'id': 'YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ',
2503         }
2504     }, {
2505         'note': 'Buggy playlist: the webpage has a "Load more" button but it doesn\'t have more videos',
2506         'url': 'https://www.youtube.com/playlist?list=UUXw-G3eDE9trcvY2sBMM_aA',
2507         'info_dict': {
2508             'title': 'Uploads from Interstellar Movie',
2509             'id': 'UUXw-G3eDE9trcvY2sBMM_aA',
2510         },
2511         'playlist_mincount': 21,
2512     }, {
2513         # Playlist URL that does not actually serve a playlist
2514         'url': 'https://www.youtube.com/watch?v=FqZTN594JQw&list=PLMYEtVRpaqY00V9W81Cwmzp6N6vZqfUKD4',
2515         'info_dict': {
2516             'id': 'FqZTN594JQw',
2517             'ext': 'webm',
2518             'title': "Smiley's People 01 detective, Adventure Series, Action",
2519             'uploader': 'STREEM',
2520             'uploader_id': 'UCyPhqAZgwYWZfxElWVbVJng',
2521             'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCyPhqAZgwYWZfxElWVbVJng',
2522             'upload_date': '20150526',
2523             'license': 'Standard YouTube License',
2524             'description': 'md5:507cdcb5a49ac0da37a920ece610be80',
2525             'categories': ['People & Blogs'],
2526             'tags': list,
2527             'view_count': int,
2528             'like_count': int,
2529             'dislike_count': int,
2530         },
2531         'params': {
2532             'skip_download': True,
2533         },
2534         'add_ie': [YoutubeIE.ie_key()],
2535     }, {
2536         'url': 'https://youtu.be/yeWKywCrFtk?list=PL2qgrgXsNUG5ig9cat4ohreBjYLAPC0J5',
2537         'info_dict': {
2538             'id': 'yeWKywCrFtk',
2539             'ext': 'mp4',
2540             'title': 'Small Scale Baler and Braiding Rugs',
2541             'uploader': 'Backus-Page House Museum',
2542             'uploader_id': 'backuspagemuseum',
2543             'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/backuspagemuseum',
2544             'upload_date': '20161008',
2545             'license': 'Standard YouTube License',
2546             'description': 'md5:800c0c78d5eb128500bffd4f0b4f2e8a',
2547             'categories': ['Nonprofits & Activism'],
2548             'tags': list,
2549             'like_count': int,
2550             'dislike_count': int,
2551         },
2552         'params': {
2553             'noplaylist': True,
2554             'skip_download': True,
2555         },
2556     }, {
2557         'url': 'https://youtu.be/uWyaPkt-VOI?list=PL9D9FC436B881BA21',
2558         'only_matching': True,
2559     }, {
2560         'url': 'TLGGrESM50VT6acwMjAyMjAxNw',
2561         'only_matching': True,
2562     }, {
2563         # music album playlist
2564         'url': 'OLAK5uy_m4xAFdmMC5rX3Ji3g93pQe3hqLZw_9LhM',
2565         'only_matching': True,
2566     }, {
2567         'url': 'https://invidio.us/playlist?list=PLDIoUOhQQPlXr63I_vwF9GD8sAKh77dWU',
2568         'only_matching': True,
2569     }]
2570
2571     def _real_initialize(self):
2572         self._login()
2573
2574     def _extract_mix(self, playlist_id):
2575         # The mixes are generated from a single video
2576         # the id of the playlist is just 'RD' + video_id
2577         ids = []
2578         last_id = playlist_id[-11:]
2579         for n in itertools.count(1):
2580             url = 'https://youtube.com/watch?v=%s&list=%s' % (last_id, playlist_id)
2581             webpage = self._download_webpage(
2582                 url, playlist_id, 'Downloading page {0} of Youtube mix'.format(n))
2583             new_ids = orderedSet(re.findall(
2584                 r'''(?xs)data-video-username=".*?".*?
2585                            href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
2586                 webpage))
2587             # Fetch new pages until all the videos are repeated, it seems that
2588             # there are always 51 unique videos.
2589             new_ids = [_id for _id in new_ids if _id not in ids]
2590             if not new_ids:
2591                 break
2592             ids.extend(new_ids)
2593             last_id = ids[-1]
2594
2595         url_results = self._ids_to_results(ids)
2596
2597         search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
2598         title_span = (
2599             search_title('playlist-title')
2600             or search_title('title long-title')
2601             or search_title('title'))
2602         title = clean_html(title_span)
2603
2604         return self.playlist_result(url_results, playlist_id, title)
2605
2606     def _extract_playlist(self, playlist_id):
2607         url = self._TEMPLATE_URL % playlist_id
2608         page = self._download_webpage(url, playlist_id)
2609
2610         # the yt-alert-message now has tabindex attribute (see https://github.com/ytdl-org/youtube-dl/issues/11604)
2611         for match in re.findall(r'<div class="yt-alert-message"[^>]*>([^<]+)</div>', page):
2612             match = match.strip()
2613             # Check if the playlist exists or is private
2614             mobj = re.match(r'[^<]*(?:The|This) playlist (?P<reason>does not exist|is private)[^<]*', match)
2615             if mobj:
2616                 reason = mobj.group('reason')
2617                 message = 'This playlist %s' % reason
2618                 if 'private' in reason:
2619                     message += ', use --username or --netrc to access it'
2620                 message += '.'
2621                 raise ExtractorError(message, expected=True)
2622             elif re.match(r'[^<]*Invalid parameters[^<]*', match):
2623                 raise ExtractorError(
2624                     'Invalid parameters. Maybe URL is incorrect.',
2625                     expected=True)
2626             elif re.match(r'[^<]*Choose your language[^<]*', match):
2627                 continue
2628             else:
2629                 self.report_warning('Youtube gives an alert message: ' + match)
2630
2631         playlist_title = self._html_search_regex(
2632             r'(?s)<h1 class="pl-header-title[^"]*"[^>]*>\s*(.*?)\s*</h1>',
2633             page, 'title', default=None)
2634
2635         _UPLOADER_BASE = r'class=["\']pl-header-details[^>]+>\s*<li>\s*<a[^>]+\bhref='
2636         uploader = self._search_regex(
2637             r'%s["\']/(?:user|channel)/[^>]+>([^<]+)' % _UPLOADER_BASE,
2638             page, 'uploader', default=None)
2639         mobj = re.search(
2640             r'%s(["\'])(?P<path>/(?:user|channel)/(?P<uploader_id>.+?))\1' % _UPLOADER_BASE,
2641             page)
2642         if mobj:
2643             uploader_id = mobj.group('uploader_id')
2644             uploader_url = compat_urlparse.urljoin(url, mobj.group('path'))
2645         else:
2646             uploader_id = uploader_url = None
2647
2648         has_videos = True
2649
2650         if not playlist_title:
2651             try:
2652                 # Some playlist URLs don't actually serve a playlist (e.g.
2653                 # https://www.youtube.com/watch?v=FqZTN594JQw&list=PLMYEtVRpaqY00V9W81Cwmzp6N6vZqfUKD4)
2654                 next(self._entries(page, playlist_id))
2655             except StopIteration:
2656                 has_videos = False
2657
2658         playlist = self.playlist_result(
2659             self._entries(page, playlist_id), playlist_id, playlist_title)
2660         playlist.update({
2661             'uploader': uploader,
2662             'uploader_id': uploader_id,
2663             'uploader_url': uploader_url,
2664         })
2665
2666         return has_videos, playlist
2667
2668     def _check_download_just_video(self, url, playlist_id):
2669         # Check if it's a video-specific URL
2670         query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
2671         video_id = query_dict.get('v', [None])[0] or self._search_regex(
2672             r'(?:(?:^|//)youtu\.be/|youtube\.com/embed/(?!videoseries))([0-9A-Za-z_-]{11})', url,
2673             'video id', default=None)
2674         if video_id:
2675             if self._downloader.params.get('noplaylist'):
2676                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
2677                 return video_id, self.url_result(video_id, 'Youtube', video_id=video_id)
2678             else:
2679                 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
2680                 return video_id, None
2681         return None, None
2682
2683     def _real_extract(self, url):
2684         # Extract playlist id
2685         mobj = re.match(self._VALID_URL, url)
2686         if mobj is None:
2687             raise ExtractorError('Invalid URL: %s' % url)
2688         playlist_id = mobj.group(1) or mobj.group(2)
2689
2690         video_id, video = self._check_download_just_video(url, playlist_id)
2691         if video:
2692             return video
2693
2694         if playlist_id.startswith(('RD', 'UL', 'PU')):
2695             # Mixes require a custom extraction process
2696             return self._extract_mix(playlist_id)
2697
2698         has_videos, playlist = self._extract_playlist(playlist_id)
2699         if has_videos or not video_id:
2700             return playlist
2701
2702         # Some playlist URLs don't actually serve a playlist (see
2703         # https://github.com/ytdl-org/youtube-dl/issues/10537).
2704         # Fallback to plain video extraction if there is a video id
2705         # along with playlist id.
2706         return self.url_result(video_id, 'Youtube', video_id=video_id)
2707
2708
2709 class YoutubeChannelIE(YoutubePlaylistBaseInfoExtractor):
2710     IE_DESC = 'YouTube.com channels'
2711     _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com|(?:www\.)?invidio\.us)/channel/(?P<id>[0-9A-Za-z_-]+)'
2712     _TEMPLATE_URL = 'https://www.youtube.com/channel/%s/videos'
2713     _VIDEO_RE = r'(?:title="(?P<title>[^"]+)"[^>]+)?href="/watch\?v=(?P<id>[0-9A-Za-z_-]+)&?'
2714     IE_NAME = 'youtube:channel'
2715     _TESTS = [{
2716         'note': 'paginated channel',
2717         'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
2718         'playlist_mincount': 91,
2719         'info_dict': {
2720             'id': 'UUKfVa3S1e4PHvxWcwyMMg8w',
2721             'title': 'Uploads from lex will',
2722         }
2723     }, {
2724         'note': 'Age restricted channel',
2725         # from https://www.youtube.com/user/DeusExOfficial
2726         'url': 'https://www.youtube.com/channel/UCs0ifCMCm1icqRbqhUINa0w',
2727         'playlist_mincount': 64,
2728         'info_dict': {
2729             'id': 'UUs0ifCMCm1icqRbqhUINa0w',
2730             'title': 'Uploads from Deus Ex',
2731         },
2732     }, {
2733         'url': 'https://invidio.us/channel/UC23qupoDRn9YOAVzeoxjOQA',
2734         'only_matching': True,
2735     }]
2736
2737     @classmethod
2738     def suitable(cls, url):
2739         return (False if YoutubePlaylistsIE.suitable(url) or YoutubeLiveIE.suitable(url)
2740                 else super(YoutubeChannelIE, cls).suitable(url))
2741
2742     def _build_template_url(self, url, channel_id):
2743         return self._TEMPLATE_URL % channel_id
2744
2745     def _real_extract(self, url):
2746         channel_id = self._match_id(url)
2747
2748         url = self._build_template_url(url, channel_id)
2749
2750         # Channel by page listing is restricted to 35 pages of 30 items, i.e. 1050 videos total (see #5778)
2751         # Workaround by extracting as a playlist if managed to obtain channel playlist URL
2752         # otherwise fallback on channel by page extraction
2753         channel_page = self._download_webpage(
2754             url + '?view=57', channel_id,
2755             'Downloading channel page', fatal=False)
2756         if channel_page is False:
2757             channel_playlist_id = False
2758         else:
2759             channel_playlist_id = self._html_search_meta(
2760                 'channelId', channel_page, 'channel id', default=None)
2761             if not channel_playlist_id:
2762                 channel_url = self._html_search_meta(
2763                     ('al:ios:url', 'twitter:app:url:iphone', 'twitter:app:url:ipad'),
2764                     channel_page, 'channel url', default=None)
2765                 if channel_url:
2766                     channel_playlist_id = self._search_regex(
2767                         r'vnd\.youtube://user/([0-9A-Za-z_-]+)',
2768                         channel_url, 'channel id', default=None)
2769         if channel_playlist_id and channel_playlist_id.startswith('UC'):
2770             playlist_id = 'UU' + channel_playlist_id[2:]
2771             return self.url_result(
2772                 compat_urlparse.urljoin(url, '/playlist?list=%s' % playlist_id), 'YoutubePlaylist')
2773
2774         channel_page = self._download_webpage(url, channel_id, 'Downloading page #1')
2775         autogenerated = re.search(r'''(?x)
2776                 class="[^"]*?(?:
2777                     channel-header-autogenerated-label|
2778                     yt-channel-title-autogenerated
2779                 )[^"]*"''', channel_page) is not None
2780
2781         if autogenerated:
2782             # The videos are contained in a single page
2783             # the ajax pages can't be used, they are empty
2784             entries = [
2785                 self.url_result(
2786                     video_id, 'Youtube', video_id=video_id,
2787                     video_title=video_title)
2788                 for video_id, video_title in self.extract_videos_from_page(channel_page)]
2789             return self.playlist_result(entries, channel_id)
2790
2791         try:
2792             next(self._entries(channel_page, channel_id))
2793         except StopIteration:
2794             alert_message = self._html_search_regex(
2795                 r'(?s)<div[^>]+class=(["\']).*?\byt-alert-message\b.*?\1[^>]*>(?P<alert>[^<]+)</div>',
2796                 channel_page, 'alert', default=None, group='alert')
2797             if alert_message:
2798                 raise ExtractorError('Youtube said: %s' % alert_message, expected=True)
2799
2800         return self.playlist_result(self._entries(channel_page, channel_id), channel_id)
2801
2802
2803 class YoutubeUserIE(YoutubeChannelIE):
2804     IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
2805     _VALID_URL = r'(?:(?:https?://(?:\w+\.)?youtube\.com/(?:(?P<user>user|c)/)?(?!(?:attribution_link|watch|results|shared)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)'
2806     _TEMPLATE_URL = 'https://www.youtube.com/%s/%s/videos'
2807     IE_NAME = 'youtube:user'
2808
2809     _TESTS = [{
2810         'url': 'https://www.youtube.com/user/TheLinuxFoundation',
2811         'playlist_mincount': 320,
2812         'info_dict': {
2813             'id': 'UUfX55Sx5hEFjoC3cNs6mCUQ',
2814             'title': 'Uploads from The Linux Foundation',
2815         }
2816     }, {
2817         # Only available via https://www.youtube.com/c/12minuteathlete/videos
2818         # but not https://www.youtube.com/user/12minuteathlete/videos
2819         'url': 'https://www.youtube.com/c/12minuteathlete/videos',
2820         'playlist_mincount': 249,
2821         'info_dict': {
2822             'id': 'UUVjM-zV6_opMDx7WYxnjZiQ',
2823             'title': 'Uploads from 12 Minute Athlete',
2824         }
2825     }, {
2826         'url': 'ytuser:phihag',
2827         'only_matching': True,
2828     }, {
2829         'url': 'https://www.youtube.com/c/gametrailers',
2830         'only_matching': True,
2831     }, {
2832         'url': 'https://www.youtube.com/gametrailers',
2833         'only_matching': True,
2834     }, {
2835         # This channel is not available, geo restricted to JP
2836         'url': 'https://www.youtube.com/user/kananishinoSMEJ/videos',
2837         'only_matching': True,
2838     }]
2839
2840     @classmethod
2841     def suitable(cls, url):
2842         # Don't return True if the url can be extracted with other youtube
2843         # extractor, the regex would is too permissive and it would match.
2844         other_yt_ies = iter(klass for (name, klass) in globals().items() if name.startswith('Youtube') and name.endswith('IE') and klass is not cls)
2845         if any(ie.suitable(url) for ie in other_yt_ies):
2846             return False
2847         else:
2848             return super(YoutubeUserIE, cls).suitable(url)
2849
2850     def _build_template_url(self, url, channel_id):
2851         mobj = re.match(self._VALID_URL, url)
2852         return self._TEMPLATE_URL % (mobj.group('user') or 'user', mobj.group('id'))
2853
2854
2855 class YoutubeLiveIE(YoutubeBaseInfoExtractor):
2856     IE_DESC = 'YouTube.com live streams'
2857     _VALID_URL = r'(?P<base_url>https?://(?:\w+\.)?youtube\.com/(?:(?:user|channel|c)/)?(?P<id>[^/]+))/live'
2858     IE_NAME = 'youtube:live'
2859
2860     _TESTS = [{
2861         'url': 'https://www.youtube.com/user/TheYoungTurks/live',
2862         'info_dict': {
2863             'id': 'a48o2S1cPoo',
2864             'ext': 'mp4',
2865             'title': 'The Young Turks - Live Main Show',
2866             'uploader': 'The Young Turks',
2867             'uploader_id': 'TheYoungTurks',
2868             'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/TheYoungTurks',
2869             'upload_date': '20150715',
2870             'license': 'Standard YouTube License',
2871             'description': 'md5:438179573adcdff3c97ebb1ee632b891',
2872             'categories': ['News & Politics'],
2873             'tags': ['Cenk Uygur (TV Program Creator)', 'The Young Turks (Award-Winning Work)', 'Talk Show (TV Genre)'],
2874             'like_count': int,
2875             'dislike_count': int,
2876         },
2877         'params': {
2878             'skip_download': True,
2879         },
2880     }, {
2881         'url': 'https://www.youtube.com/channel/UC1yBKRuGpC1tSM73A0ZjYjQ/live',
2882         'only_matching': True,
2883     }, {
2884         'url': 'https://www.youtube.com/c/CommanderVideoHq/live',
2885         'only_matching': True,
2886     }, {
2887         'url': 'https://www.youtube.com/TheYoungTurks/live',
2888         'only_matching': True,
2889     }]
2890
2891     def _real_extract(self, url):
2892         mobj = re.match(self._VALID_URL, url)
2893         channel_id = mobj.group('id')
2894         base_url = mobj.group('base_url')
2895         webpage = self._download_webpage(url, channel_id, fatal=False)
2896         if webpage:
2897             page_type = self._og_search_property(
2898                 'type', webpage, 'page type', default='')
2899             video_id = self._html_search_meta(
2900                 'videoId', webpage, 'video id', default=None)
2901             if page_type.startswith('video') and video_id and re.match(
2902                     r'^[0-9A-Za-z_-]{11}$', video_id):
2903                 return self.url_result(video_id, YoutubeIE.ie_key())
2904         return self.url_result(base_url)
2905
2906
2907 class YoutubePlaylistsIE(YoutubePlaylistsBaseInfoExtractor):
2908     IE_DESC = 'YouTube.com user/channel playlists'
2909     _VALID_URL = r'https?://(?:\w+\.)?youtube\.com/(?:user|channel)/(?P<id>[^/]+)/playlists'
2910     IE_NAME = 'youtube:playlists'
2911
2912     _TESTS = [{
2913         'url': 'https://www.youtube.com/user/ThirstForScience/playlists',
2914         'playlist_mincount': 4,
2915         'info_dict': {
2916             'id': 'ThirstForScience',
2917             'title': 'Thirst for Science',
2918         },
2919     }, {
2920         # with "Load more" button
2921         'url': 'https://www.youtube.com/user/igorkle1/playlists?view=1&sort=dd',
2922         'playlist_mincount': 70,
2923         'info_dict': {
2924             'id': 'igorkle1',
2925             'title': 'Игорь Клейнер',
2926         },
2927     }, {
2928         'url': 'https://www.youtube.com/channel/UCiU1dHvZObB2iP6xkJ__Icw/playlists',
2929         'playlist_mincount': 17,
2930         'info_dict': {
2931             'id': 'UCiU1dHvZObB2iP6xkJ__Icw',
2932             'title': 'Chem Player',
2933         },
2934     }]
2935
2936
2937 class YoutubeSearchBaseInfoExtractor(YoutubePlaylistBaseInfoExtractor):
2938     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?'
2939
2940
2941 class YoutubeSearchIE(SearchInfoExtractor, YoutubeSearchBaseInfoExtractor):
2942     IE_DESC = 'YouTube.com searches'
2943     # there doesn't appear to be a real limit, for example if you search for
2944     # 'python' you get more than 8.000.000 results
2945     _MAX_RESULTS = float('inf')
2946     IE_NAME = 'youtube:search'
2947     _SEARCH_KEY = 'ytsearch'
2948     _EXTRA_QUERY_ARGS = {}
2949     _TESTS = []
2950
2951     def _get_n_results(self, query, n):
2952         """Get a specified number of results for a query"""
2953
2954         videos = []
2955         limit = n
2956
2957         url_query = {
2958             'search_query': query.encode('utf-8'),
2959         }
2960         url_query.update(self._EXTRA_QUERY_ARGS)
2961         result_url = 'https://www.youtube.com/results?' + compat_urllib_parse_urlencode(url_query)
2962
2963         for pagenum in itertools.count(1):
2964             data = self._download_json(
2965                 result_url, video_id='query "%s"' % query,
2966                 note='Downloading page %s' % pagenum,
2967                 errnote='Unable to download API page',
2968                 query={'spf': 'navigate'})
2969             html_content = data[1]['body']['content']
2970
2971             if 'class="search-message' in html_content:
2972                 raise ExtractorError(
2973                     '[youtube] No video results', expected=True)
2974
2975             new_videos = list(self._process_page(html_content))
2976             videos += new_videos
2977             if not new_videos or len(videos) > limit:
2978                 break
2979             next_link = self._html_search_regex(
2980                 r'href="(/results\?[^"]*\bsp=[^"]+)"[^>]*>\s*<span[^>]+class="[^"]*\byt-uix-button-content\b[^"]*"[^>]*>Next',
2981                 html_content, 'next link', default=None)
2982             if next_link is None:
2983                 break
2984             result_url = compat_urlparse.urljoin('https://www.youtube.com/', next_link)
2985
2986         if len(videos) > n:
2987             videos = videos[:n]
2988         return self.playlist_result(videos, query)
2989
2990
2991 class YoutubeSearchDateIE(YoutubeSearchIE):
2992     IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
2993     _SEARCH_KEY = 'ytsearchdate'
2994     IE_DESC = 'YouTube.com searches, newest videos first'
2995     _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
2996
2997
2998 class YoutubeSearchURLIE(YoutubeSearchBaseInfoExtractor):
2999     IE_DESC = 'YouTube.com search URLs'
3000     IE_NAME = 'youtube:search_url'
3001     _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?P<query>[^&]+)(?:[&]|$)'
3002     _TESTS = [{
3003         'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
3004         'playlist_mincount': 5,
3005         'info_dict': {
3006             'title': 'youtube-dl test video',
3007         }
3008     }, {
3009         'url': 'https://www.youtube.com/results?q=test&sp=EgQIBBgB',
3010         'only_matching': True,
3011     }]
3012
3013     def _real_extract(self, url):
3014         mobj = re.match(self._VALID_URL, url)
3015         query = compat_urllib_parse_unquote_plus(mobj.group('query'))
3016         webpage = self._download_webpage(url, query)
3017         return self.playlist_result(self._process_page(webpage), playlist_title=query)
3018
3019
3020 class YoutubeShowIE(YoutubePlaylistsBaseInfoExtractor):
3021     IE_DESC = 'YouTube.com (multi-season) shows'
3022     _VALID_URL = r'https?://(?:www\.)?youtube\.com/show/(?P<id>[^?#]*)'
3023     IE_NAME = 'youtube:show'
3024     _TESTS = [{
3025         'url': 'https://www.youtube.com/show/airdisasters',
3026         'playlist_mincount': 5,
3027         'info_dict': {
3028             'id': 'airdisasters',
3029             'title': 'Air Disasters',
3030         }
3031     }]
3032
3033     def _real_extract(self, url):
3034         playlist_id = self._match_id(url)
3035         return super(YoutubeShowIE, self)._real_extract(
3036             'https://www.youtube.com/show/%s/playlists' % playlist_id)
3037
3038
3039 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
3040     """
3041     Base class for feed extractors
3042     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
3043     """
3044     _LOGIN_REQUIRED = True
3045
3046     @property
3047     def IE_NAME(self):
3048         return 'youtube:%s' % self._FEED_NAME
3049
3050     def _real_initialize(self):
3051         self._login()
3052
3053     def _entries(self, page):
3054         # The extraction process is the same as for playlists, but the regex
3055         # for the video ids doesn't contain an index
3056         ids = []
3057         more_widget_html = content_html = page
3058         for page_num in itertools.count(1):
3059             matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
3060
3061             # 'recommended' feed has infinite 'load more' and each new portion spins
3062             # the same videos in (sometimes) slightly different order, so we'll check
3063             # for unicity and break when portion has no new videos
3064             new_ids = list(filter(lambda video_id: video_id not in ids, orderedSet(matches)))
3065             if not new_ids:
3066                 break
3067
3068             ids.extend(new_ids)
3069
3070             for entry in self._ids_to_results(new_ids):
3071                 yield entry
3072
3073             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
3074             if not mobj:
3075                 break
3076
3077             more = self._download_json(
3078                 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE,
3079                 'Downloading page #%s' % page_num,
3080                 transform_source=uppercase_escape)
3081             content_html = more['content_html']
3082             more_widget_html = more['load_more_widget_html']
3083
3084     def _real_extract(self, url):
3085         page = self._download_webpage(
3086             'https://www.youtube.com/feed/%s' % self._FEED_NAME,
3087             self._PLAYLIST_TITLE)
3088         return self.playlist_result(
3089             self._entries(page), playlist_title=self._PLAYLIST_TITLE)
3090
3091
3092 class YoutubeWatchLaterIE(YoutubePlaylistIE):
3093     IE_NAME = 'youtube:watchlater'
3094     IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
3095     _VALID_URL = r'https?://(?:www\.)?youtube\.com/(?:feed/watch_later|(?:playlist|watch)\?(?:.+&)?list=WL)|:ytwatchlater'
3096
3097     _TESTS = [{
3098         'url': 'https://www.youtube.com/playlist?list=WL',
3099         'only_matching': True,
3100     }, {
3101         'url': 'https://www.youtube.com/watch?v=bCNU9TrbiRk&index=1&list=WL',
3102         'only_matching': True,
3103     }]
3104
3105     def _real_extract(self, url):
3106         _, video = self._check_download_just_video(url, 'WL')
3107         if video:
3108             return video
3109         _, playlist = self._extract_playlist('WL')
3110         return playlist
3111
3112
3113 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
3114     IE_NAME = 'youtube:favorites'
3115     IE_DESC = 'YouTube.com favourite videos, ":ytfav" for short (requires authentication)'
3116     _VALID_URL = r'https?://(?:www\.)?youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
3117     _LOGIN_REQUIRED = True
3118
3119     def _real_extract(self, url):
3120         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
3121         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
3122         return self.url_result(playlist_id, 'YoutubePlaylist')
3123
3124
3125 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
3126     IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
3127     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/recommended|:ytrec(?:ommended)?'
3128     _FEED_NAME = 'recommended'
3129     _PLAYLIST_TITLE = 'Youtube Recommended videos'
3130
3131
3132 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
3133     IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
3134     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
3135     _FEED_NAME = 'subscriptions'
3136     _PLAYLIST_TITLE = 'Youtube Subscriptions'
3137
3138
3139 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
3140     IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
3141     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/history|:ythistory'
3142     _FEED_NAME = 'history'
3143     _PLAYLIST_TITLE = 'Youtube History'
3144
3145
3146 class YoutubeTruncatedURLIE(InfoExtractor):
3147     IE_NAME = 'youtube:truncated_url'
3148     IE_DESC = False  # Do not list
3149     _VALID_URL = r'''(?x)
3150         (?:https?://)?
3151         (?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/
3152         (?:watch\?(?:
3153             feature=[a-z_]+|
3154             annotation_id=annotation_[^&]+|
3155             x-yt-cl=[0-9]+|
3156             hl=[^&]*|
3157             t=[0-9]+
3158         )?
3159         |
3160             attribution_link\?a=[^&]+
3161         )
3162         $
3163     '''
3164
3165     _TESTS = [{
3166         'url': 'https://www.youtube.com/watch?annotation_id=annotation_3951667041',
3167         'only_matching': True,
3168     }, {
3169         'url': 'https://www.youtube.com/watch?',
3170         'only_matching': True,
3171     }, {
3172         'url': 'https://www.youtube.com/watch?x-yt-cl=84503534',
3173         'only_matching': True,
3174     }, {
3175         'url': 'https://www.youtube.com/watch?feature=foo',
3176         'only_matching': True,
3177     }, {
3178         'url': 'https://www.youtube.com/watch?hl=en-GB',
3179         'only_matching': True,
3180     }, {
3181         'url': 'https://www.youtube.com/watch?t=2372',
3182         'only_matching': True,
3183     }]
3184
3185     def _real_extract(self, url):
3186         raise ExtractorError(
3187             'Did you forget to quote the URL? Remember that & is a meta '
3188             'character in most shells, so you want to put the URL in quotes, '
3189             'like  youtube-dl '
3190             '"https://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
3191             ' or simply  youtube-dl BaW_jenozKc  .',
3192             expected=True)
3193
3194
3195 class YoutubeTruncatedIDIE(InfoExtractor):
3196     IE_NAME = 'youtube:truncated_id'
3197     IE_DESC = False  # Do not list
3198     _VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
3199
3200     _TESTS = [{
3201         'url': 'https://www.youtube.com/watch?v=N_708QY7Ob',
3202         'only_matching': True,
3203     }]
3204
3205     def _real_extract(self, url):
3206         video_id = self._match_id(url)
3207         raise ExtractorError(
3208             'Incomplete YouTube ID %s. URL %s looks truncated.' % (video_id, url),
3209             expected=True)