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 re
  10 import traceback
  11
  12 from .common import InfoExtractor, SearchInfoExtractor
  13 from .subtitles import SubtitlesInfoExtractor
  14 from ..jsinterp import JSInterpreter
  15 from ..swfinterp import SWFInterpreter
  16 from ..utils import (
  17     compat_chr,
  18     compat_parse_qs,
  19     compat_urllib_parse,
  20     compat_urllib_request,
  21     compat_urlparse,
  22     compat_str,
  23
  24     clean_html,
  25     get_element_by_id,
  26     get_element_by_attribute,
  27     ExtractorError,
  28     int_or_none,
  29     OnDemandPagedList,
  30     unescapeHTML,
  31     unified_strdate,
  32     orderedSet,
  33     uppercase_escape,
  34 )
  35
  36 class YoutubeBaseInfoExtractor(InfoExtractor):
  37     """Provide base functions for Youtube extractors"""
  38     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  39     _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor'
  40     _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
  41     _AGE_URL = 'https://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
  42     _NETRC_MACHINE = 'youtube'
  43     # If True it will raise an error if no login info is provided
  44     _LOGIN_REQUIRED = False
  45
  46     def _set_language(self):
  47         return bool(self._download_webpage(
  48             self._LANG_URL, None,
  49             note='Setting language', errnote='unable to set language',
  50             fatal=False))
  51
  52     def _login(self):
  53         """
  54         Attempt to log in to YouTube.
  55         True is returned if successful or skipped.
  56         False is returned if login failed.
  57
  58         If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
  59         """
  60         (username, password) = self._get_login_info()
  61         # No authentication to be performed
  62         if username is None:
  63             if self._LOGIN_REQUIRED:
  64                 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  65             return True
  66
  67         login_page = self._download_webpage(
  68             self._LOGIN_URL, None,
  69             note='Downloading login page',
  70             errnote='unable to fetch login page', fatal=False)
  71         if login_page is False:
  72             return
  73
  74         galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
  75                                   login_page, 'Login GALX parameter')
  76
  77         # Log in
  78         login_form_strs = {
  79                 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
  80                 'Email': username,
  81                 'GALX': galx,
  82                 'Passwd': password,
  83
  84                 'PersistentCookie': 'yes',
  85                 '_utf8': '霱',
  86                 'bgresponse': 'js_disabled',
  87                 'checkConnection': '',
  88                 'checkedDomains': 'youtube',
  89                 'dnConn': '',
  90                 'pstMsg': '0',
  91                 'rmShown': '1',
  92                 'secTok': '',
  93                 'signIn': 'Sign in',
  94                 'timeStmp': '',
  95                 'service': 'youtube',
  96                 'uilel': '3',
  97                 'hl': 'en_US',
  98         }
  99
 100         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
 101         # chokes on unicode
 102         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
 103         login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
 104
 105         req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
 106         login_results = self._download_webpage(
 107             req, None,
 108             note='Logging in', errnote='unable to log in', fatal=False)
 109         if login_results is False:
 110             return False
 111
 112         if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
 113             raise ExtractorError('Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
 114
 115         # Two-Factor
 116         # TODO add SMS and phone call support - these require making a request and then prompting the user
 117
 118         if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', login_results) is not None:
 119             tfa_code = self._get_tfa_info()
 120
 121             if tfa_code is None:
 122                 self._downloader.report_warning('Two-factor authentication required. Provide it with --twofactor <code>')
 123                 self._downloader.report_warning('(Note that only TOTP (Google Authenticator App) codes work at this time.)')
 124                 return False
 125
 126             # Unlike the first login form, secTok and timeStmp are both required for the TFA form
 127
 128             match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
 129             if match is None:
 130                 self._downloader.report_warning('Failed to get secTok - did the page structure change?')
 131             secTok = match.group(1)
 132             match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
 133             if match is None:
 134                 self._downloader.report_warning('Failed to get timeStmp - did the page structure change?')
 135             timeStmp = match.group(1)
 136
 137             tfa_form_strs = {
 138                 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
 139                 'smsToken': '',
 140                 'smsUserPin': tfa_code,
 141                 'smsVerifyPin': 'Verify',
 142
 143                 'PersistentCookie': 'yes',
 144                 'checkConnection': '',
 145                 'checkedDomains': 'youtube',
 146                 'pstMsg': '1',
 147                 'secTok': secTok,
 148                 'timeStmp': timeStmp,
 149                 'service': 'youtube',
 150                 'hl': 'en_US',
 151             }
 152             tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in tfa_form_strs.items())
 153             tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
 154
 155             tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
 156             tfa_results = self._download_webpage(
 157                 tfa_req, None,
 158                 note='Submitting TFA code', errnote='unable to submit tfa', fatal=False)
 159
 160             if tfa_results is False:
 161                 return False
 162
 163             if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', tfa_results) is not None:
 164                 self._downloader.report_warning('Two-factor code expired. Please try again, or use a one-use backup code instead.')
 165                 return False
 166             if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
 167                 self._downloader.report_warning('unable to log in - did the page structure change?')
 168                 return False
 169             if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
 170                 self._downloader.report_warning('Your Google account has a security notice. Please log in on your web browser, resolve the notice, and try again.')
 171                 return False
 172
 173         if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
 174             self._downloader.report_warning('unable to log in: bad username or password')
 175             return False
 176         return True
 177
 178     def _confirm_age(self):
 179         age_form = {
 180             'next_url': '/',
 181             'action_confirm': 'Confirm',
 182         }
 183         req = compat_urllib_request.Request(self._AGE_URL,
 184             compat_urllib_parse.urlencode(age_form).encode('ascii'))
 185
 186         self._download_webpage(
 187             req, None,
 188             note='Confirming age', errnote='Unable to confirm age',
 189             fatal=False)
 190
 191     def _real_initialize(self):
 192         if self._downloader is None:
 193             return
 194         if self._get_login_info()[0] is not None:
 195             if not self._set_language():
 196                 return
 197         if not self._login():
 198             return
 199         self._confirm_age()
 200
 201
 202 class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
 203     IE_DESC = 'YouTube.com'
 204     _VALID_URL = r"""(?x)^
 205                      (
 206                          (?:https?://|//)                                    # http(s):// or protocol-independent URL
 207                          (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
 208                             (?:www\.)?deturl\.com/www\.youtube\.com/|
 209                             (?:www\.)?pwnyoutube\.com/|
 210                             (?:www\.)?yourepeat\.com/|
 211                             tube\.majestyc\.net/|
 212                             youtube\.googleapis\.com/)                        # the various hostnames, with wildcard subdomains
 213                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
 214                          (?:                                                  # the various things that can precede the ID:
 215                              (?:(?:v|embed|e)/(?!videoseries))                # v/ or embed/ or e/
 216                              |(?:                                             # or the v= param in all its forms
 217                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)?  # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
 218                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
 219                                  (?:.*?&)?                                    # any other preceding param (like /?s=tuff&v=xxxx)
 220                                  v=
 221                              )
 222                          ))
 223                          |youtu\.be/                                          # just youtu.be/xxxx
 224                          |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
 225                          )
 226                      )?                                                       # all until now is optional -> you can pass the naked ID
 227                      ([0-9A-Za-z_-]{11})                                      # here is it! the YouTube video ID
 228                      (?!.*?&list=)                                            # combined list/video URLs are handled by the playlist IE
 229                      (?(1).+)?                                                # if we found the ID, everything can follow
 230                      $"""
 231     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
 232     _formats = {
 233         '5': {'ext': 'flv', 'width': 400, 'height': 240},
 234         '6': {'ext': 'flv', 'width': 450, 'height': 270},
 235         '13': {'ext': '3gp'},
 236         '17': {'ext': '3gp', 'width': 176, 'height': 144},
 237         '18': {'ext': 'mp4', 'width': 640, 'height': 360},
 238         '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
 239         '34': {'ext': 'flv', 'width': 640, 'height': 360},
 240         '35': {'ext': 'flv', 'width': 854, 'height': 480},
 241         '36': {'ext': '3gp', 'width': 320, 'height': 240},
 242         '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
 243         '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
 244         '43': {'ext': 'webm', 'width': 640, 'height': 360},
 245         '44': {'ext': 'webm', 'width': 854, 'height': 480},
 246         '45': {'ext': 'webm', 'width': 1280, 'height': 720},
 247         '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
 248
 249
 250         # 3d videos
 251         '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20},
 252         '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
 253         '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
 254         '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
 255         '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
 256         '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
 257         '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
 258
 259         # Apple HTTP Live Streaming
 260         '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
 261         '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
 262         '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
 263         '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
 264         '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
 265         '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
 266         '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
 267
 268         # DASH mp4 video
 269         '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 270         '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 271         '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 272         '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 273         '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 274         '138': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 275         '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 276         '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 277         '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
 278         '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
 279         '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'h264'},
 280
 281         # Dash mp4 audio
 282         '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
 283         '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
 284         '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
 285
 286         # Dash webm
 287         '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 288         '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 289         '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 290         '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 291         '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 292         '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 293         '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'container': 'webm', 'vcodec': 'VP9'},
 294         '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 295         '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 296         '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 297         '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 298         '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 299         '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 300         '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 301         '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 302         '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 303         '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 304         '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 305
 306         # Dash webm audio
 307         '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50},
 308         '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
 309
 310         # RTMP (unnamed)
 311         '_rtmp': {'protocol': 'rtmp'},
 312     }
 313
 314     IE_NAME = 'youtube'
 315     _TESTS = [
 316         {
 317             'url': 'http://www.youtube.com/watch?v=BaW_jenozKc',
 318             'info_dict': {
 319                 'id': 'BaW_jenozKc',
 320                 'ext': 'mp4',
 321                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 322                 'uploader': 'Philipp Hagemeister',
 323                 'uploader_id': 'phihag',
 324                 'upload_date': '20121002',
 325                 '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 .',
 326                 'categories': ['Science & Technology'],
 327                 'like_count': int,
 328                 'dislike_count': int,
 329             }
 330         },
 331         {
 332             'url': 'http://www.youtube.com/watch?v=UxxajLWwzqY',
 333             'note': 'Test generic use_cipher_signature video (#897)',
 334             'info_dict': {
 335                 'id': 'UxxajLWwzqY',
 336                 'ext': 'mp4',
 337                 'upload_date': '20120506',
 338                 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
 339                 'description': 'md5:fea86fda2d5a5784273df5c7cc994d9f',
 340                 'uploader': 'Icona Pop',
 341                 'uploader_id': 'IconaPop',
 342             }
 343         },
 344         {
 345             'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
 346             'note': 'Test VEVO video with age protection (#956)',
 347             'info_dict': {
 348                 'id': '07FYdnEawAQ',
 349                 'ext': 'mp4',
 350                 'upload_date': '20130703',
 351                 'title': 'Justin Timberlake - Tunnel Vision (Explicit)',
 352                 'description': 'md5:64249768eec3bc4276236606ea996373',
 353                 'uploader': 'justintimberlakeVEVO',
 354                 'uploader_id': 'justintimberlakeVEVO',
 355             }
 356         },
 357         {
 358             'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
 359             'note': 'Embed-only video (#1746)',
 360             'info_dict': {
 361                 'id': 'yZIXLfi8CZQ',
 362                 'ext': 'mp4',
 363                 'upload_date': '20120608',
 364                 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
 365                 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
 366                 'uploader': 'SET India',
 367                 'uploader_id': 'setindia'
 368             }
 369         },
 370         {
 371             'url': 'http://www.youtube.com/watch?v=a9LDPn-MO4I',
 372             'note': '256k DASH audio (format 141) via DASH manifest',
 373             'info_dict': {
 374                 'id': 'a9LDPn-MO4I',
 375                 'ext': 'm4a',
 376                 'upload_date': '20121002',
 377                 'uploader_id': '8KVIDEO',
 378                 'description': '',
 379                 'uploader': '8KVIDEO',
 380                 'title': 'UHDTV TEST 8K VIDEO.mp4'
 381             },
 382             'params': {
 383                 'youtube_include_dash_manifest': True,
 384                 'format': '141',
 385             },
 386         },
 387         # DASH manifest with encrypted signature
 388         {
 389             'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
 390             'info_dict': {
 391                 'id': 'IB3lcPjvWLA',
 392                 'ext': 'm4a',
 393                 'title': 'Afrojack - The Spark ft. Spree Wilson',
 394                 'description': 'md5:9717375db5a9a3992be4668bbf3bc0a8',
 395                 'uploader': 'AfrojackVEVO',
 396                 'uploader_id': 'AfrojackVEVO',
 397                 'upload_date': '20131011',
 398             },
 399             'params': {
 400                 'youtube_include_dash_manifest': True,
 401                 'format': '141',
 402             },
 403         },
 404     ]
 405
 406     def __init__(self, *args, **kwargs):
 407         super(YoutubeIE, self).__init__(*args, **kwargs)
 408         self._player_cache = {}
 409
 410     def report_video_info_webpage_download(self, video_id):
 411         """Report attempt to download video info webpage."""
 412         self.to_screen('%s: Downloading video info webpage' % video_id)
 413
 414     def report_information_extraction(self, video_id):
 415         """Report attempt to extract video information."""
 416         self.to_screen('%s: Extracting video information' % video_id)
 417
 418     def report_unavailable_format(self, video_id, format):
 419         """Report extracted video URL."""
 420         self.to_screen('%s: Format %s not available' % (video_id, format))
 421
 422     def report_rtmp_download(self):
 423         """Indicate the download will use the RTMP protocol."""
 424         self.to_screen('RTMP download detected')
 425
 426     def _signature_cache_id(self, example_sig):
 427         """ Return a string representation of a signature """
 428         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
 429
 430     def _extract_signature_function(self, video_id, player_url, example_sig):
 431         id_m = re.match(
 432             r'.*-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P<ext>[a-z]+)$',
 433             player_url)
 434         if not id_m:
 435             raise ExtractorError('Cannot identify player %r' % player_url)
 436         player_type = id_m.group('ext')
 437         player_id = id_m.group('id')
 438
 439         # Read from filesystem cache
 440         func_id = '%s_%s_%s' % (
 441             player_type, player_id, self._signature_cache_id(example_sig))
 442         assert os.path.basename(func_id) == func_id
 443
 444         cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
 445         if cache_spec is not None:
 446             return lambda s: ''.join(s[i] for i in cache_spec)
 447
 448         if player_type == 'js':
 449             code = self._download_webpage(
 450                 player_url, video_id,
 451                 note='Downloading %s player %s' % (player_type, player_id),
 452                 errnote='Download of %s failed' % player_url)
 453             res = self._parse_sig_js(code)
 454         elif player_type == 'swf':
 455             urlh = self._request_webpage(
 456                 player_url, video_id,
 457                 note='Downloading %s player %s' % (player_type, player_id),
 458                 errnote='Download of %s failed' % player_url)
 459             code = urlh.read()
 460             res = self._parse_sig_swf(code)
 461         else:
 462             assert False, 'Invalid player type %r' % player_type
 463
 464         if cache_spec is None:
 465             test_string = ''.join(map(compat_chr, range(len(example_sig))))
 466             cache_res = res(test_string)
 467             cache_spec = [ord(c) for c in cache_res]
 468
 469         self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
 470         return res
 471
 472     def _print_sig_code(self, func, example_sig):
 473         def gen_sig_code(idxs):
 474             def _genslice(start, end, step):
 475                 starts = '' if start == 0 else str(start)
 476                 ends = (':%d' % (end+step)) if end + step >= 0 else ':'
 477                 steps = '' if step == 1 else (':%d' % step)
 478                 return 's[%s%s%s]' % (starts, ends, steps)
 479
 480             step = None
 481             start = '(Never used)'  # Quelch pyflakes warnings - start will be
 482                                     # set as soon as step is set
 483             for i, prev in zip(idxs[1:], idxs[:-1]):
 484                 if step is not None:
 485                     if i - prev == step:
 486                         continue
 487                     yield _genslice(start, prev, step)
 488                     step = None
 489                     continue
 490                 if i - prev in [-1, 1]:
 491                     step = i - prev
 492                     start = prev
 493                     continue
 494                 else:
 495                     yield 's[%d]' % prev
 496             if step is None:
 497                 yield 's[%d]' % i
 498             else:
 499                 yield _genslice(start, i, step)
 500
 501         test_string = ''.join(map(compat_chr, range(len(example_sig))))
 502         cache_res = func(test_string)
 503         cache_spec = [ord(c) for c in cache_res]
 504         expr_code = ' + '.join(gen_sig_code(cache_spec))
 505         signature_id_tuple = '(%s)' % (
 506             ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
 507         code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
 508                 '    return %s\n') % (signature_id_tuple, expr_code)
 509         self.to_screen('Extracted signature function:\n' + code)
 510
 511     def _parse_sig_js(self, jscode):
 512         funcname = self._search_regex(
 513             r'signature=([$a-zA-Z]+)', jscode,
 514              'Initial JS player signature function name')
 515
 516         jsi = JSInterpreter(jscode)
 517         initial_function = jsi.extract_function(funcname)
 518         return lambda s: initial_function([s])
 519
 520     def _parse_sig_swf(self, file_contents):
 521         swfi = SWFInterpreter(file_contents)
 522         TARGET_CLASSNAME = 'SignatureDecipher'
 523         searched_class = swfi.extract_class(TARGET_CLASSNAME)
 524         initial_function = swfi.extract_function(searched_class, 'decipher')
 525         return lambda s: initial_function([s])
 526
 527     def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
 528         """Turn the encrypted s field into a working signature"""
 529
 530         if player_url is None:
 531             raise ExtractorError('Cannot decrypt signature without player_url')
 532
 533         if player_url.startswith('//'):
 534             player_url = 'https:' + player_url
 535         try:
 536             player_id = (player_url, self._signature_cache_id(s))
 537             if player_id not in self._player_cache:
 538                 func = self._extract_signature_function(
 539                     video_id, player_url, s
 540                 )
 541                 self._player_cache[player_id] = func
 542             func = self._player_cache[player_id]
 543             if self._downloader.params.get('youtube_print_sig_code'):
 544                 self._print_sig_code(func, s)
 545             return func(s)
 546         except Exception as e:
 547             tb = traceback.format_exc()
 548             raise ExtractorError(
 549                 'Signature extraction failed: ' + tb, cause=e)
 550
 551     def _get_available_subtitles(self, video_id, webpage):
 552         try:
 553             sub_list = self._download_webpage(
 554                 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
 555                 video_id, note=False)
 556         except ExtractorError as err:
 557             self._downloader.report_warning('unable to download video subtitles: %s' % compat_str(err))
 558             return {}
 559         lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
 560
 561         sub_lang_list = {}
 562         for l in lang_list:
 563             lang = l[1]
 564             if lang in sub_lang_list:
 565                 continue
 566             params = compat_urllib_parse.urlencode({
 567                 'lang': lang,
 568                 'v': video_id,
 569                 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
 570                 'name': unescapeHTML(l[0]).encode('utf-8'),
 571             })
 572             url = 'https://www.youtube.com/api/timedtext?' + params
 573             sub_lang_list[lang] = url
 574         if not sub_lang_list:
 575             self._downloader.report_warning('video doesn\'t have subtitles')
 576             return {}
 577         return sub_lang_list
 578
 579     def _get_available_automatic_caption(self, video_id, webpage):
 580         """We need the webpage for getting the captions url, pass it as an
 581            argument to speed up the process."""
 582         sub_format = self._downloader.params.get('subtitlesformat', 'srt')
 583         self.to_screen('%s: Looking for automatic captions' % video_id)
 584         mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
 585         err_msg = 'Couldn\'t find automatic captions for %s' % video_id
 586         if mobj is None:
 587             self._downloader.report_warning(err_msg)
 588             return {}
 589         player_config = json.loads(mobj.group(1))
 590         try:
 591             args = player_config[u'args']
 592             caption_url = args[u'ttsurl']
 593             timestamp = args[u'timestamp']
 594             # We get the available subtitles
 595             list_params = compat_urllib_parse.urlencode({
 596                 'type': 'list',
 597                 'tlangs': 1,
 598                 'asrs': 1,
 599             })
 600             list_url = caption_url + '&' + list_params
 601             caption_list = self._download_xml(list_url, video_id)
 602             original_lang_node = caption_list.find('track')
 603             if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
 604                 self._downloader.report_warning('Video doesn\'t have automatic captions')
 605                 return {}
 606             original_lang = original_lang_node.attrib['lang_code']
 607
 608             sub_lang_list = {}
 609             for lang_node in caption_list.findall('target'):
 610                 sub_lang = lang_node.attrib['lang_code']
 611                 params = compat_urllib_parse.urlencode({
 612                     'lang': original_lang,
 613                     'tlang': sub_lang,
 614                     'fmt': sub_format,
 615                     'ts': timestamp,
 616                     'kind': 'asr',
 617                 })
 618                 sub_lang_list[sub_lang] = caption_url + '&' + params
 619             return sub_lang_list
 620         # An extractor error can be raise by the download process if there are
 621         # no automatic captions but there are subtitles
 622         except (KeyError, ExtractorError):
 623             self._downloader.report_warning(err_msg)
 624             return {}
 625
 626     @classmethod
 627     def extract_id(cls, url):
 628         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
 629         if mobj is None:
 630             raise ExtractorError('Invalid URL: %s' % url)
 631         video_id = mobj.group(2)
 632         return video_id
 633
 634     def _extract_from_m3u8(self, manifest_url, video_id):
 635         url_map = {}
 636         def _get_urls(_manifest):
 637             lines = _manifest.split('\n')
 638             urls = filter(lambda l: l and not l.startswith('#'),
 639                             lines)
 640             return urls
 641         manifest = self._download_webpage(manifest_url, video_id, 'Downloading formats manifest')
 642         formats_urls = _get_urls(manifest)
 643         for format_url in formats_urls:
 644             itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
 645             url_map[itag] = format_url
 646         return url_map
 647
 648     def _extract_annotations(self, video_id):
 649         url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
 650         return self._download_webpage(url, video_id, note='Searching for annotations.', errnote='Unable to download video annotations.')
 651
 652     def _real_extract(self, url):
 653         proto = (
 654             'http' if self._downloader.params.get('prefer_insecure', False)
 655             else 'https')
 656
 657         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
 658         mobj = re.search(self._NEXT_URL_RE, url)
 659         if mobj:
 660             url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
 661         video_id = self.extract_id(url)
 662
 663         # Get video webpage
 664         url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
 665         pref_cookies = [
 666             c for c in self._downloader.cookiejar
 667             if c.domain == '.youtube.com' and c.name == 'PREF']
 668         for pc in pref_cookies:
 669             if 'hl=' in pc.value:
 670                 pc.value = re.sub(r'hl=[^&]+', 'hl=en', pc.value)
 671             else:
 672                 if pc.value:
 673                     pc.value += '&'
 674                 pc.value += 'hl=en'
 675         video_webpage = self._download_webpage(url, video_id)
 676
 677         # Attempt to extract SWF player URL
 678         mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
 679         if mobj is not None:
 680             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
 681         else:
 682             player_url = None
 683
 684         # Get video info
 685         self.report_video_info_webpage_download(video_id)
 686         if re.search(r'player-age-gate-content">', video_webpage) is not None:
 687             age_gate = True
 688             # We simulate the access to the video from www.youtube.com/v/{video_id}
 689             # this can be viewed without login into Youtube
 690             data = compat_urllib_parse.urlencode({
 691                 'video_id': video_id,
 692                 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
 693                 'sts': self._search_regex(
 694                     r'"sts"\s*:\s*(\d+)', video_webpage, 'sts', default=''),
 695             })
 696             video_info_url = proto + '://www.youtube.com/get_video_info?' + data
 697             video_info_webpage = self._download_webpage(
 698                 video_info_url, video_id,
 699                 note='Refetching age-gated info webpage',
 700                 errnote='unable to download video info webpage')
 701             video_info = compat_parse_qs(video_info_webpage)
 702         else:
 703             age_gate = False
 704             for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
 705                 video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
 706                         % (video_id, el_type))
 707                 video_info_webpage = self._download_webpage(video_info_url, video_id,
 708                                         note=False,
 709                                         errnote='unable to download video info webpage')
 710                 video_info = compat_parse_qs(video_info_webpage)
 711                 if 'token' in video_info:
 712                     break
 713         if 'token' not in video_info:
 714             if 'reason' in video_info:
 715                 raise ExtractorError(
 716                     'YouTube said: %s' % video_info['reason'][0],
 717                     expected=True, video_id=video_id)
 718             else:
 719                 raise ExtractorError(
 720                     '"token" parameter not in video info for unknown reason',
 721                     video_id=video_id)
 722
 723         if 'view_count' in video_info:
 724             view_count = int(video_info['view_count'][0])
 725         else:
 726             view_count = None
 727
 728         # Check for "rental" videos
 729         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 730             raise ExtractorError('"rental" videos not supported')
 731
 732         # Start extracting information
 733         self.report_information_extraction(video_id)
 734
 735         # uploader
 736         if 'author' not in video_info:
 737             raise ExtractorError('Unable to extract uploader name')
 738         video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
 739
 740         # uploader_id
 741         video_uploader_id = None
 742         mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
 743         if mobj is not None:
 744             video_uploader_id = mobj.group(1)
 745         else:
 746             self._downloader.report_warning('unable to extract uploader nickname')
 747
 748         # title
 749         if 'title' in video_info:
 750             video_title = video_info['title'][0]
 751         else:
 752             self._downloader.report_warning('Unable to extract video title')
 753             video_title = '_'
 754
 755         # thumbnail image
 756         # We try first to get a high quality image:
 757         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
 758                             video_webpage, re.DOTALL)
 759         if m_thumb is not None:
 760             video_thumbnail = m_thumb.group(1)
 761         elif 'thumbnail_url' not in video_info:
 762             self._downloader.report_warning('unable to extract video thumbnail')
 763             video_thumbnail = None
 764         else:   # don't panic if we can't find it
 765             video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
 766
 767         # upload date
 768         upload_date = None
 769         mobj = re.search(r'(?s)id="eow-date.*?>(.*?)</span>', video_webpage)
 770         if mobj is None:
 771             mobj = re.search(
 772                 r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
 773                 video_webpage)
 774         if mobj is not None:
 775             upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
 776             upload_date = unified_strdate(upload_date)
 777
 778         m_cat_container = self._search_regex(
 779             r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
 780             video_webpage, 'categories', fatal=False)
 781         if m_cat_container:
 782             category = self._html_search_regex(
 783                 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
 784                 default=None)
 785             video_categories = None if category is None else [category]
 786         else:
 787             video_categories = None
 788
 789         # description
 790         video_description = get_element_by_id("eow-description", video_webpage)
 791         if video_description:
 792             video_description = re.sub(r'''(?x)
 793                 <a\s+
 794                     (?:[a-zA-Z-]+="[^"]+"\s+)*?
 795                     title="([^"]+)"\s+
 796                     (?:[a-zA-Z-]+="[^"]+"\s+)*?
 797                     class="yt-uix-redirect-link"\s*>
 798                 [^<]+
 799                 </a>
 800             ''', r'\1', video_description)
 801             video_description = clean_html(video_description)
 802         else:
 803             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
 804             if fd_mobj:
 805                 video_description = unescapeHTML(fd_mobj.group(1))
 806             else:
 807                 video_description = ''
 808
 809         def _extract_count(count_name):
 810             count = self._search_regex(
 811                 r'id="watch-%s"[^>]*>.*?([\d,]+)\s*</span>' % re.escape(count_name),
 812                 video_webpage, count_name, default=None)
 813             if count is not None:
 814                 return int(count.replace(',', ''))
 815             return None
 816         like_count = _extract_count('like')
 817         dislike_count = _extract_count('dislike')
 818
 819         # subtitles
 820         video_subtitles = self.extract_subtitles(video_id, video_webpage)
 821
 822         if self._downloader.params.get('listsubtitles', False):
 823             self._list_available_subtitles(video_id, video_webpage)
 824             return
 825
 826         if 'length_seconds' not in video_info:
 827             self._downloader.report_warning('unable to extract video duration')
 828             video_duration = None
 829         else:
 830             video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
 831
 832         # annotations
 833         video_annotations = None
 834         if self._downloader.params.get('writeannotations', False):
 835                 video_annotations = self._extract_annotations(video_id)
 836
 837         # Decide which formats to download
 838         try:
 839             mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
 840             if not mobj:
 841                 raise ValueError('Could not find vevo ID')
 842             json_code = uppercase_escape(mobj.group(1))
 843             ytplayer_config = json.loads(json_code)
 844             args = ytplayer_config['args']
 845             # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
 846             # this signatures are encrypted
 847             if 'url_encoded_fmt_stream_map' not in args:
 848                 raise ValueError('No stream_map present')  # caught below
 849             re_signature = re.compile(r'[&,]s=')
 850             m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
 851             if m_s is not None:
 852                 self.to_screen('%s: Encrypted signatures detected.' % video_id)
 853                 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
 854             m_s = re_signature.search(args.get('adaptive_fmts', ''))
 855             if m_s is not None:
 856                 if 'adaptive_fmts' in video_info:
 857                     video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
 858                 else:
 859                     video_info['adaptive_fmts'] = [args['adaptive_fmts']]
 860         except ValueError:
 861             pass
 862
 863         def _map_to_format_list(urlmap):
 864             formats = []
 865             for itag, video_real_url in urlmap.items():
 866                 dct = {
 867                     'format_id': itag,
 868                     'url': video_real_url,
 869                     'player_url': player_url,
 870                 }
 871                 if itag in self._formats:
 872                     dct.update(self._formats[itag])
 873                 formats.append(dct)
 874             return formats
 875
 876         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
 877             self.report_rtmp_download()
 878             formats = [{
 879                 'format_id': '_rtmp',
 880                 'protocol': 'rtmp',
 881                 'url': video_info['conn'][0],
 882                 'player_url': player_url,
 883             }]
 884         elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
 885             encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
 886             if 'rtmpe%3Dyes' in encoded_url_map:
 887                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
 888             url_map = {}
 889             for url_data_str in encoded_url_map.split(','):
 890                 url_data = compat_parse_qs(url_data_str)
 891                 if 'itag' not in url_data or 'url' not in url_data:
 892                     continue
 893                 format_id = url_data['itag'][0]
 894                 url = url_data['url'][0]
 895
 896                 if 'sig' in url_data:
 897                     url += '&signature=' + url_data['sig'][0]
 898                 elif 's' in url_data:
 899                     encrypted_sig = url_data['s'][0]
 900
 901                     if not age_gate:
 902                         jsplayer_url_json = self._search_regex(
 903                             r'"assets":.+?"js":\s*("[^"]+")',
 904                             video_webpage, 'JS player URL')
 905                         player_url = json.loads(jsplayer_url_json)
 906                     if player_url is None:
 907                         player_url_json = self._search_regex(
 908                             r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
 909                             video_webpage, 'age gate player URL')
 910                         player_url = json.loads(player_url_json)
 911
 912                     if self._downloader.params.get('verbose'):
 913                         if player_url is None:
 914                             player_version = 'unknown'
 915                             player_desc = 'unknown'
 916                         else:
 917                             if player_url.endswith('swf'):
 918                                 player_version = self._search_regex(
 919                                     r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
 920                                     'flash player', fatal=False)
 921                                 player_desc = 'flash player %s' % player_version
 922                             else:
 923                                 player_version = self._search_regex(
 924                                     r'html5player-([^/]+?)(?:/html5player)?\.js',
 925                                     player_url,
 926                                     'html5 player', fatal=False)
 927                                 player_desc = 'html5 player %s' % player_version
 928
 929                         parts_sizes = self._signature_cache_id(encrypted_sig)
 930                         self.to_screen('{%s} signature length %s, %s' %
 931                             (format_id, parts_sizes, player_desc))
 932
 933                     signature = self._decrypt_signature(
 934                         encrypted_sig, video_id, player_url, age_gate)
 935                     url += '&signature=' + signature
 936                 if 'ratebypass' not in url:
 937                     url += '&ratebypass=yes'
 938                 url_map[format_id] = url
 939             formats = _map_to_format_list(url_map)
 940         elif video_info.get('hlsvp'):
 941             manifest_url = video_info['hlsvp'][0]
 942             url_map = self._extract_from_m3u8(manifest_url, video_id)
 943             formats = _map_to_format_list(url_map)
 944         else:
 945             raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
 946
 947         # Look for the DASH manifest
 948         if self._downloader.params.get('youtube_include_dash_manifest', True):
 949             try:
 950                 # The DASH manifest used needs to be the one from the original video_webpage.
 951                 # The one found in get_video_info seems to be using different signatures.
 952                 # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
 953                 # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
 954                 # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
 955                 if age_gate:
 956                     dash_manifest_url = video_info.get('dashmpd')[0]
 957                 else:
 958                     dash_manifest_url = ytplayer_config['args']['dashmpd']
 959                 def decrypt_sig(mobj):
 960                     s = mobj.group(1)
 961                     dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
 962                     return '/signature/%s' % dec_s
 963                 dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
 964                 dash_doc = self._download_xml(
 965                     dash_manifest_url, video_id,
 966                     note='Downloading DASH manifest',
 967                     errnote='Could not download DASH manifest')
 968                 for r in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
 969                     url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
 970                     if url_el is None:
 971                         continue
 972                     format_id = r.attrib['id']
 973                     video_url = url_el.text
 974                     filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
 975                     f = {
 976                         'format_id': format_id,
 977                         'url': video_url,
 978                         'width': int_or_none(r.attrib.get('width')),
 979                         'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
 980                         'asr': int_or_none(r.attrib.get('audioSamplingRate')),
 981                         'filesize': filesize,
 982                     }
 983                     try:
 984                         existing_format = next(
 985                             fo for fo in formats
 986                             if fo['format_id'] == format_id)
 987                     except StopIteration:
 988                         f.update(self._formats.get(format_id, {}))
 989                         formats.append(f)
 990                     else:
 991                         existing_format.update(f)
 992
 993             except (ExtractorError, KeyError) as e:
 994                 self.report_warning('Skipping DASH manifest: %s' % e, video_id)
 995
 996         self._sort_formats(formats)
 997
 998         return {
 999             'id':           video_id,
1000             'uploader':     video_uploader,
1001             'uploader_id':  video_uploader_id,
1002             'upload_date':  upload_date,
1003             'title':        video_title,
1004             'thumbnail':    video_thumbnail,
1005             'description':  video_description,
1006             'categories':   video_categories,
1007             'subtitles':    video_subtitles,
1008             'duration':     video_duration,
1009             'age_limit':    18 if age_gate else 0,
1010             'annotations':  video_annotations,
1011             'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1012             'view_count':   view_count,
1013             'like_count': like_count,
1014             'dislike_count': dislike_count,
1015             'formats':      formats,
1016         }
1017
1018 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1019     IE_DESC = 'YouTube.com playlists'
1020     _VALID_URL = r"""(?x)(?:
1021                         (?:https?://)?
1022                         (?:\w+\.)?
1023                         youtube\.com/
1024                         (?:
1025                            (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/videoseries)
1026                            \? (?:.*?&)*? (?:p|a|list)=
1027                         |  p/
1028                         )
1029                         (
1030                             (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
1031                             # Top tracks, they can also include dots
1032                             |(?:MC)[\w\.]*
1033                         )
1034                         .*
1035                      |
1036                         ((?:PL|LL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,})
1037                      )"""
1038     _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1039     _MORE_PAGES_INDICATOR = r'data-link-type="next"'
1040     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)'
1041     IE_NAME = 'youtube:playlist'
1042     _TESTS = [{
1043         'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1044         'info_dict': {
1045             'title': 'ytdl test PL',
1046         },
1047         'playlist_count': 3,
1048     }, {
1049         'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1050         'info_dict': {
1051             'title': 'YDL_Empty_List',
1052         },
1053         'playlist_count': 0,
1054     }, {
1055         'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
1056         'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1057         'info_dict': {
1058             'title': '29C3: Not my department',
1059         },
1060         'playlist_count': 95,
1061     }, {
1062         'note': 'issue #673',
1063         'url': 'PLBB231211A4F62143',
1064         'info_dict': {
1065             'title': '[OLD]Team Fortress 2 (Class-based LP)',
1066         },
1067         'playlist_mincount': 26,
1068     }, {
1069         'note': 'Large playlist',
1070         'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
1071         'info_dict': {
1072             'title': 'Uploads from Cauchemar',
1073         },
1074         'playlist_mincount': 799,
1075     }, {
1076         'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1077         'info_dict': {
1078             'title': 'YDL_safe_search',
1079         },
1080         'playlist_count': 2,
1081     }, {
1082         'note': 'embedded',
1083         'url': 'http://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1084         'playlist_count': 4,
1085         'info_dict': {
1086             'title': 'JODA15',
1087         }
1088     }, {
1089         'note': 'Embedded SWF player',
1090         'url': 'http://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
1091         'playlist_count': 4,
1092         'info_dict': {
1093             'title': 'JODA7',
1094         }
1095     }]
1096
1097     def _real_initialize(self):
1098         self._login()
1099
1100     def _ids_to_results(self, ids):
1101         return [
1102             self.url_result(vid_id, 'Youtube', video_id=vid_id)
1103             for vid_id in ids]
1104
1105     def _extract_mix(self, playlist_id):
1106         # The mixes are generated from a a single video
1107         # the id of the playlist is just 'RD' + video_id
1108         url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1109         webpage = self._download_webpage(
1110             url, playlist_id, 'Downloading Youtube mix')
1111         search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1112         title_span = (
1113             search_title('playlist-title') or
1114             search_title('title long-title') or
1115             search_title('title'))
1116         title = clean_html(title_span)
1117         ids = orderedSet(re.findall(
1118             r'''(?xs)data-video-username=".*?".*?
1119                        href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
1120             webpage))
1121         url_results = self._ids_to_results(ids)
1122
1123         return self.playlist_result(url_results, playlist_id, title)
1124
1125     def _real_extract(self, url):
1126         # Extract playlist id
1127         mobj = re.match(self._VALID_URL, url)
1128         if mobj is None:
1129             raise ExtractorError('Invalid URL: %s' % url)
1130         playlist_id = mobj.group(1) or mobj.group(2)
1131
1132         # Check if it's a video-specific URL
1133         query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1134         if 'v' in query_dict:
1135             video_id = query_dict['v'][0]
1136             if self._downloader.params.get('noplaylist'):
1137                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1138                 return self.url_result(video_id, 'Youtube', video_id=video_id)
1139             else:
1140                 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1141
1142         if playlist_id.startswith('RD'):
1143             # Mixes require a custom extraction process
1144             return self._extract_mix(playlist_id)
1145         if playlist_id.startswith('TL'):
1146             raise ExtractorError('For downloading YouTube.com top lists, use '
1147                 'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
1148
1149         url = self._TEMPLATE_URL % playlist_id
1150         page = self._download_webpage(url, playlist_id)
1151         more_widget_html = content_html = page
1152
1153         # Check if the playlist exists or is private
1154         if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None:
1155             raise ExtractorError(
1156                 'The playlist doesn\'t exist or is private, use --username or '
1157                 '--netrc to access it.',
1158                 expected=True)
1159
1160         # Extract the video ids from the playlist pages
1161         ids = []
1162
1163         for page_num in itertools.count(1):
1164             matches = re.finditer(self._VIDEO_RE, content_html)
1165             # We remove the duplicates and the link with index 0
1166             # (it's not the first video of the playlist)
1167             new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1168             ids.extend(new_ids)
1169
1170             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1171             if not mobj:
1172                 break
1173
1174             more = self._download_json(
1175                 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1176                 'Downloading page #%s' % page_num,
1177                 transform_source=uppercase_escape)
1178             content_html = more['content_html']
1179             more_widget_html = more['load_more_widget_html']
1180
1181         playlist_title = self._html_search_regex(
1182             r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1183             page, 'title')
1184
1185         url_results = self._ids_to_results(ids)
1186         return self.playlist_result(url_results, playlist_id, playlist_title)
1187
1188
1189 class YoutubeTopListIE(YoutubePlaylistIE):
1190     IE_NAME = 'youtube:toplist'
1191     IE_DESC = ('YouTube.com top lists, "yttoplist:{channel}:{list title}"'
1192         ' (Example: "yttoplist:music:Top Tracks")')
1193     _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
1194     _TESTS = [{
1195         'url': 'yttoplist:music:Trending',
1196         'playlist_mincount': 5,
1197         'skip': 'Only works for logged-in users',
1198     }]
1199
1200     def _real_extract(self, url):
1201         mobj = re.match(self._VALID_URL, url)
1202         channel = mobj.group('chann')
1203         title = mobj.group('title')
1204         query = compat_urllib_parse.urlencode({'title': title})
1205         channel_page = self._download_webpage(
1206             'https://www.youtube.com/%s' % channel, title)
1207         link = self._html_search_regex(
1208             r'''(?x)
1209                 <a\s+href="([^"]+)".*?>\s*
1210                 <span\s+class="branded-page-module-title-text">\s*
1211                 <span[^>]*>.*?%s.*?</span>''' % re.escape(query),
1212             channel_page, 'list')
1213         url = compat_urlparse.urljoin('https://www.youtube.com/', link)
1214
1215         video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
1216         ids = []
1217         # sometimes the webpage doesn't contain the videos
1218         # retry until we get them
1219         for i in itertools.count(0):
1220             msg = 'Downloading Youtube mix'
1221             if i > 0:
1222                 msg += ', retry #%d' % i
1223
1224             webpage = self._download_webpage(url, title, msg)
1225             ids = orderedSet(re.findall(video_re, webpage))
1226             if ids:
1227                 break
1228         url_results = self._ids_to_results(ids)
1229         return self.playlist_result(url_results, playlist_title=title)
1230
1231
1232 class YoutubeChannelIE(InfoExtractor):
1233     IE_DESC = 'YouTube.com channels'
1234     _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
1235     _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
1236     _MORE_PAGES_URL = 'https://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
1237     IE_NAME = 'youtube:channel'
1238     _TESTS = [{
1239         'note': 'paginated channel',
1240         'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
1241         'playlist_mincount': 91,
1242     }]
1243
1244     def extract_videos_from_page(self, page):
1245         ids_in_page = []
1246         for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1247             if mobj.group(1) not in ids_in_page:
1248                 ids_in_page.append(mobj.group(1))
1249         return ids_in_page
1250
1251     def _real_extract(self, url):
1252         # Extract channel id
1253         mobj = re.match(self._VALID_URL, url)
1254         if mobj is None:
1255             raise ExtractorError('Invalid URL: %s' % url)
1256
1257         # Download channel page
1258         channel_id = mobj.group(1)
1259         video_ids = []
1260         url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1261         channel_page = self._download_webpage(url, channel_id)
1262         autogenerated = re.search(r'''(?x)
1263                 class="[^"]*?(?:
1264                     channel-header-autogenerated-label|
1265                     yt-channel-title-autogenerated
1266                 )[^"]*"''', channel_page) is not None
1267
1268         if autogenerated:
1269             # The videos are contained in a single page
1270             # the ajax pages can't be used, they are empty
1271             video_ids = self.extract_videos_from_page(channel_page)
1272         else:
1273             # Download all channel pages using the json-based channel_ajax query
1274             for pagenum in itertools.count(1):
1275                 url = self._MORE_PAGES_URL % (pagenum, channel_id)
1276                 page = self._download_json(
1277                     url, channel_id, note='Downloading page #%s' % pagenum,
1278                     transform_source=uppercase_escape)
1279
1280                 ids_in_page = self.extract_videos_from_page(page['content_html'])
1281                 video_ids.extend(ids_in_page)
1282
1283                 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
1284                     break
1285
1286         self._downloader.to_screen('[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
1287
1288         url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
1289                        for video_id in video_ids]
1290         return self.playlist_result(url_entries, channel_id)
1291
1292
1293 class YoutubeUserIE(InfoExtractor):
1294     IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
1295     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
1296     _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
1297     _GDATA_PAGE_SIZE = 50
1298     _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1299     IE_NAME = 'youtube:user'
1300
1301     _TESTS = [{
1302         'url': 'https://www.youtube.com/user/TheLinuxFoundation',
1303         'playlist_mincount': 320,
1304         'info_dict': {
1305             'title': 'TheLinuxFoundation',
1306         }
1307     }, {
1308         'url': 'ytuser:phihag',
1309         'only_matching': True,
1310     }]
1311
1312     @classmethod
1313     def suitable(cls, url):
1314         # Don't return True if the url can be extracted with other youtube
1315         # extractor, the regex would is too permissive and it would match.
1316         other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1317         if any(ie.suitable(url) for ie in other_ies): return False
1318         else: return super(YoutubeUserIE, cls).suitable(url)
1319
1320     def _real_extract(self, url):
1321         # Extract username
1322         mobj = re.match(self._VALID_URL, url)
1323         if mobj is None:
1324             raise ExtractorError('Invalid URL: %s' % url)
1325
1326         username = mobj.group(1)
1327
1328         # Download video ids using YouTube Data API. Result size per
1329         # query is limited (currently to 50 videos) so we need to query
1330         # page by page until there are no video ids - it means we got
1331         # all of them.
1332
1333         def download_page(pagenum):
1334             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1335
1336             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1337             page = self._download_webpage(
1338                 gdata_url, username,
1339                 'Downloading video ids from %d to %d' % (
1340                     start_index, start_index + self._GDATA_PAGE_SIZE))
1341
1342             try:
1343                 response = json.loads(page)
1344             except ValueError as err:
1345                 raise ExtractorError('Invalid JSON in API response: ' + compat_str(err))
1346             if 'entry' not in response['feed']:
1347                 return
1348
1349             # Extract video identifiers
1350             entries = response['feed']['entry']
1351             for entry in entries:
1352                 title = entry['title']['$t']
1353                 video_id = entry['id']['$t'].split('/')[-1]
1354                 yield {
1355                     '_type': 'url',
1356                     'url': video_id,
1357                     'ie_key': 'Youtube',
1358                     'id': video_id,
1359                     'title': title,
1360                 }
1361         url_results = OnDemandPagedList(download_page, self._GDATA_PAGE_SIZE)
1362
1363         return self.playlist_result(url_results, playlist_title=username)
1364
1365
1366 class YoutubeSearchIE(SearchInfoExtractor):
1367     IE_DESC = 'YouTube.com searches'
1368     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1369     _MAX_RESULTS = 1000
1370     IE_NAME = 'youtube:search'
1371     _SEARCH_KEY = 'ytsearch'
1372
1373     def _get_n_results(self, query, n):
1374         """Get a specified number of results for a query"""
1375
1376         video_ids = []
1377         pagenum = 0
1378         limit = n
1379         PAGE_SIZE = 50
1380
1381         while (PAGE_SIZE * pagenum) < limit:
1382             result_url = self._API_URL % (
1383                 compat_urllib_parse.quote_plus(query.encode('utf-8')),
1384                 (PAGE_SIZE * pagenum) + 1)
1385             data_json = self._download_webpage(
1386                 result_url, video_id='query "%s"' % query,
1387                 note='Downloading page %s' % (pagenum + 1),
1388                 errnote='Unable to download API page')
1389             data = json.loads(data_json)
1390             api_response = data['data']
1391
1392             if 'items' not in api_response:
1393                 raise ExtractorError(
1394                     '[youtube] No video results', expected=True)
1395
1396             new_ids = list(video['id'] for video in api_response['items'])
1397             video_ids += new_ids
1398
1399             limit = min(n, api_response['totalItems'])
1400             pagenum += 1
1401
1402         if len(video_ids) > n:
1403             video_ids = video_ids[:n]
1404         videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1405                   for video_id in video_ids]
1406         return self.playlist_result(videos, query)
1407
1408
1409 class YoutubeSearchDateIE(YoutubeSearchIE):
1410     IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1411     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1412     _SEARCH_KEY = 'ytsearchdate'
1413     IE_DESC = 'YouTube.com searches, newest videos first'
1414
1415
1416 class YoutubeSearchURLIE(InfoExtractor):
1417     IE_DESC = 'YouTube.com search URLs'
1418     IE_NAME = 'youtube:search_url'
1419     _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1420     _TESTS = [{
1421         'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
1422         'playlist_mincount': 5,
1423         'info_dict': {
1424             'title': 'youtube-dl test video',
1425         }
1426     }]
1427
1428     def _real_extract(self, url):
1429         mobj = re.match(self._VALID_URL, url)
1430         query = compat_urllib_parse.unquote_plus(mobj.group('query'))
1431
1432         webpage = self._download_webpage(url, query)
1433         result_code = self._search_regex(
1434             r'(?s)<ol class="item-section"(.*?)</ol>', webpage, 'result HTML')
1435
1436         part_codes = re.findall(
1437             r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1438         entries = []
1439         for part_code in part_codes:
1440             part_title = self._html_search_regex(
1441                 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1442             part_url_snippet = self._html_search_regex(
1443                 r'(?s)href="([^"]+)"', part_code, 'item URL')
1444             part_url = compat_urlparse.urljoin(
1445                 'https://www.youtube.com/', part_url_snippet)
1446             entries.append({
1447                 '_type': 'url',
1448                 'url': part_url,
1449                 'title': part_title,
1450             })
1451
1452         return {
1453             '_type': 'playlist',
1454             'entries': entries,
1455             'title': query,
1456         }
1457
1458
1459 class YoutubeShowIE(InfoExtractor):
1460     IE_DESC = 'YouTube.com (multi-season) shows'
1461     _VALID_URL = r'https?://www\.youtube\.com/show/(?P<id>[^?#]*)'
1462     IE_NAME = 'youtube:show'
1463     _TESTS = [{
1464         'url': 'http://www.youtube.com/show/airdisasters',
1465         'playlist_mincount': 3,
1466         'info_dict': {
1467             'id': 'airdisasters',
1468             'title': 'Air Disasters',
1469         }
1470     }]
1471
1472     def _real_extract(self, url):
1473         mobj = re.match(self._VALID_URL, url)
1474         playlist_id = mobj.group('id')
1475         webpage = self._download_webpage(
1476             url, playlist_id, 'Downloading show webpage')
1477         # There's one playlist for each season of the show
1478         m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1479         self.to_screen('%s: Found %s seasons' % (playlist_id, len(m_seasons)))
1480         entries = [
1481             self.url_result(
1482                 'https://www.youtube.com' + season.group(1), 'YoutubePlaylist')
1483             for season in m_seasons
1484         ]
1485         title = self._og_search_title(webpage, fatal=False)
1486
1487         return {
1488             '_type': 'playlist',
1489             'id': playlist_id,
1490             'title': title,
1491             'entries': entries,
1492         }
1493
1494
1495 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1496     """
1497     Base class for extractors that fetch info from
1498     http://www.youtube.com/feed_ajax
1499     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1500     """
1501     _LOGIN_REQUIRED = True
1502     # use action_load_personal_feed instead of action_load_system_feed
1503     _PERSONAL_FEED = False
1504
1505     @property
1506     def _FEED_TEMPLATE(self):
1507         action = 'action_load_system_feed'
1508         if self._PERSONAL_FEED:
1509             action = 'action_load_personal_feed'
1510         return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1511
1512     @property
1513     def IE_NAME(self):
1514         return 'youtube:%s' % self._FEED_NAME
1515
1516     def _real_initialize(self):
1517         self._login()
1518
1519     def _real_extract(self, url):
1520         feed_entries = []
1521         paging = 0
1522         for i in itertools.count(1):
1523             info = self._download_json(self._FEED_TEMPLATE % paging,
1524                                           '%s feed' % self._FEED_NAME,
1525                                           'Downloading page %s' % i)
1526             feed_html = info.get('feed_html') or info.get('content_html')
1527             load_more_widget_html = info.get('load_more_widget_html') or feed_html
1528             m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1529             ids = orderedSet(m.group(1) for m in m_ids)
1530             feed_entries.extend(
1531                 self.url_result(video_id, 'Youtube', video_id=video_id)
1532                 for video_id in ids)
1533             mobj = re.search(
1534                 r'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)',
1535                 load_more_widget_html)
1536             if mobj is None:
1537                 break
1538             paging = mobj.group('paging')
1539         return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1540
1541 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1542     IE_DESC = 'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1543     _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1544     _FEED_NAME = 'recommended'
1545     _PLAYLIST_TITLE = 'Youtube Recommended videos'
1546
1547 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1548     IE_DESC = 'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1549     _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1550     _FEED_NAME = 'watch_later'
1551     _PLAYLIST_TITLE = 'Youtube Watch Later'
1552     _PERSONAL_FEED = True
1553
1554 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1555     IE_DESC = 'Youtube watch history, "ythistory" keyword (requires authentication)'
1556     _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory'
1557     _FEED_NAME = 'history'
1558     _PERSONAL_FEED = True
1559     _PLAYLIST_TITLE = 'Youtube Watch History'
1560
1561 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1562     IE_NAME = 'youtube:favorites'
1563     IE_DESC = 'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1564     _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1565     _LOGIN_REQUIRED = True
1566
1567     def _real_extract(self, url):
1568         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1569         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
1570         return self.url_result(playlist_id, 'YoutubePlaylist')
1571
1572
1573 class YoutubeSubscriptionsIE(YoutubePlaylistIE):
1574     IE_NAME = 'youtube:subscriptions'
1575     IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1576     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1577     _TESTS = []
1578
1579     def _real_extract(self, url):
1580         title = 'Youtube Subscriptions'
1581         page = self._download_webpage('https://www.youtube.com/feed/subscriptions', title)
1582
1583         # The extraction process is the same as for playlists, but the regex
1584         # for the video ids doesn't contain an index
1585         ids = []
1586         more_widget_html = content_html = page
1587
1588         for page_num in itertools.count(1):
1589             matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
1590             new_ids = orderedSet(matches)
1591             ids.extend(new_ids)
1592
1593             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1594             if not mobj:
1595                 break
1596
1597             more = self._download_json(
1598                 'https://youtube.com/%s' % mobj.group('more'), title,
1599                 'Downloading page #%s' % page_num,
1600                 transform_source=uppercase_escape)
1601             content_html = more['content_html']
1602             more_widget_html = more['load_more_widget_html']
1603
1604         return {
1605             '_type': 'playlist',
1606             'title': title,
1607             'entries': self._ids_to_results(ids),
1608         }
1609
1610
1611 class YoutubeTruncatedURLIE(InfoExtractor):
1612     IE_NAME = 'youtube:truncated_url'
1613     IE_DESC = False  # Do not list
1614     _VALID_URL = r'''(?x)
1615         (?:https?://)?[^/]+/watch\?(?:
1616             feature=[a-z_]+|
1617             annotation_id=annotation_[^&]+
1618         )?$|
1619         (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
1620     '''
1621
1622     _TESTS = [{
1623         'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1624         'only_matching': True,
1625     }, {
1626         'url': 'http://www.youtube.com/watch?',
1627         'only_matching': True,
1628     }]
1629
1630     def _real_extract(self, url):
1631         raise ExtractorError(
1632             'Did you forget to quote the URL? Remember that & is a meta '
1633             'character in most shells, so you want to put the URL in quotes, '
1634             'like  youtube-dl '
1635             '"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1636             ' or simply  youtube-dl BaW_jenozKc  .',
1637             expected=True)