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 time
  11 import traceback
  12
  13 from .common import InfoExtractor, SearchInfoExtractor
  14 from ..jsinterp import JSInterpreter
  15 from ..swfinterp import SWFInterpreter
  16 from ..compat import (
  17     compat_chr,
  18     compat_parse_qs,
  19     compat_urllib_parse,
  20     compat_urllib_request,
  21     compat_urlparse,
  22     compat_str,
  23 )
  24 from ..utils import (
  25     clean_html,
  26     ExtractorError,
  27     float_or_none,
  28     get_element_by_attribute,
  29     get_element_by_id,
  30     int_or_none,
  31     orderedSet,
  32     unescapeHTML,
  33     unified_strdate,
  34     uppercase_escape,
  35 )
  36
  37
  38 class YoutubeBaseInfoExtractor(InfoExtractor):
  39     """Provide base functions for Youtube extractors"""
  40     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  41     _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor'
  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         self._set_cookie(
  48             '.youtube.com', 'PREF', 'f1=50000000&hl=en',
  49             # YouTube sets the expire time to about two months
  50             expire_time=time.time() + 2 * 30 * 24 * 3600)
  51
  52     def _ids_to_results(self, ids):
  53         return [
  54             self.url_result(vid_id, 'Youtube', video_id=vid_id)
  55             for vid_id in ids]
  56
  57     def _login(self):
  58         """
  59         Attempt to log in to YouTube.
  60         True is returned if successful or skipped.
  61         False is returned if login failed.
  62
  63         If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
  64         """
  65         (username, password) = self._get_login_info()
  66         # No authentication to be performed
  67         if username is None:
  68             if self._LOGIN_REQUIRED:
  69                 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  70             return True
  71
  72         login_page = self._download_webpage(
  73             self._LOGIN_URL, None,
  74             note='Downloading login page',
  75             errnote='unable to fetch login page', fatal=False)
  76         if login_page is False:
  77             return
  78
  79         galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
  80                                   login_page, 'Login GALX parameter')
  81
  82         # Log in
  83         login_form_strs = {
  84             'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
  85             'Email': username,
  86             'GALX': galx,
  87             'Passwd': password,
  88
  89             'PersistentCookie': 'yes',
  90             '_utf8': '霱',
  91             'bgresponse': 'js_disabled',
  92             'checkConnection': '',
  93             'checkedDomains': 'youtube',
  94             'dnConn': '',
  95             'pstMsg': '0',
  96             'rmShown': '1',
  97             'secTok': '',
  98             'signIn': 'Sign in',
  99             'timeStmp': '',
 100             'service': 'youtube',
 101             'uilel': '3',
 102             'hl': 'en_US',
 103         }
 104
 105         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
 106         # chokes on unicode
 107         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items())
 108         login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
 109
 110         req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
 111         login_results = self._download_webpage(
 112             req, None,
 113             note='Logging in', errnote='unable to log in', fatal=False)
 114         if login_results is False:
 115             return False
 116
 117         if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
 118             raise ExtractorError('Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
 119
 120         # Two-Factor
 121         # TODO add SMS and phone call support - these require making a request and then prompting the user
 122
 123         if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', login_results) is not None:
 124             tfa_code = self._get_tfa_info()
 125
 126             if tfa_code is None:
 127                 self._downloader.report_warning('Two-factor authentication required. Provide it with --twofactor <code>')
 128                 self._downloader.report_warning('(Note that only TOTP (Google Authenticator App) codes work at this time.)')
 129                 return False
 130
 131             # Unlike the first login form, secTok and timeStmp are both required for the TFA form
 132
 133             match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
 134             if match is None:
 135                 self._downloader.report_warning('Failed to get secTok - did the page structure change?')
 136             secTok = match.group(1)
 137             match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
 138             if match is None:
 139                 self._downloader.report_warning('Failed to get timeStmp - did the page structure change?')
 140             timeStmp = match.group(1)
 141
 142             tfa_form_strs = {
 143                 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
 144                 'smsToken': '',
 145                 'smsUserPin': tfa_code,
 146                 'smsVerifyPin': 'Verify',
 147
 148                 'PersistentCookie': 'yes',
 149                 'checkConnection': '',
 150                 'checkedDomains': 'youtube',
 151                 'pstMsg': '1',
 152                 'secTok': secTok,
 153                 'timeStmp': timeStmp,
 154                 'service': 'youtube',
 155                 'hl': 'en_US',
 156             }
 157             tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in tfa_form_strs.items())
 158             tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
 159
 160             tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
 161             tfa_results = self._download_webpage(
 162                 tfa_req, None,
 163                 note='Submitting TFA code', errnote='unable to submit tfa', fatal=False)
 164
 165             if tfa_results is False:
 166                 return False
 167
 168             if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', tfa_results) is not None:
 169                 self._downloader.report_warning('Two-factor code expired. Please try again, or use a one-use backup code instead.')
 170                 return False
 171             if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
 172                 self._downloader.report_warning('unable to log in - did the page structure change?')
 173                 return False
 174             if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
 175                 self._downloader.report_warning('Your Google account has a security notice. Please log in on your web browser, resolve the notice, and try again.')
 176                 return False
 177
 178         if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
 179             self._downloader.report_warning('unable to log in: bad username or password')
 180             return False
 181         return True
 182
 183     def _real_initialize(self):
 184         if self._downloader is None:
 185             return
 186         self._set_language()
 187         if not self._login():
 188             return
 189
 190
 191 class YoutubeIE(YoutubeBaseInfoExtractor):
 192     IE_DESC = 'YouTube.com'
 193     _VALID_URL = r"""(?x)^
 194                      (
 195                          (?:https?://|//)                                    # http(s):// or protocol-independent URL
 196                          (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
 197                             (?:www\.)?deturl\.com/www\.youtube\.com/|
 198                             (?:www\.)?pwnyoutube\.com/|
 199                             (?:www\.)?yourepeat\.com/|
 200                             tube\.majestyc\.net/|
 201                             youtube\.googleapis\.com/)                        # the various hostnames, with wildcard subdomains
 202                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
 203                          (?:                                                  # the various things that can precede the ID:
 204                              (?:(?:v|embed|e)/(?!videoseries))                # v/ or embed/ or e/
 205                              |(?:                                             # or the v= param in all its forms
 206                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)?  # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
 207                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
 208                                  (?:.*?&)?                                    # any other preceding param (like /?s=tuff&v=xxxx)
 209                                  v=
 210                              )
 211                          ))
 212                          |youtu\.be/                                          # just youtu.be/xxxx
 213                          |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
 214                          )
 215                      )?                                                       # all until now is optional -> you can pass the naked ID
 216                      ([0-9A-Za-z_-]{11})                                      # here is it! the YouTube video ID
 217                      (?!.*?&list=)                                            # combined list/video URLs are handled by the playlist IE
 218                      (?(1).+)?                                                # if we found the ID, everything can follow
 219                      $"""
 220     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
 221     _formats = {
 222         '5': {'ext': 'flv', 'width': 400, 'height': 240},
 223         '6': {'ext': 'flv', 'width': 450, 'height': 270},
 224         '13': {'ext': '3gp'},
 225         '17': {'ext': '3gp', 'width': 176, 'height': 144},
 226         '18': {'ext': 'mp4', 'width': 640, 'height': 360},
 227         '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
 228         '34': {'ext': 'flv', 'width': 640, 'height': 360},
 229         '35': {'ext': 'flv', 'width': 854, 'height': 480},
 230         '36': {'ext': '3gp', 'width': 320, 'height': 240},
 231         '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
 232         '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
 233         '43': {'ext': 'webm', 'width': 640, 'height': 360},
 234         '44': {'ext': 'webm', 'width': 854, 'height': 480},
 235         '45': {'ext': 'webm', 'width': 1280, 'height': 720},
 236         '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
 237
 238
 239         # 3d videos
 240         '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20},
 241         '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
 242         '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
 243         '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
 244         '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
 245         '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
 246         '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
 247
 248         # Apple HTTP Live Streaming
 249         '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
 250         '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
 251         '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
 252         '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
 253         '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
 254         '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
 255         '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
 256
 257         # DASH mp4 video
 258         '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 259         '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 260         '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 261         '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 262         '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 263         '138': {'ext': 'mp4', 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},  # Height can vary (https://github.com/rg3/youtube-dl/issues/4559)
 264         '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 265         '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 266         '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
 267         '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
 268         '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'h264'},
 269
 270         # Dash mp4 audio
 271         '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 48, 'preference': -50, 'container': 'm4a_dash'},
 272         '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 128, 'preference': -50, 'container': 'm4a_dash'},
 273         '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 256, 'preference': -50, 'container': 'm4a_dash'},
 274
 275         # Dash webm
 276         '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 277         '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 278         '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 279         '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 280         '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 281         '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 282         '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'container': 'webm', 'vcodec': 'VP9'},
 283         '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 284         '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 285         '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 286         '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 287         '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 288         '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 289         '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 290         '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 291         '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 292         '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 293         '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 294         '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 295         '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'VP9'},
 296         '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 297
 298         # Dash webm audio
 299         '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50},
 300         '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
 301
 302         # Dash webm audio with opus inside
 303         '249': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50, 'preference': -50},
 304         '250': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70, 'preference': -50},
 305         '251': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160, 'preference': -50},
 306
 307         # RTMP (unnamed)
 308         '_rtmp': {'protocol': 'rtmp'},
 309     }
 310
 311     IE_NAME = 'youtube'
 312     _TESTS = [
 313         {
 314             'url': 'http://www.youtube.com/watch?v=BaW_jenozKc',
 315             'info_dict': {
 316                 'id': 'BaW_jenozKc',
 317                 'ext': 'mp4',
 318                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 319                 'uploader': 'Philipp Hagemeister',
 320                 'uploader_id': 'phihag',
 321                 'upload_date': '20121002',
 322                 '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 .',
 323                 'categories': ['Science & Technology'],
 324                 'like_count': int,
 325                 'dislike_count': int,
 326             }
 327         },
 328         {
 329             'url': 'http://www.youtube.com/watch?v=UxxajLWwzqY',
 330             'note': 'Test generic use_cipher_signature video (#897)',
 331             'info_dict': {
 332                 'id': 'UxxajLWwzqY',
 333                 'ext': 'mp4',
 334                 'upload_date': '20120506',
 335                 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
 336                 'description': 'md5:fea86fda2d5a5784273df5c7cc994d9f',
 337                 'uploader': 'Icona Pop',
 338                 'uploader_id': 'IconaPop',
 339             }
 340         },
 341         {
 342             'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
 343             'note': 'Test VEVO video with age protection (#956)',
 344             'info_dict': {
 345                 'id': '07FYdnEawAQ',
 346                 'ext': 'mp4',
 347                 'upload_date': '20130703',
 348                 'title': 'Justin Timberlake - Tunnel Vision (Explicit)',
 349                 'description': 'md5:64249768eec3bc4276236606ea996373',
 350                 'uploader': 'justintimberlakeVEVO',
 351                 'uploader_id': 'justintimberlakeVEVO',
 352             }
 353         },
 354         {
 355             'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
 356             'note': 'Embed-only video (#1746)',
 357             'info_dict': {
 358                 'id': 'yZIXLfi8CZQ',
 359                 'ext': 'mp4',
 360                 'upload_date': '20120608',
 361                 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
 362                 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
 363                 'uploader': 'SET India',
 364                 'uploader_id': 'setindia'
 365             }
 366         },
 367         {
 368             'url': 'http://www.youtube.com/watch?v=a9LDPn-MO4I',
 369             'note': '256k DASH audio (format 141) via DASH manifest',
 370             'info_dict': {
 371                 'id': 'a9LDPn-MO4I',
 372                 'ext': 'm4a',
 373                 'upload_date': '20121002',
 374                 'uploader_id': '8KVIDEO',
 375                 'description': '',
 376                 'uploader': '8KVIDEO',
 377                 'title': 'UHDTV TEST 8K VIDEO.mp4'
 378             },
 379             'params': {
 380                 'youtube_include_dash_manifest': True,
 381                 'format': '141',
 382             },
 383         },
 384         # DASH manifest with encrypted signature
 385         {
 386             'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
 387             'info_dict': {
 388                 'id': 'IB3lcPjvWLA',
 389                 'ext': 'm4a',
 390                 'title': 'Afrojack, Spree Wilson - The Spark ft. Spree Wilson',
 391                 'description': 'md5:12e7067fa6735a77bdcbb58cb1187d2d',
 392                 'uploader': 'AfrojackVEVO',
 393                 'uploader_id': 'AfrojackVEVO',
 394                 'upload_date': '20131011',
 395             },
 396             'params': {
 397                 'youtube_include_dash_manifest': True,
 398                 'format': '141',
 399             },
 400         },
 401         # JS player signature function name containing $
 402         {
 403             'url': 'https://www.youtube.com/watch?v=nfWlot6h_JM',
 404             'info_dict': {
 405                 'id': 'nfWlot6h_JM',
 406                 'ext': 'm4a',
 407                 'title': 'Taylor Swift - Shake It Off',
 408                 'description': 'md5:2acfda1b285bdd478ccec22f9918199d',
 409                 'uploader': 'TaylorSwiftVEVO',
 410                 'uploader_id': 'TaylorSwiftVEVO',
 411                 'upload_date': '20140818',
 412             },
 413             'params': {
 414                 'youtube_include_dash_manifest': True,
 415                 'format': '141',
 416             },
 417         },
 418         # Controversy video
 419         {
 420             'url': 'https://www.youtube.com/watch?v=T4XJQO3qol8',
 421             'info_dict': {
 422                 'id': 'T4XJQO3qol8',
 423                 'ext': 'mp4',
 424                 'upload_date': '20100909',
 425                 'uploader': 'The Amazing Atheist',
 426                 'uploader_id': 'TheAmazingAtheist',
 427                 'title': 'Burning Everyone\'s Koran',
 428                 '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',
 429             }
 430         },
 431         # Normal age-gate video (No vevo, embed allowed)
 432         {
 433             'url': 'http://youtube.com/watch?v=HtVdAasjOgU',
 434             'info_dict': {
 435                 'id': 'HtVdAasjOgU',
 436                 'ext': 'mp4',
 437                 'title': 'The Witcher 3: Wild Hunt - The Sword Of Destiny Trailer',
 438                 'description': 're:(?s).{100,}About the Game\n.*?The Witcher 3: Wild Hunt.{100,}',
 439                 'uploader': 'The Witcher',
 440                 'uploader_id': 'WitcherGame',
 441                 'upload_date': '20140605',
 442             },
 443         },
 444         # Age-gate video with encrypted signature
 445         {
 446             'url': 'http://www.youtube.com/watch?v=6kLq3WMV1nU',
 447             'info_dict': {
 448                 'id': '6kLq3WMV1nU',
 449                 'ext': 'mp4',
 450                 'title': 'Dedication To My Ex (Miss That) (Lyric Video)',
 451                 'description': 'md5:33765bb339e1b47e7e72b5490139bb41',
 452                 'uploader': 'LloydVEVO',
 453                 'uploader_id': 'LloydVEVO',
 454                 'upload_date': '20110629',
 455             },
 456         },
 457         # video_info is None (https://github.com/rg3/youtube-dl/issues/4421)
 458         {
 459             'url': '__2ABJjxzNo',
 460             'info_dict': {
 461                 'id': '__2ABJjxzNo',
 462                 'ext': 'mp4',
 463                 'upload_date': '20100430',
 464                 'uploader_id': 'deadmau5',
 465                 'description': 'md5:12c56784b8032162bb936a5f76d55360',
 466                 'uploader': 'deadmau5',
 467                 'title': 'Deadmau5 - Some Chords (HD)',
 468             },
 469             'expected_warnings': [
 470                 'DASH manifest missing',
 471             ]
 472         },
 473         # Olympics (https://github.com/rg3/youtube-dl/issues/4431)
 474         {
 475             'url': 'lqQg6PlCWgI',
 476             'info_dict': {
 477                 'id': 'lqQg6PlCWgI',
 478                 'ext': 'mp4',
 479                 'upload_date': '20120731',
 480                 'uploader_id': 'olympic',
 481                 'description': 'HO09  - Women -  GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games',
 482                 'uploader': 'Olympics',
 483                 'title': 'Hockey - Women -  GER-AUS - London 2012 Olympic Games',
 484             },
 485             'params': {
 486                 'skip_download': 'requires avconv',
 487             }
 488         },
 489         # Non-square pixels
 490         {
 491             'url': 'https://www.youtube.com/watch?v=_b-2C3KPAM0',
 492             'info_dict': {
 493                 'id': '_b-2C3KPAM0',
 494                 'ext': 'mp4',
 495                 'stretched_ratio': 16 / 9.,
 496                 'upload_date': '20110310',
 497                 'uploader_id': 'AllenMeow',
 498                 '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',
 499                 'uploader': '孫艾倫',
 500                 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人',
 501             },
 502         },
 503         # url_encoded_fmt_stream_map is empty string
 504         {
 505             'url': 'qEJwOuvDf7I',
 506             'info_dict': {
 507                 'id': 'qEJwOuvDf7I',
 508                 'ext': 'mp4',
 509                 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге',
 510                 'description': '',
 511                 'upload_date': '20150404',
 512                 'uploader_id': 'spbelect',
 513                 'uploader': 'Наблюдатели Петербурга',
 514             },
 515             'params': {
 516                 'skip_download': 'requires avconv',
 517             }
 518         },
 519     ]
 520
 521     def __init__(self, *args, **kwargs):
 522         super(YoutubeIE, self).__init__(*args, **kwargs)
 523         self._player_cache = {}
 524
 525     def report_video_info_webpage_download(self, video_id):
 526         """Report attempt to download video info webpage."""
 527         self.to_screen('%s: Downloading video info webpage' % video_id)
 528
 529     def report_information_extraction(self, video_id):
 530         """Report attempt to extract video information."""
 531         self.to_screen('%s: Extracting video information' % video_id)
 532
 533     def report_unavailable_format(self, video_id, format):
 534         """Report extracted video URL."""
 535         self.to_screen('%s: Format %s not available' % (video_id, format))
 536
 537     def report_rtmp_download(self):
 538         """Indicate the download will use the RTMP protocol."""
 539         self.to_screen('RTMP download detected')
 540
 541     def _signature_cache_id(self, example_sig):
 542         """ Return a string representation of a signature """
 543         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
 544
 545     def _extract_signature_function(self, video_id, player_url, example_sig):
 546         id_m = re.match(
 547             r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P<ext>[a-z]+)$',
 548             player_url)
 549         if not id_m:
 550             raise ExtractorError('Cannot identify player %r' % player_url)
 551         player_type = id_m.group('ext')
 552         player_id = id_m.group('id')
 553
 554         # Read from filesystem cache
 555         func_id = '%s_%s_%s' % (
 556             player_type, player_id, self._signature_cache_id(example_sig))
 557         assert os.path.basename(func_id) == func_id
 558
 559         cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
 560         if cache_spec is not None:
 561             return lambda s: ''.join(s[i] for i in cache_spec)
 562
 563         download_note = (
 564             'Downloading player %s' % player_url
 565             if self._downloader.params.get('verbose') else
 566             'Downloading %s player %s' % (player_type, player_id)
 567         )
 568         if player_type == 'js':
 569             code = self._download_webpage(
 570                 player_url, video_id,
 571                 note=download_note,
 572                 errnote='Download of %s failed' % player_url)
 573             res = self._parse_sig_js(code)
 574         elif player_type == 'swf':
 575             urlh = self._request_webpage(
 576                 player_url, video_id,
 577                 note=download_note,
 578                 errnote='Download of %s failed' % player_url)
 579             code = urlh.read()
 580             res = self._parse_sig_swf(code)
 581         else:
 582             assert False, 'Invalid player type %r' % player_type
 583
 584         test_string = ''.join(map(compat_chr, range(len(example_sig))))
 585         cache_res = res(test_string)
 586         cache_spec = [ord(c) for c in cache_res]
 587
 588         self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
 589         return res
 590
 591     def _print_sig_code(self, func, example_sig):
 592         def gen_sig_code(idxs):
 593             def _genslice(start, end, step):
 594                 starts = '' if start == 0 else str(start)
 595                 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
 596                 steps = '' if step == 1 else (':%d' % step)
 597                 return 's[%s%s%s]' % (starts, ends, steps)
 598
 599             step = None
 600             # Quelch pyflakes warnings - start will be set when step is set
 601             start = '(Never used)'
 602             for i, prev in zip(idxs[1:], idxs[:-1]):
 603                 if step is not None:
 604                     if i - prev == step:
 605                         continue
 606                     yield _genslice(start, prev, step)
 607                     step = None
 608                     continue
 609                 if i - prev in [-1, 1]:
 610                     step = i - prev
 611                     start = prev
 612                     continue
 613                 else:
 614                     yield 's[%d]' % prev
 615             if step is None:
 616                 yield 's[%d]' % i
 617             else:
 618                 yield _genslice(start, i, step)
 619
 620         test_string = ''.join(map(compat_chr, range(len(example_sig))))
 621         cache_res = func(test_string)
 622         cache_spec = [ord(c) for c in cache_res]
 623         expr_code = ' + '.join(gen_sig_code(cache_spec))
 624         signature_id_tuple = '(%s)' % (
 625             ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
 626         code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
 627                 '    return %s\n') % (signature_id_tuple, expr_code)
 628         self.to_screen('Extracted signature function:\n' + code)
 629
 630     def _parse_sig_js(self, jscode):
 631         funcname = self._search_regex(
 632             r'\.sig\|\|([a-zA-Z0-9$]+)\(', jscode,
 633             'Initial JS player signature function name')
 634
 635         jsi = JSInterpreter(jscode)
 636         initial_function = jsi.extract_function(funcname)
 637         return lambda s: initial_function([s])
 638
 639     def _parse_sig_swf(self, file_contents):
 640         swfi = SWFInterpreter(file_contents)
 641         TARGET_CLASSNAME = 'SignatureDecipher'
 642         searched_class = swfi.extract_class(TARGET_CLASSNAME)
 643         initial_function = swfi.extract_function(searched_class, 'decipher')
 644         return lambda s: initial_function([s])
 645
 646     def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
 647         """Turn the encrypted s field into a working signature"""
 648
 649         if player_url is None:
 650             raise ExtractorError('Cannot decrypt signature without player_url')
 651
 652         if player_url.startswith('//'):
 653             player_url = 'https:' + player_url
 654         try:
 655             player_id = (player_url, self._signature_cache_id(s))
 656             if player_id not in self._player_cache:
 657                 func = self._extract_signature_function(
 658                     video_id, player_url, s
 659                 )
 660                 self._player_cache[player_id] = func
 661             func = self._player_cache[player_id]
 662             if self._downloader.params.get('youtube_print_sig_code'):
 663                 self._print_sig_code(func, s)
 664             return func(s)
 665         except Exception as e:
 666             tb = traceback.format_exc()
 667             raise ExtractorError(
 668                 'Signature extraction failed: ' + tb, cause=e)
 669
 670     def _get_subtitles(self, video_id, webpage):
 671         try:
 672             subs_doc = self._download_xml(
 673                 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
 674                 video_id, note=False)
 675         except ExtractorError as err:
 676             self._downloader.report_warning('unable to download video subtitles: %s' % compat_str(err))
 677             return {}
 678
 679         sub_lang_list = {}
 680         for track in subs_doc.findall('track'):
 681             lang = track.attrib['lang_code']
 682             if lang in sub_lang_list:
 683                 continue
 684             sub_formats = []
 685             for ext in ['sbv', 'vtt', 'srt']:
 686                 params = compat_urllib_parse.urlencode({
 687                     'lang': lang,
 688                     'v': video_id,
 689                     'fmt': ext,
 690                     'name': track.attrib['name'].encode('utf-8'),
 691                 })
 692                 sub_formats.append({
 693                     'url': 'https://www.youtube.com/api/timedtext?' + params,
 694                     'ext': ext,
 695                 })
 696             sub_lang_list[lang] = sub_formats
 697         if not sub_lang_list:
 698             self._downloader.report_warning('video doesn\'t have subtitles')
 699             return {}
 700         return sub_lang_list
 701
 702     def _get_automatic_captions(self, video_id, webpage):
 703         """We need the webpage for getting the captions url, pass it as an
 704            argument to speed up the process."""
 705         self.to_screen('%s: Looking for automatic captions' % video_id)
 706         mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
 707         err_msg = 'Couldn\'t find automatic captions for %s' % video_id
 708         if mobj is None:
 709             self._downloader.report_warning(err_msg)
 710             return {}
 711         player_config = json.loads(mobj.group(1))
 712         try:
 713             args = player_config['args']
 714             caption_url = args['ttsurl']
 715             timestamp = args['timestamp']
 716             # We get the available subtitles
 717             list_params = compat_urllib_parse.urlencode({
 718                 'type': 'list',
 719                 'tlangs': 1,
 720                 'asrs': 1,
 721             })
 722             list_url = caption_url + '&' + list_params
 723             caption_list = self._download_xml(list_url, video_id)
 724             original_lang_node = caption_list.find('track')
 725             if original_lang_node is None:
 726                 self._downloader.report_warning('Video doesn\'t have automatic captions')
 727                 return {}
 728             original_lang = original_lang_node.attrib['lang_code']
 729             caption_kind = original_lang_node.attrib.get('kind', '')
 730
 731             sub_lang_list = {}
 732             for lang_node in caption_list.findall('target'):
 733                 sub_lang = lang_node.attrib['lang_code']
 734                 sub_formats = []
 735                 for ext in ['sbv', 'vtt', 'srt']:
 736                     params = compat_urllib_parse.urlencode({
 737                         'lang': original_lang,
 738                         'tlang': sub_lang,
 739                         'fmt': ext,
 740                         'ts': timestamp,
 741                         'kind': caption_kind,
 742                     })
 743                     sub_formats.append({
 744                         'url': caption_url + '&' + params,
 745                         'ext': ext,
 746                     })
 747                 sub_lang_list[sub_lang] = sub_formats
 748             return sub_lang_list
 749         # An extractor error can be raise by the download process if there are
 750         # no automatic captions but there are subtitles
 751         except (KeyError, ExtractorError):
 752             self._downloader.report_warning(err_msg)
 753             return {}
 754
 755     @classmethod
 756     def extract_id(cls, url):
 757         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
 758         if mobj is None:
 759             raise ExtractorError('Invalid URL: %s' % url)
 760         video_id = mobj.group(2)
 761         return video_id
 762
 763     def _extract_from_m3u8(self, manifest_url, video_id):
 764         url_map = {}
 765
 766         def _get_urls(_manifest):
 767             lines = _manifest.split('\n')
 768             urls = filter(lambda l: l and not l.startswith('#'),
 769                           lines)
 770             return urls
 771         manifest = self._download_webpage(manifest_url, video_id, 'Downloading formats manifest')
 772         formats_urls = _get_urls(manifest)
 773         for format_url in formats_urls:
 774             itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
 775             url_map[itag] = format_url
 776         return url_map
 777
 778     def _extract_annotations(self, video_id):
 779         url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
 780         return self._download_webpage(url, video_id, note='Searching for annotations.', errnote='Unable to download video annotations.')
 781
 782     def _parse_dash_manifest(
 783             self, video_id, dash_manifest_url, player_url, age_gate):
 784         def decrypt_sig(mobj):
 785             s = mobj.group(1)
 786             dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
 787             return '/signature/%s' % dec_s
 788         dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
 789         dash_doc = self._download_xml(
 790             dash_manifest_url, video_id,
 791             note='Downloading DASH manifest',
 792             errnote='Could not download DASH manifest')
 793
 794         formats = []
 795         for a in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}AdaptationSet'):
 796             mime_type = a.attrib.get('mimeType')
 797             for r in a.findall('{urn:mpeg:DASH:schema:MPD:2011}Representation'):
 798                 url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
 799                 if url_el is None:
 800                     continue
 801                 if mime_type == 'text/vtt':
 802                     # TODO implement WebVTT downloading
 803                     pass
 804                 elif mime_type.startswith('audio/') or mime_type.startswith('video/'):
 805                     segment_list = r.find('{urn:mpeg:DASH:schema:MPD:2011}SegmentList')
 806                     format_id = r.attrib['id']
 807                     video_url = url_el.text
 808                     filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
 809                     f = {
 810                         'format_id': format_id,
 811                         'url': video_url,
 812                         'width': int_or_none(r.attrib.get('width')),
 813                         'height': int_or_none(r.attrib.get('height')),
 814                         'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
 815                         'asr': int_or_none(r.attrib.get('audioSamplingRate')),
 816                         'filesize': filesize,
 817                         'fps': int_or_none(r.attrib.get('frameRate')),
 818                     }
 819                     if segment_list is not None:
 820                         f.update({
 821                             'initialization_url': segment_list.find('{urn:mpeg:DASH:schema:MPD:2011}Initialization').attrib['sourceURL'],
 822                             'segment_urls': [segment.attrib.get('media') for segment in segment_list.findall('{urn:mpeg:DASH:schema:MPD:2011}SegmentURL')],
 823                             'protocol': 'http_dash_segments',
 824                         })
 825                     try:
 826                         existing_format = next(
 827                             fo for fo in formats
 828                             if fo['format_id'] == format_id)
 829                     except StopIteration:
 830                         full_info = self._formats.get(format_id, {}).copy()
 831                         full_info.update(f)
 832                         formats.append(full_info)
 833                     else:
 834                         existing_format.update(f)
 835                 else:
 836                     self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
 837         return formats
 838
 839     def _real_extract(self, url):
 840         proto = (
 841             'http' if self._downloader.params.get('prefer_insecure', False)
 842             else 'https')
 843
 844         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
 845         mobj = re.search(self._NEXT_URL_RE, url)
 846         if mobj:
 847             url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
 848         video_id = self.extract_id(url)
 849
 850         # Get video webpage
 851         url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
 852         video_webpage = self._download_webpage(url, video_id)
 853
 854         # Attempt to extract SWF player URL
 855         mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
 856         if mobj is not None:
 857             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
 858         else:
 859             player_url = None
 860
 861         # Get video info
 862         embed_webpage = None
 863         if re.search(r'player-age-gate-content">', video_webpage) is not None:
 864             age_gate = True
 865             # We simulate the access to the video from www.youtube.com/v/{video_id}
 866             # this can be viewed without login into Youtube
 867             url = proto + '://www.youtube.com/embed/%s' % video_id
 868             embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage')
 869             data = compat_urllib_parse.urlencode({
 870                 'video_id': video_id,
 871                 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
 872                 'sts': self._search_regex(
 873                     r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
 874             })
 875             video_info_url = proto + '://www.youtube.com/get_video_info?' + data
 876             video_info_webpage = self._download_webpage(
 877                 video_info_url, video_id,
 878                 note='Refetching age-gated info webpage',
 879                 errnote='unable to download video info webpage')
 880             video_info = compat_parse_qs(video_info_webpage)
 881         else:
 882             age_gate = False
 883             try:
 884                 # Try looking directly into the video webpage
 885                 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
 886                 if not mobj:
 887                     raise ValueError('Could not find ytplayer.config')  # caught below
 888                 json_code = uppercase_escape(mobj.group(1))
 889                 ytplayer_config = json.loads(json_code)
 890                 args = ytplayer_config['args']
 891                 # Convert to the same format returned by compat_parse_qs
 892                 video_info = dict((k, [v]) for k, v in args.items())
 893                 if not args.get('url_encoded_fmt_stream_map'):
 894                     raise ValueError('No stream_map present')  # caught below
 895             except ValueError:
 896                 # We fallback to the get_video_info pages (used by the embed page)
 897                 self.report_video_info_webpage_download(video_id)
 898                 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
 899                     video_info_url = (
 900                         '%s://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
 901                         % (proto, video_id, el_type))
 902                     video_info_webpage = self._download_webpage(
 903                         video_info_url,
 904                         video_id, note=False,
 905                         errnote='unable to download video info webpage')
 906                     video_info = compat_parse_qs(video_info_webpage)
 907                     if 'token' in video_info:
 908                         break
 909         if 'token' not in video_info:
 910             if 'reason' in video_info:
 911                 raise ExtractorError(
 912                     'YouTube said: %s' % video_info['reason'][0],
 913                     expected=True, video_id=video_id)
 914             else:
 915                 raise ExtractorError(
 916                     '"token" parameter not in video info for unknown reason',
 917                     video_id=video_id)
 918
 919         if 'view_count' in video_info:
 920             view_count = int(video_info['view_count'][0])
 921         else:
 922             view_count = None
 923
 924         # Check for "rental" videos
 925         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 926             raise ExtractorError('"rental" videos not supported')
 927
 928         # Start extracting information
 929         self.report_information_extraction(video_id)
 930
 931         # uploader
 932         if 'author' not in video_info:
 933             raise ExtractorError('Unable to extract uploader name')
 934         video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
 935
 936         # uploader_id
 937         video_uploader_id = None
 938         mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
 939         if mobj is not None:
 940             video_uploader_id = mobj.group(1)
 941         else:
 942             self._downloader.report_warning('unable to extract uploader nickname')
 943
 944         # title
 945         if 'title' in video_info:
 946             video_title = video_info['title'][0]
 947         else:
 948             self._downloader.report_warning('Unable to extract video title')
 949             video_title = '_'
 950
 951         # thumbnail image
 952         # We try first to get a high quality image:
 953         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
 954                             video_webpage, re.DOTALL)
 955         if m_thumb is not None:
 956             video_thumbnail = m_thumb.group(1)
 957         elif 'thumbnail_url' not in video_info:
 958             self._downloader.report_warning('unable to extract video thumbnail')
 959             video_thumbnail = None
 960         else:   # don't panic if we can't find it
 961             video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
 962
 963         # upload date
 964         upload_date = None
 965         mobj = re.search(r'(?s)id="eow-date.*?>(.*?)</span>', video_webpage)
 966         if mobj is None:
 967             mobj = re.search(
 968                 r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
 969                 video_webpage)
 970         if mobj is not None:
 971             upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
 972             upload_date = unified_strdate(upload_date)
 973
 974         m_cat_container = self._search_regex(
 975             r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
 976             video_webpage, 'categories', default=None)
 977         if m_cat_container:
 978             category = self._html_search_regex(
 979                 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
 980                 default=None)
 981             video_categories = None if category is None else [category]
 982         else:
 983             video_categories = None
 984
 985         # description
 986         video_description = get_element_by_id("eow-description", video_webpage)
 987         if video_description:
 988             video_description = re.sub(r'''(?x)
 989                 <a\s+
 990                     (?:[a-zA-Z-]+="[^"]+"\s+)*?
 991                     title="([^"]+)"\s+
 992                     (?:[a-zA-Z-]+="[^"]+"\s+)*?
 993                     class="yt-uix-redirect-link"\s*>
 994                 [^<]+
 995                 </a>
 996             ''', r'\1', video_description)
 997             video_description = clean_html(video_description)
 998         else:
 999             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1000             if fd_mobj:
1001                 video_description = unescapeHTML(fd_mobj.group(1))
1002             else:
1003                 video_description = ''
1004
1005         def _extract_count(count_name):
1006             count = self._search_regex(
1007                 r'id="watch-%s"[^>]*>.*?([\d,]+)\s*</span>' % re.escape(count_name),
1008                 video_webpage, count_name, default=None)
1009             if count is not None:
1010                 return int(count.replace(',', ''))
1011             return None
1012         like_count = _extract_count('like')
1013         dislike_count = _extract_count('dislike')
1014
1015         # subtitles
1016         video_subtitles = self.extract_subtitles(video_id, video_webpage)
1017         automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
1018
1019         if 'length_seconds' not in video_info:
1020             self._downloader.report_warning('unable to extract video duration')
1021             video_duration = None
1022         else:
1023             video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1024
1025         # annotations
1026         video_annotations = None
1027         if self._downloader.params.get('writeannotations', False):
1028             video_annotations = self._extract_annotations(video_id)
1029
1030         def _map_to_format_list(urlmap):
1031             formats = []
1032             for itag, video_real_url in urlmap.items():
1033                 dct = {
1034                     'format_id': itag,
1035                     'url': video_real_url,
1036                     'player_url': player_url,
1037                 }
1038                 if itag in self._formats:
1039                     dct.update(self._formats[itag])
1040                 formats.append(dct)
1041             return formats
1042
1043         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1044             self.report_rtmp_download()
1045             formats = [{
1046                 'format_id': '_rtmp',
1047                 'protocol': 'rtmp',
1048                 'url': video_info['conn'][0],
1049                 'player_url': player_url,
1050             }]
1051         elif len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1:
1052             encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
1053             if 'rtmpe%3Dyes' in encoded_url_map:
1054                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1055             url_map = {}
1056             for url_data_str in encoded_url_map.split(','):
1057                 url_data = compat_parse_qs(url_data_str)
1058                 if 'itag' not in url_data or 'url' not in url_data:
1059                     continue
1060                 format_id = url_data['itag'][0]
1061                 url = url_data['url'][0]
1062
1063                 if 'sig' in url_data:
1064                     url += '&signature=' + url_data['sig'][0]
1065                 elif 's' in url_data:
1066                     encrypted_sig = url_data['s'][0]
1067                     ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
1068
1069                     jsplayer_url_json = self._search_regex(
1070                         ASSETS_RE,
1071                         embed_webpage if age_gate else video_webpage,
1072                         'JS player URL (1)', default=None)
1073                     if not jsplayer_url_json and not age_gate:
1074                         # We need the embed website after all
1075                         if embed_webpage is None:
1076                             embed_url = proto + '://www.youtube.com/embed/%s' % video_id
1077                             embed_webpage = self._download_webpage(
1078                                 embed_url, video_id, 'Downloading embed webpage')
1079                         jsplayer_url_json = self._search_regex(
1080                             ASSETS_RE, embed_webpage, 'JS player URL')
1081
1082                     player_url = json.loads(jsplayer_url_json)
1083                     if player_url is None:
1084                         player_url_json = self._search_regex(
1085                             r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
1086                             video_webpage, 'age gate player URL')
1087                         player_url = json.loads(player_url_json)
1088
1089                     if self._downloader.params.get('verbose'):
1090                         if player_url is None:
1091                             player_version = 'unknown'
1092                             player_desc = 'unknown'
1093                         else:
1094                             if player_url.endswith('swf'):
1095                                 player_version = self._search_regex(
1096                                     r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
1097                                     'flash player', fatal=False)
1098                                 player_desc = 'flash player %s' % player_version
1099                             else:
1100                                 player_version = self._search_regex(
1101                                     r'html5player-([^/]+?)(?:/html5player)?\.js',
1102                                     player_url,
1103                                     'html5 player', fatal=False)
1104                                 player_desc = 'html5 player %s' % player_version
1105
1106                         parts_sizes = self._signature_cache_id(encrypted_sig)
1107                         self.to_screen('{%s} signature length %s, %s' %
1108                                        (format_id, parts_sizes, player_desc))
1109
1110                     signature = self._decrypt_signature(
1111                         encrypted_sig, video_id, player_url, age_gate)
1112                     url += '&signature=' + signature
1113                 if 'ratebypass' not in url:
1114                     url += '&ratebypass=yes'
1115                 url_map[format_id] = url
1116             formats = _map_to_format_list(url_map)
1117         elif video_info.get('hlsvp'):
1118             manifest_url = video_info['hlsvp'][0]
1119             url_map = self._extract_from_m3u8(manifest_url, video_id)
1120             formats = _map_to_format_list(url_map)
1121         else:
1122             raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1123
1124         # Look for the DASH manifest
1125         if self._downloader.params.get('youtube_include_dash_manifest', True):
1126             dash_mpd = video_info.get('dashmpd')
1127             if dash_mpd:
1128                 dash_manifest_url = dash_mpd[0]
1129                 try:
1130                     dash_formats = self._parse_dash_manifest(
1131                         video_id, dash_manifest_url, player_url, age_gate)
1132                 except (ExtractorError, KeyError) as e:
1133                     self.report_warning(
1134                         'Skipping DASH manifest: %r' % e, video_id)
1135                 else:
1136                     # Remove the formats we found through non-DASH, they
1137                     # contain less info and it can be wrong, because we use
1138                     # fixed values (for example the resolution). See
1139                     # https://github.com/rg3/youtube-dl/issues/5774 for an
1140                     # example.
1141                     dash_keys = set(df['format_id'] for df in dash_formats)
1142                     formats = [f for f in formats if f['format_id'] not in dash_keys]
1143                     formats.extend(dash_formats)
1144
1145         # Check for malformed aspect ratio
1146         stretched_m = re.search(
1147             r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">',
1148             video_webpage)
1149         if stretched_m:
1150             ratio = float(stretched_m.group('w')) / float(stretched_m.group('h'))
1151             for f in formats:
1152                 if f.get('vcodec') != 'none':
1153                     f['stretched_ratio'] = ratio
1154
1155         self._sort_formats(formats)
1156
1157         return {
1158             'id': video_id,
1159             'uploader': video_uploader,
1160             'uploader_id': video_uploader_id,
1161             'upload_date': upload_date,
1162             'title': video_title,
1163             'thumbnail': video_thumbnail,
1164             'description': video_description,
1165             'categories': video_categories,
1166             'subtitles': video_subtitles,
1167             'automatic_captions': automatic_captions,
1168             'duration': video_duration,
1169             'age_limit': 18 if age_gate else 0,
1170             'annotations': video_annotations,
1171             'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1172             'view_count': view_count,
1173             'like_count': like_count,
1174             'dislike_count': dislike_count,
1175             'average_rating': float_or_none(video_info.get('avg_rating', [None])[0]),
1176             'formats': formats,
1177         }
1178
1179
1180 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1181     IE_DESC = 'YouTube.com playlists'
1182     _VALID_URL = r"""(?x)(?:
1183                         (?:https?://)?
1184                         (?:\w+\.)?
1185                         youtube\.com/
1186                         (?:
1187                            (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/videoseries)
1188                            \? (?:.*?&)*? (?:p|a|list)=
1189                         |  p/
1190                         )
1191                         (
1192                             (?:PL|LL|EC|UU|FL|RD|UL)?[0-9A-Za-z-_]{10,}
1193                             # Top tracks, they can also include dots
1194                             |(?:MC)[\w\.]*
1195                         )
1196                         .*
1197                      |
1198                         ((?:PL|LL|EC|UU|FL|RD|UL)[0-9A-Za-z-_]{10,})
1199                      )"""
1200     _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1201     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)'
1202     IE_NAME = 'youtube:playlist'
1203     _TESTS = [{
1204         'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1205         'info_dict': {
1206             'title': 'ytdl test PL',
1207             'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1208         },
1209         'playlist_count': 3,
1210     }, {
1211         'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1212         'info_dict': {
1213             'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1214             'title': 'YDL_Empty_List',
1215         },
1216         'playlist_count': 0,
1217     }, {
1218         'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
1219         'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1220         'info_dict': {
1221             'title': '29C3: Not my department',
1222             'id': 'PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1223         },
1224         'playlist_count': 95,
1225     }, {
1226         'note': 'issue #673',
1227         'url': 'PLBB231211A4F62143',
1228         'info_dict': {
1229             'title': '[OLD]Team Fortress 2 (Class-based LP)',
1230             'id': 'PLBB231211A4F62143',
1231         },
1232         'playlist_mincount': 26,
1233     }, {
1234         'note': 'Large playlist',
1235         'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
1236         'info_dict': {
1237             'title': 'Uploads from Cauchemar',
1238             'id': 'UUBABnxM4Ar9ten8Mdjj1j0Q',
1239         },
1240         'playlist_mincount': 799,
1241     }, {
1242         'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1243         'info_dict': {
1244             'title': 'YDL_safe_search',
1245             'id': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1246         },
1247         'playlist_count': 2,
1248     }, {
1249         'note': 'embedded',
1250         'url': 'http://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1251         'playlist_count': 4,
1252         'info_dict': {
1253             'title': 'JODA15',
1254             'id': 'PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1255         }
1256     }, {
1257         'note': 'Embedded SWF player',
1258         'url': 'http://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
1259         'playlist_count': 4,
1260         'info_dict': {
1261             'title': 'JODA7',
1262             'id': 'YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ',
1263         }
1264     }, {
1265         'note': 'Buggy playlist: the webpage has a "Load more" button but it doesn\'t have more videos',
1266         'url': 'https://www.youtube.com/playlist?list=UUXw-G3eDE9trcvY2sBMM_aA',
1267         'info_dict': {
1268             'title': 'Uploads from Interstellar Movie',
1269             'id': 'UUXw-G3eDE9trcvY2sBMM_aA',
1270         },
1271         'playlist_mincout': 21,
1272     }]
1273
1274     def _real_initialize(self):
1275         self._login()
1276
1277     def _extract_mix(self, playlist_id):
1278         # The mixes are generated from a single video
1279         # the id of the playlist is just 'RD' + video_id
1280         url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1281         webpage = self._download_webpage(
1282             url, playlist_id, 'Downloading Youtube mix')
1283         search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1284         title_span = (
1285             search_title('playlist-title') or
1286             search_title('title long-title') or
1287             search_title('title'))
1288         title = clean_html(title_span)
1289         ids = orderedSet(re.findall(
1290             r'''(?xs)data-video-username=".*?".*?
1291                        href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
1292             webpage))
1293         url_results = self._ids_to_results(ids)
1294
1295         return self.playlist_result(url_results, playlist_id, title)
1296
1297     def _extract_playlist(self, playlist_id):
1298         url = self._TEMPLATE_URL % playlist_id
1299         page = self._download_webpage(url, playlist_id)
1300         more_widget_html = content_html = page
1301
1302         for match in re.findall(r'<div class="yt-alert-message">([^<]+)</div>', page):
1303             match = match.strip()
1304             # Check if the playlist exists or is private
1305             if re.match(r'[^<]*(The|This) playlist (does not exist|is private)[^<]*', match):
1306                 raise ExtractorError(
1307                     'The playlist doesn\'t exist or is private, use --username or '
1308                     '--netrc to access it.',
1309                     expected=True)
1310             elif re.match(r'[^<]*Invalid parameters[^<]*', match):
1311                 raise ExtractorError(
1312                     'Invalid parameters. Maybe URL is incorrect.',
1313                     expected=True)
1314             elif re.match(r'[^<]*Choose your language[^<]*', match):
1315                 continue
1316             else:
1317                 self.report_warning('Youtube gives an alert message: ' + match)
1318
1319         # Extract the video ids from the playlist pages
1320         ids = []
1321
1322         for page_num in itertools.count(1):
1323             matches = re.finditer(self._VIDEO_RE, content_html)
1324             # We remove the duplicates and the link with index 0
1325             # (it's not the first video of the playlist)
1326             new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1327             ids.extend(new_ids)
1328
1329             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1330             if not mobj:
1331                 break
1332
1333             more = self._download_json(
1334                 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1335                 'Downloading page #%s' % page_num,
1336                 transform_source=uppercase_escape)
1337             content_html = more['content_html']
1338             if not content_html.strip():
1339                 # Some webpages show a "Load more" button but they don't
1340                 # have more videos
1341                 break
1342             more_widget_html = more['load_more_widget_html']
1343
1344         playlist_title = self._html_search_regex(
1345             r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1346             page, 'title')
1347
1348         url_results = self._ids_to_results(ids)
1349         return self.playlist_result(url_results, playlist_id, playlist_title)
1350
1351     def _real_extract(self, url):
1352         # Extract playlist id
1353         mobj = re.match(self._VALID_URL, url)
1354         if mobj is None:
1355             raise ExtractorError('Invalid URL: %s' % url)
1356         playlist_id = mobj.group(1) or mobj.group(2)
1357
1358         # Check if it's a video-specific URL
1359         query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1360         if 'v' in query_dict:
1361             video_id = query_dict['v'][0]
1362             if self._downloader.params.get('noplaylist'):
1363                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1364                 return self.url_result(video_id, 'Youtube', video_id=video_id)
1365             else:
1366                 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1367
1368         if playlist_id.startswith('RD') or playlist_id.startswith('UL'):
1369             # Mixes require a custom extraction process
1370             return self._extract_mix(playlist_id)
1371
1372         return self._extract_playlist(playlist_id)
1373
1374
1375 class YoutubeChannelIE(InfoExtractor):
1376     IE_DESC = 'YouTube.com channels'
1377     _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/(?P<id>[0-9A-Za-z_-]+)'
1378     _TEMPLATE_URL = 'https://www.youtube.com/channel/%s/videos'
1379     IE_NAME = 'youtube:channel'
1380     _TESTS = [{
1381         'note': 'paginated channel',
1382         'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
1383         'playlist_mincount': 91,
1384         'info_dict': {
1385             'id': 'UCKfVa3S1e4PHvxWcwyMMg8w',
1386         }
1387     }]
1388
1389     @staticmethod
1390     def extract_videos_from_page(page):
1391         ids_in_page = []
1392         titles_in_page = []
1393         for mobj in re.finditer(r'(?:title="(?P<title>[^"]+)"[^>]+)?href="/watch\?v=(?P<id>[0-9A-Za-z_-]+)&?', page):
1394             video_id = mobj.group('id')
1395             video_title = unescapeHTML(mobj.group('title'))
1396             try:
1397                 idx = ids_in_page.index(video_id)
1398                 if video_title and not titles_in_page[idx]:
1399                     titles_in_page[idx] = video_title
1400             except ValueError:
1401                 ids_in_page.append(video_id)
1402                 titles_in_page.append(video_title)
1403         return zip(ids_in_page, titles_in_page)
1404
1405     def _real_extract(self, url):
1406         channel_id = self._match_id(url)
1407
1408         url = self._TEMPLATE_URL % channel_id
1409
1410         # Channel by page listing is restricted to 35 pages of 30 items, i.e. 1050 videos total (see #5778)
1411         # Workaround by extracting as a playlist if managed to obtain channel playlist URL
1412         # otherwise fallback on channel by page extraction
1413         channel_page = self._download_webpage(
1414             url + '?view=57', channel_id,
1415             'Downloading channel page', fatal=False)
1416         channel_playlist_id = self._search_regex(
1417             [r'<meta itemprop="channelId" content="([^"]+)">',
1418              r'data-channel-external-id="([^"]+)"'],
1419             channel_page, 'channel id', default=None)
1420         if channel_playlist_id and channel_playlist_id.startswith('UC'):
1421             playlist_id = 'UU' + channel_playlist_id[2:]
1422             return self.url_result(
1423                 compat_urlparse.urljoin(url, '/playlist?list=%s' % playlist_id), 'YoutubePlaylist')
1424
1425         channel_page = self._download_webpage(url, channel_id, 'Downloading page #1')
1426         autogenerated = re.search(r'''(?x)
1427                 class="[^"]*?(?:
1428                     channel-header-autogenerated-label|
1429                     yt-channel-title-autogenerated
1430                 )[^"]*"''', channel_page) is not None
1431
1432         if autogenerated:
1433             # The videos are contained in a single page
1434             # the ajax pages can't be used, they are empty
1435             entries = [
1436                 self.url_result(
1437                     video_id, 'Youtube', video_id=video_id,
1438                     video_title=video_title)
1439                 for video_id, video_title in self.extract_videos_from_page(channel_page)]
1440             return self.playlist_result(entries, channel_id)
1441
1442         def _entries():
1443             more_widget_html = content_html = channel_page
1444             for pagenum in itertools.count(1):
1445
1446                 for video_id, video_title in self.extract_videos_from_page(content_html):
1447                     yield self.url_result(
1448                         video_id, 'Youtube', video_id=video_id,
1449                         video_title=video_title)
1450
1451                 mobj = re.search(
1452                     r'data-uix-load-more-href="/?(?P<more>[^"]+)"',
1453                     more_widget_html)
1454                 if not mobj:
1455                     break
1456
1457                 more = self._download_json(
1458                     'https://youtube.com/%s' % mobj.group('more'), channel_id,
1459                     'Downloading page #%s' % (pagenum + 1),
1460                     transform_source=uppercase_escape)
1461                 content_html = more['content_html']
1462                 more_widget_html = more['load_more_widget_html']
1463
1464         return self.playlist_result(_entries(), channel_id)
1465
1466
1467 class YoutubeUserIE(YoutubeChannelIE):
1468     IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
1469     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)'
1470     _TEMPLATE_URL = 'https://www.youtube.com/user/%s/videos'
1471     IE_NAME = 'youtube:user'
1472
1473     _TESTS = [{
1474         'url': 'https://www.youtube.com/user/TheLinuxFoundation',
1475         'playlist_mincount': 320,
1476         'info_dict': {
1477             'title': 'TheLinuxFoundation',
1478         }
1479     }, {
1480         'url': 'ytuser:phihag',
1481         'only_matching': True,
1482     }]
1483
1484     @classmethod
1485     def suitable(cls, url):
1486         # Don't return True if the url can be extracted with other youtube
1487         # extractor, the regex would is too permissive and it would match.
1488         other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1489         if any(ie.suitable(url) for ie in other_ies):
1490             return False
1491         else:
1492             return super(YoutubeUserIE, cls).suitable(url)
1493
1494
1495 class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE):
1496     IE_DESC = 'YouTube.com searches'
1497     # there doesn't appear to be a real limit, for example if you search for
1498     # 'python' you get more than 8.000.000 results
1499     _MAX_RESULTS = float('inf')
1500     IE_NAME = 'youtube:search'
1501     _SEARCH_KEY = 'ytsearch'
1502     _EXTRA_QUERY_ARGS = {}
1503     _TESTS = []
1504
1505     def _get_n_results(self, query, n):
1506         """Get a specified number of results for a query"""
1507
1508         videos = []
1509         limit = n
1510
1511         for pagenum in itertools.count(1):
1512             url_query = {
1513                 'search_query': query,
1514                 'page': pagenum,
1515                 'spf': 'navigate',
1516             }
1517             url_query.update(self._EXTRA_QUERY_ARGS)
1518             result_url = 'https://www.youtube.com/results?' + compat_urllib_parse.urlencode(url_query)
1519             data = self._download_json(
1520                 result_url, video_id='query "%s"' % query,
1521                 note='Downloading page %s' % pagenum,
1522                 errnote='Unable to download API page')
1523             html_content = data[1]['body']['content']
1524
1525             if 'class="search-message' in html_content:
1526                 raise ExtractorError(
1527                     '[youtube] No video results', expected=True)
1528
1529             new_videos = self._ids_to_results(orderedSet(re.findall(
1530                 r'href="/watch\?v=(.{11})', html_content)))
1531             videos += new_videos
1532             if not new_videos or len(videos) > limit:
1533                 break
1534
1535         if len(videos) > n:
1536             videos = videos[:n]
1537         return self.playlist_result(videos, query)
1538
1539
1540 class YoutubeSearchDateIE(YoutubeSearchIE):
1541     IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1542     _SEARCH_KEY = 'ytsearchdate'
1543     IE_DESC = 'YouTube.com searches, newest videos first'
1544     _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
1545
1546
1547 class YoutubeSearchURLIE(InfoExtractor):
1548     IE_DESC = 'YouTube.com search URLs'
1549     IE_NAME = 'youtube:search_url'
1550     _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1551     _TESTS = [{
1552         'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
1553         'playlist_mincount': 5,
1554         'info_dict': {
1555             'title': 'youtube-dl test video',
1556         }
1557     }]
1558
1559     def _real_extract(self, url):
1560         mobj = re.match(self._VALID_URL, url)
1561         query = compat_urllib_parse.unquote_plus(mobj.group('query'))
1562
1563         webpage = self._download_webpage(url, query)
1564         result_code = self._search_regex(
1565             r'(?s)<ol[^>]+class="item-section"(.*?)</ol>', webpage, 'result HTML')
1566
1567         part_codes = re.findall(
1568             r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1569         entries = []
1570         for part_code in part_codes:
1571             part_title = self._html_search_regex(
1572                 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1573             part_url_snippet = self._html_search_regex(
1574                 r'(?s)href="([^"]+)"', part_code, 'item URL')
1575             part_url = compat_urlparse.urljoin(
1576                 'https://www.youtube.com/', part_url_snippet)
1577             entries.append({
1578                 '_type': 'url',
1579                 'url': part_url,
1580                 'title': part_title,
1581             })
1582
1583         return {
1584             '_type': 'playlist',
1585             'entries': entries,
1586             'title': query,
1587         }
1588
1589
1590 class YoutubeShowIE(InfoExtractor):
1591     IE_DESC = 'YouTube.com (multi-season) shows'
1592     _VALID_URL = r'https?://www\.youtube\.com/show/(?P<id>[^?#]*)'
1593     IE_NAME = 'youtube:show'
1594     _TESTS = [{
1595         'url': 'http://www.youtube.com/show/airdisasters',
1596         'playlist_mincount': 3,
1597         'info_dict': {
1598             'id': 'airdisasters',
1599             'title': 'Air Disasters',
1600         }
1601     }]
1602
1603     def _real_extract(self, url):
1604         mobj = re.match(self._VALID_URL, url)
1605         playlist_id = mobj.group('id')
1606         webpage = self._download_webpage(
1607             url, playlist_id, 'Downloading show webpage')
1608         # There's one playlist for each season of the show
1609         m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1610         self.to_screen('%s: Found %s seasons' % (playlist_id, len(m_seasons)))
1611         entries = [
1612             self.url_result(
1613                 'https://www.youtube.com' + season.group(1), 'YoutubePlaylist')
1614             for season in m_seasons
1615         ]
1616         title = self._og_search_title(webpage, fatal=False)
1617
1618         return {
1619             '_type': 'playlist',
1620             'id': playlist_id,
1621             'title': title,
1622             'entries': entries,
1623         }
1624
1625
1626 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1627     """
1628     Base class for feed extractors
1629     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1630     """
1631     _LOGIN_REQUIRED = True
1632
1633     @property
1634     def IE_NAME(self):
1635         return 'youtube:%s' % self._FEED_NAME
1636
1637     def _real_initialize(self):
1638         self._login()
1639
1640     def _real_extract(self, url):
1641         page = self._download_webpage(
1642             'https://www.youtube.com/feed/%s' % self._FEED_NAME, self._PLAYLIST_TITLE)
1643
1644         # The extraction process is the same as for playlists, but the regex
1645         # for the video ids doesn't contain an index
1646         ids = []
1647         more_widget_html = content_html = page
1648         for page_num in itertools.count(1):
1649             matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
1650
1651             # 'recommended' feed has infinite 'load more' and each new portion spins
1652             # the same videos in (sometimes) slightly different order, so we'll check
1653             # for unicity and break when portion has no new videos
1654             new_ids = filter(lambda video_id: video_id not in ids, orderedSet(matches))
1655             if not new_ids:
1656                 break
1657
1658             ids.extend(new_ids)
1659
1660             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1661             if not mobj:
1662                 break
1663
1664             more = self._download_json(
1665                 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE,
1666                 'Downloading page #%s' % page_num,
1667                 transform_source=uppercase_escape)
1668             content_html = more['content_html']
1669             more_widget_html = more['load_more_widget_html']
1670
1671         return self.playlist_result(
1672             self._ids_to_results(ids), playlist_title=self._PLAYLIST_TITLE)
1673
1674
1675 class YoutubeWatchLaterIE(YoutubePlaylistIE):
1676     IE_NAME = 'youtube:watchlater'
1677     IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
1678     _VALID_URL = r'https?://www\.youtube\.com/(?:feed/watch_later|playlist\?list=WL)|:ytwatchlater'
1679
1680     _TESTS = []  # override PlaylistIE tests
1681
1682     def _real_extract(self, url):
1683         return self._extract_playlist('WL')
1684
1685
1686 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1687     IE_NAME = 'youtube:favorites'
1688     IE_DESC = 'YouTube.com favourite videos, ":ytfav" for short (requires authentication)'
1689     _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1690     _LOGIN_REQUIRED = True
1691
1692     def _real_extract(self, url):
1693         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1694         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
1695         return self.url_result(playlist_id, 'YoutubePlaylist')
1696
1697
1698 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1699     IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
1700     _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1701     _FEED_NAME = 'recommended'
1702     _PLAYLIST_TITLE = 'Youtube Recommended videos'
1703
1704
1705 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1706     IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1707     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1708     _FEED_NAME = 'subscriptions'
1709     _PLAYLIST_TITLE = 'Youtube Subscriptions'
1710
1711
1712 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1713     IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
1714     _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory'
1715     _FEED_NAME = 'history'
1716     _PLAYLIST_TITLE = 'Youtube History'
1717
1718
1719 class YoutubeTruncatedURLIE(InfoExtractor):
1720     IE_NAME = 'youtube:truncated_url'
1721     IE_DESC = False  # Do not list
1722     _VALID_URL = r'''(?x)
1723         (?:https?://)?
1724         (?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/
1725         (?:watch\?(?:
1726             feature=[a-z_]+|
1727             annotation_id=annotation_[^&]+|
1728             x-yt-cl=[0-9]+|
1729             hl=[^&]*|
1730         )?
1731         |
1732             attribution_link\?a=[^&]+
1733         )
1734         $
1735     '''
1736
1737     _TESTS = [{
1738         'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1739         'only_matching': True,
1740     }, {
1741         'url': 'http://www.youtube.com/watch?',
1742         'only_matching': True,
1743     }, {
1744         'url': 'https://www.youtube.com/watch?x-yt-cl=84503534',
1745         'only_matching': True,
1746     }, {
1747         'url': 'https://www.youtube.com/watch?feature=foo',
1748         'only_matching': True,
1749     }, {
1750         'url': 'https://www.youtube.com/watch?hl=en-GB',
1751         'only_matching': True,
1752     }]
1753
1754     def _real_extract(self, url):
1755         raise ExtractorError(
1756             'Did you forget to quote the URL? Remember that & is a meta '
1757             'character in most shells, so you want to put the URL in quotes, '
1758             'like  youtube-dl '
1759             '"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1760             ' or simply  youtube-dl BaW_jenozKc  .',
1761             expected=True)
1762
1763
1764 class YoutubeTruncatedIDIE(InfoExtractor):
1765     IE_NAME = 'youtube:truncated_id'
1766     IE_DESC = False  # Do not list
1767     _VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
1768
1769     _TESTS = [{
1770         'url': 'https://www.youtube.com/watch?v=N_708QY7Ob',
1771         'only_matching': True,
1772     }]
1773
1774     def _real_extract(self, url):
1775         video_id = self._match_id(url)
1776         raise ExtractorError(
1777             'Incomplete YouTube ID %s. URL %s looks truncated.' % (video_id, url),
1778             expected=True)