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     OnDemandPagedList,
  32     orderedSet,
  33     unescapeHTML,
  34     unified_strdate,
  35     uppercase_escape,
  36 )
  37
  38
  39 class YoutubeBaseInfoExtractor(InfoExtractor):
  40     """Provide base functions for Youtube extractors"""
  41     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  42     _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor'
  43     _NETRC_MACHINE = 'youtube'
  44     # If True it will raise an error if no login info is provided
  45     _LOGIN_REQUIRED = False
  46
  47     def _set_language(self):
  48         self._set_cookie(
  49             '.youtube.com', 'PREF', 'f1=50000000&hl=en',
  50             # YouTube sets the expire time to about two months
  51             expire_time=time.time() + 2 * 30 * 24 * 3600)
  52
  53     def _login(self):
  54         """
  55         Attempt to log in to YouTube.
  56         True is returned if successful or skipped.
  57         False is returned if login failed.
  58
  59         If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
  60         """
  61         (username, password) = self._get_login_info()
  62         # No authentication to be performed
  63         if username is None:
  64             if self._LOGIN_REQUIRED:
  65                 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  66             return True
  67
  68         login_page = self._download_webpage(
  69             self._LOGIN_URL, None,
  70             note='Downloading login page',
  71             errnote='unable to fetch login page', fatal=False)
  72         if login_page is False:
  73             return
  74
  75         galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
  76                                   login_page, 'Login GALX parameter')
  77
  78         # Log in
  79         login_form_strs = {
  80             'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
  81             'Email': username,
  82             'GALX': galx,
  83             'Passwd': password,
  84
  85             'PersistentCookie': 'yes',
  86             '_utf8': '霱',
  87             'bgresponse': 'js_disabled',
  88             'checkConnection': '',
  89             'checkedDomains': 'youtube',
  90             'dnConn': '',
  91             'pstMsg': '0',
  92             'rmShown': '1',
  93             'secTok': '',
  94             'signIn': 'Sign in',
  95             'timeStmp': '',
  96             'service': 'youtube',
  97             'uilel': '3',
  98             'hl': 'en_US',
  99         }
 100
 101         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
 102         # chokes on unicode
 103         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items())
 104         login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
 105
 106         req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
 107         login_results = self._download_webpage(
 108             req, None,
 109             note='Logging in', errnote='unable to log in', fatal=False)
 110         if login_results is False:
 111             return False
 112
 113         if re.search(r'id="errormsg_0_Passwd"', login_results) is not None:
 114             raise ExtractorError('Please use your account password and a two-factor code instead of an application-specific password.', expected=True)
 115
 116         # Two-Factor
 117         # TODO add SMS and phone call support - these require making a request and then prompting the user
 118
 119         if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', login_results) is not None:
 120             tfa_code = self._get_tfa_info()
 121
 122             if tfa_code is None:
 123                 self._downloader.report_warning('Two-factor authentication required. Provide it with --twofactor <code>')
 124                 self._downloader.report_warning('(Note that only TOTP (Google Authenticator App) codes work at this time.)')
 125                 return False
 126
 127             # Unlike the first login form, secTok and timeStmp are both required for the TFA form
 128
 129             match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
 130             if match is None:
 131                 self._downloader.report_warning('Failed to get secTok - did the page structure change?')
 132             secTok = match.group(1)
 133             match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U)
 134             if match is None:
 135                 self._downloader.report_warning('Failed to get timeStmp - did the page structure change?')
 136             timeStmp = match.group(1)
 137
 138             tfa_form_strs = {
 139                 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
 140                 'smsToken': '',
 141                 'smsUserPin': tfa_code,
 142                 'smsVerifyPin': 'Verify',
 143
 144                 'PersistentCookie': 'yes',
 145                 'checkConnection': '',
 146                 'checkedDomains': 'youtube',
 147                 'pstMsg': '1',
 148                 'secTok': secTok,
 149                 'timeStmp': timeStmp,
 150                 'service': 'youtube',
 151                 'hl': 'en_US',
 152             }
 153             tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in tfa_form_strs.items())
 154             tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii')
 155
 156             tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data)
 157             tfa_results = self._download_webpage(
 158                 tfa_req, None,
 159                 note='Submitting TFA code', errnote='unable to submit tfa', fatal=False)
 160
 161             if tfa_results is False:
 162                 return False
 163
 164             if re.search(r'(?i)<form[^>]* id="gaia_secondfactorform"', tfa_results) is not None:
 165                 self._downloader.report_warning('Two-factor code expired. Please try again, or use a one-use backup code instead.')
 166                 return False
 167             if re.search(r'(?i)<form[^>]* id="gaia_loginform"', tfa_results) is not None:
 168                 self._downloader.report_warning('unable to log in - did the page structure change?')
 169                 return False
 170             if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None:
 171                 self._downloader.report_warning('Your Google account has a security notice. Please log in on your web browser, resolve the notice, and try again.')
 172                 return False
 173
 174         if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
 175             self._downloader.report_warning('unable to log in: bad username or password')
 176             return False
 177         return True
 178
 179     def _real_initialize(self):
 180         if self._downloader is None:
 181             return
 182         self._set_language()
 183         if not self._login():
 184             return
 185
 186
 187 class YoutubeIE(YoutubeBaseInfoExtractor):
 188     IE_DESC = 'YouTube.com'
 189     _VALID_URL = r"""(?x)^
 190                      (
 191                          (?:https?://|//)                                    # http(s):// or protocol-independent URL
 192                          (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
 193                             (?:www\.)?deturl\.com/www\.youtube\.com/|
 194                             (?:www\.)?pwnyoutube\.com/|
 195                             (?:www\.)?yourepeat\.com/|
 196                             tube\.majestyc\.net/|
 197                             youtube\.googleapis\.com/)                        # the various hostnames, with wildcard subdomains
 198                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
 199                          (?:                                                  # the various things that can precede the ID:
 200                              (?:(?:v|embed|e)/(?!videoseries))                # v/ or embed/ or e/
 201                              |(?:                                             # or the v= param in all its forms
 202                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)?  # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
 203                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
 204                                  (?:.*?&)?                                    # any other preceding param (like /?s=tuff&v=xxxx)
 205                                  v=
 206                              )
 207                          ))
 208                          |youtu\.be/                                          # just youtu.be/xxxx
 209                          |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
 210                          )
 211                      )?                                                       # all until now is optional -> you can pass the naked ID
 212                      ([0-9A-Za-z_-]{11})                                      # here is it! the YouTube video ID
 213                      (?!.*?&list=)                                            # combined list/video URLs are handled by the playlist IE
 214                      (?(1).+)?                                                # if we found the ID, everything can follow
 215                      $"""
 216     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
 217     _formats = {
 218         '5': {'ext': 'flv', 'width': 400, 'height': 240},
 219         '6': {'ext': 'flv', 'width': 450, 'height': 270},
 220         '13': {'ext': '3gp'},
 221         '17': {'ext': '3gp', 'width': 176, 'height': 144},
 222         '18': {'ext': 'mp4', 'width': 640, 'height': 360},
 223         '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
 224         '34': {'ext': 'flv', 'width': 640, 'height': 360},
 225         '35': {'ext': 'flv', 'width': 854, 'height': 480},
 226         '36': {'ext': '3gp', 'width': 320, 'height': 240},
 227         '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
 228         '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
 229         '43': {'ext': 'webm', 'width': 640, 'height': 360},
 230         '44': {'ext': 'webm', 'width': 854, 'height': 480},
 231         '45': {'ext': 'webm', 'width': 1280, 'height': 720},
 232         '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
 233
 234
 235         # 3d videos
 236         '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20},
 237         '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
 238         '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
 239         '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
 240         '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
 241         '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
 242         '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
 243
 244         # Apple HTTP Live Streaming
 245         '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
 246         '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
 247         '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
 248         '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
 249         '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
 250         '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
 251         '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
 252
 253         # DASH mp4 video
 254         '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 255         '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 256         '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 257         '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 258         '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 259         '138': {'ext': 'mp4', 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},  # Height can vary (https://github.com/rg3/youtube-dl/issues/4559)
 260         '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 261         '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 262         '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
 263         '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'h264'},
 264         '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'h264'},
 265
 266         # Dash mp4 audio
 267         '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 48, 'preference': -50, 'container': 'm4a_dash'},
 268         '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 128, 'preference': -50, 'container': 'm4a_dash'},
 269         '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 256, 'preference': -50, 'container': 'm4a_dash'},
 270
 271         # Dash webm
 272         '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 273         '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 274         '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 275         '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 276         '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 277         '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
 278         '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'container': 'webm', 'vcodec': 'VP9'},
 279         '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 280         '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 281         '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 282         '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 283         '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 284         '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 285         '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 286         '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 287         '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
 288         '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 289         '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 290         '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 291         '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'VP9'},
 292         '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'},
 293
 294         # Dash webm audio
 295         '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50},
 296         '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
 297
 298         # Dash webm audio with opus inside
 299         '249': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50, 'preference': -50},
 300         '250': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70, 'preference': -50},
 301         '251': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160, 'preference': -50},
 302
 303         # RTMP (unnamed)
 304         '_rtmp': {'protocol': 'rtmp'},
 305     }
 306
 307     IE_NAME = 'youtube'
 308     _TESTS = [
 309         {
 310             'url': 'http://www.youtube.com/watch?v=BaW_jenozKc',
 311             'info_dict': {
 312                 'id': 'BaW_jenozKc',
 313                 'ext': 'mp4',
 314                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 315                 'uploader': 'Philipp Hagemeister',
 316                 'uploader_id': 'phihag',
 317                 'upload_date': '20121002',
 318                 '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 .',
 319                 'categories': ['Science & Technology'],
 320                 'like_count': int,
 321                 'dislike_count': int,
 322             }
 323         },
 324         {
 325             'url': 'http://www.youtube.com/watch?v=UxxajLWwzqY',
 326             'note': 'Test generic use_cipher_signature video (#897)',
 327             'info_dict': {
 328                 'id': 'UxxajLWwzqY',
 329                 'ext': 'mp4',
 330                 'upload_date': '20120506',
 331                 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
 332                 'description': 'md5:fea86fda2d5a5784273df5c7cc994d9f',
 333                 'uploader': 'Icona Pop',
 334                 'uploader_id': 'IconaPop',
 335             }
 336         },
 337         {
 338             'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
 339             'note': 'Test VEVO video with age protection (#956)',
 340             'info_dict': {
 341                 'id': '07FYdnEawAQ',
 342                 'ext': 'mp4',
 343                 'upload_date': '20130703',
 344                 'title': 'Justin Timberlake - Tunnel Vision (Explicit)',
 345                 'description': 'md5:64249768eec3bc4276236606ea996373',
 346                 'uploader': 'justintimberlakeVEVO',
 347                 'uploader_id': 'justintimberlakeVEVO',
 348             }
 349         },
 350         {
 351             'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
 352             'note': 'Embed-only video (#1746)',
 353             'info_dict': {
 354                 'id': 'yZIXLfi8CZQ',
 355                 'ext': 'mp4',
 356                 'upload_date': '20120608',
 357                 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
 358                 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
 359                 'uploader': 'SET India',
 360                 'uploader_id': 'setindia'
 361             }
 362         },
 363         {
 364             'url': 'http://www.youtube.com/watch?v=a9LDPn-MO4I',
 365             'note': '256k DASH audio (format 141) via DASH manifest',
 366             'info_dict': {
 367                 'id': 'a9LDPn-MO4I',
 368                 'ext': 'm4a',
 369                 'upload_date': '20121002',
 370                 'uploader_id': '8KVIDEO',
 371                 'description': '',
 372                 'uploader': '8KVIDEO',
 373                 'title': 'UHDTV TEST 8K VIDEO.mp4'
 374             },
 375             'params': {
 376                 'youtube_include_dash_manifest': True,
 377                 'format': '141',
 378             },
 379         },
 380         # DASH manifest with encrypted signature
 381         {
 382             'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
 383             'info_dict': {
 384                 'id': 'IB3lcPjvWLA',
 385                 'ext': 'm4a',
 386                 'title': 'Afrojack, Spree Wilson - The Spark ft. Spree Wilson',
 387                 'description': 'md5:12e7067fa6735a77bdcbb58cb1187d2d',
 388                 'uploader': 'AfrojackVEVO',
 389                 'uploader_id': 'AfrojackVEVO',
 390                 'upload_date': '20131011',
 391             },
 392             'params': {
 393                 'youtube_include_dash_manifest': True,
 394                 'format': '141',
 395             },
 396         },
 397         # JS player signature function name containing $
 398         {
 399             'url': 'https://www.youtube.com/watch?v=nfWlot6h_JM',
 400             'info_dict': {
 401                 'id': 'nfWlot6h_JM',
 402                 'ext': 'm4a',
 403                 'title': 'Taylor Swift - Shake It Off',
 404                 'description': 'md5:2acfda1b285bdd478ccec22f9918199d',
 405                 'uploader': 'TaylorSwiftVEVO',
 406                 'uploader_id': 'TaylorSwiftVEVO',
 407                 'upload_date': '20140818',
 408             },
 409             'params': {
 410                 'youtube_include_dash_manifest': True,
 411                 'format': '141',
 412             },
 413         },
 414         # Controversy video
 415         {
 416             'url': 'https://www.youtube.com/watch?v=T4XJQO3qol8',
 417             'info_dict': {
 418                 'id': 'T4XJQO3qol8',
 419                 'ext': 'mp4',
 420                 'upload_date': '20100909',
 421                 'uploader': 'The Amazing Atheist',
 422                 'uploader_id': 'TheAmazingAtheist',
 423                 'title': 'Burning Everyone\'s Koran',
 424                 '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',
 425             }
 426         },
 427         # Normal age-gate video (No vevo, embed allowed)
 428         {
 429             'url': 'http://youtube.com/watch?v=HtVdAasjOgU',
 430             'info_dict': {
 431                 'id': 'HtVdAasjOgU',
 432                 'ext': 'mp4',
 433                 'title': 'The Witcher 3: Wild Hunt - The Sword Of Destiny Trailer',
 434                 'description': 're:(?s).{100,}About the Game\n.*?The Witcher 3: Wild Hunt.{100,}',
 435                 'uploader': 'The Witcher',
 436                 'uploader_id': 'WitcherGame',
 437                 'upload_date': '20140605',
 438             },
 439         },
 440         # Age-gate video with encrypted signature
 441         {
 442             'url': 'http://www.youtube.com/watch?v=6kLq3WMV1nU',
 443             'info_dict': {
 444                 'id': '6kLq3WMV1nU',
 445                 'ext': 'mp4',
 446                 'title': 'Dedication To My Ex (Miss That) (Lyric Video)',
 447                 'description': 'md5:33765bb339e1b47e7e72b5490139bb41',
 448                 'uploader': 'LloydVEVO',
 449                 'uploader_id': 'LloydVEVO',
 450                 'upload_date': '20110629',
 451             },
 452         },
 453         # video_info is None (https://github.com/rg3/youtube-dl/issues/4421)
 454         {
 455             'url': '__2ABJjxzNo',
 456             'info_dict': {
 457                 'id': '__2ABJjxzNo',
 458                 'ext': 'mp4',
 459                 'upload_date': '20100430',
 460                 'uploader_id': 'deadmau5',
 461                 'description': 'md5:12c56784b8032162bb936a5f76d55360',
 462                 'uploader': 'deadmau5',
 463                 'title': 'Deadmau5 - Some Chords (HD)',
 464             },
 465             'expected_warnings': [
 466                 'DASH manifest missing',
 467             ]
 468         },
 469         # Olympics (https://github.com/rg3/youtube-dl/issues/4431)
 470         {
 471             'url': 'lqQg6PlCWgI',
 472             'info_dict': {
 473                 'id': 'lqQg6PlCWgI',
 474                 'ext': 'mp4',
 475                 'upload_date': '20120731',
 476                 'uploader_id': 'olympic',
 477                 'description': 'HO09  - Women -  GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games',
 478                 'uploader': 'Olympics',
 479                 'title': 'Hockey - Women -  GER-AUS - London 2012 Olympic Games',
 480             },
 481             'params': {
 482                 'skip_download': 'requires avconv',
 483             }
 484         },
 485         # Non-square pixels
 486         {
 487             'url': 'https://www.youtube.com/watch?v=_b-2C3KPAM0',
 488             'info_dict': {
 489                 'id': '_b-2C3KPAM0',
 490                 'ext': 'mp4',
 491                 'stretched_ratio': 16 / 9.,
 492                 'upload_date': '20110310',
 493                 'uploader_id': 'AllenMeow',
 494                 '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',
 495                 'uploader': '孫艾倫',
 496                 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人',
 497             },
 498         },
 499         # url_encoded_fmt_stream_map is empty string
 500         {
 501             'url': 'qEJwOuvDf7I',
 502             'info_dict': {
 503                 'id': 'qEJwOuvDf7I',
 504                 'ext': 'mp4',
 505                 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге',
 506                 'description': '',
 507                 'upload_date': '20150404',
 508                 'uploader_id': 'spbelect',
 509                 'uploader': 'Наблюдатели Петербурга',
 510             },
 511             'params': {
 512                 'skip_download': 'requires avconv',
 513             }
 514         },
 515     ]
 516
 517     def __init__(self, *args, **kwargs):
 518         super(YoutubeIE, self).__init__(*args, **kwargs)
 519         self._player_cache = {}
 520
 521     def report_video_info_webpage_download(self, video_id):
 522         """Report attempt to download video info webpage."""
 523         self.to_screen('%s: Downloading video info webpage' % video_id)
 524
 525     def report_information_extraction(self, video_id):
 526         """Report attempt to extract video information."""
 527         self.to_screen('%s: Extracting video information' % video_id)
 528
 529     def report_unavailable_format(self, video_id, format):
 530         """Report extracted video URL."""
 531         self.to_screen('%s: Format %s not available' % (video_id, format))
 532
 533     def report_rtmp_download(self):
 534         """Indicate the download will use the RTMP protocol."""
 535         self.to_screen('RTMP download detected')
 536
 537     def _signature_cache_id(self, example_sig):
 538         """ Return a string representation of a signature """
 539         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
 540
 541     def _extract_signature_function(self, video_id, player_url, example_sig):
 542         id_m = re.match(
 543             r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P<ext>[a-z]+)$',
 544             player_url)
 545         if not id_m:
 546             raise ExtractorError('Cannot identify player %r' % player_url)
 547         player_type = id_m.group('ext')
 548         player_id = id_m.group('id')
 549
 550         # Read from filesystem cache
 551         func_id = '%s_%s_%s' % (
 552             player_type, player_id, self._signature_cache_id(example_sig))
 553         assert os.path.basename(func_id) == func_id
 554
 555         cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
 556         if cache_spec is not None:
 557             return lambda s: ''.join(s[i] for i in cache_spec)
 558
 559         download_note = (
 560             'Downloading player %s' % player_url
 561             if self._downloader.params.get('verbose') else
 562             'Downloading %s player %s' % (player_type, player_id)
 563         )
 564         if player_type == 'js':
 565             code = self._download_webpage(
 566                 player_url, video_id,
 567                 note=download_note,
 568                 errnote='Download of %s failed' % player_url)
 569             res = self._parse_sig_js(code)
 570         elif player_type == 'swf':
 571             urlh = self._request_webpage(
 572                 player_url, video_id,
 573                 note=download_note,
 574                 errnote='Download of %s failed' % player_url)
 575             code = urlh.read()
 576             res = self._parse_sig_swf(code)
 577         else:
 578             assert False, 'Invalid player type %r' % player_type
 579
 580         test_string = ''.join(map(compat_chr, range(len(example_sig))))
 581         cache_res = res(test_string)
 582         cache_spec = [ord(c) for c in cache_res]
 583
 584         self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
 585         return res
 586
 587     def _print_sig_code(self, func, example_sig):
 588         def gen_sig_code(idxs):
 589             def _genslice(start, end, step):
 590                 starts = '' if start == 0 else str(start)
 591                 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
 592                 steps = '' if step == 1 else (':%d' % step)
 593                 return 's[%s%s%s]' % (starts, ends, steps)
 594
 595             step = None
 596             # Quelch pyflakes warnings - start will be set when step is set
 597             start = '(Never used)'
 598             for i, prev in zip(idxs[1:], idxs[:-1]):
 599                 if step is not None:
 600                     if i - prev == step:
 601                         continue
 602                     yield _genslice(start, prev, step)
 603                     step = None
 604                     continue
 605                 if i - prev in [-1, 1]:
 606                     step = i - prev
 607                     start = prev
 608                     continue
 609                 else:
 610                     yield 's[%d]' % prev
 611             if step is None:
 612                 yield 's[%d]' % i
 613             else:
 614                 yield _genslice(start, i, step)
 615
 616         test_string = ''.join(map(compat_chr, range(len(example_sig))))
 617         cache_res = func(test_string)
 618         cache_spec = [ord(c) for c in cache_res]
 619         expr_code = ' + '.join(gen_sig_code(cache_spec))
 620         signature_id_tuple = '(%s)' % (
 621             ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
 622         code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
 623                 '    return %s\n') % (signature_id_tuple, expr_code)
 624         self.to_screen('Extracted signature function:\n' + code)
 625
 626     def _parse_sig_js(self, jscode):
 627         funcname = self._search_regex(
 628             r'\.sig\|\|([a-zA-Z0-9$]+)\(', jscode,
 629             'Initial JS player signature function name')
 630
 631         jsi = JSInterpreter(jscode)
 632         initial_function = jsi.extract_function(funcname)
 633         return lambda s: initial_function([s])
 634
 635     def _parse_sig_swf(self, file_contents):
 636         swfi = SWFInterpreter(file_contents)
 637         TARGET_CLASSNAME = 'SignatureDecipher'
 638         searched_class = swfi.extract_class(TARGET_CLASSNAME)
 639         initial_function = swfi.extract_function(searched_class, 'decipher')
 640         return lambda s: initial_function([s])
 641
 642     def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
 643         """Turn the encrypted s field into a working signature"""
 644
 645         if player_url is None:
 646             raise ExtractorError('Cannot decrypt signature without player_url')
 647
 648         if player_url.startswith('//'):
 649             player_url = 'https:' + player_url
 650         try:
 651             player_id = (player_url, self._signature_cache_id(s))
 652             if player_id not in self._player_cache:
 653                 func = self._extract_signature_function(
 654                     video_id, player_url, s
 655                 )
 656                 self._player_cache[player_id] = func
 657             func = self._player_cache[player_id]
 658             if self._downloader.params.get('youtube_print_sig_code'):
 659                 self._print_sig_code(func, s)
 660             return func(s)
 661         except Exception as e:
 662             tb = traceback.format_exc()
 663             raise ExtractorError(
 664                 'Signature extraction failed: ' + tb, cause=e)
 665
 666     def _get_subtitles(self, video_id, webpage):
 667         try:
 668             subs_doc = self._download_xml(
 669                 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
 670                 video_id, note=False)
 671         except ExtractorError as err:
 672             self._downloader.report_warning('unable to download video subtitles: %s' % compat_str(err))
 673             return {}
 674
 675         sub_lang_list = {}
 676         for track in subs_doc.findall('track'):
 677             lang = track.attrib['lang_code']
 678             if lang in sub_lang_list:
 679                 continue
 680             sub_formats = []
 681             for ext in ['sbv', 'vtt', 'srt']:
 682                 params = compat_urllib_parse.urlencode({
 683                     'lang': lang,
 684                     'v': video_id,
 685                     'fmt': ext,
 686                     'name': track.attrib['name'].encode('utf-8'),
 687                 })
 688                 sub_formats.append({
 689                     'url': 'https://www.youtube.com/api/timedtext?' + params,
 690                     'ext': ext,
 691                 })
 692             sub_lang_list[lang] = sub_formats
 693         if not sub_lang_list:
 694             self._downloader.report_warning('video doesn\'t have subtitles')
 695             return {}
 696         return sub_lang_list
 697
 698     def _get_automatic_captions(self, video_id, webpage):
 699         """We need the webpage for getting the captions url, pass it as an
 700            argument to speed up the process."""
 701         self.to_screen('%s: Looking for automatic captions' % video_id)
 702         mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
 703         err_msg = 'Couldn\'t find automatic captions for %s' % video_id
 704         if mobj is None:
 705             self._downloader.report_warning(err_msg)
 706             return {}
 707         player_config = json.loads(mobj.group(1))
 708         try:
 709             args = player_config['args']
 710             caption_url = args['ttsurl']
 711             timestamp = args['timestamp']
 712             # We get the available subtitles
 713             list_params = compat_urllib_parse.urlencode({
 714                 'type': 'list',
 715                 'tlangs': 1,
 716                 'asrs': 1,
 717             })
 718             list_url = caption_url + '&' + list_params
 719             caption_list = self._download_xml(list_url, video_id)
 720             original_lang_node = caption_list.find('track')
 721             if original_lang_node is None:
 722                 self._downloader.report_warning('Video doesn\'t have automatic captions')
 723                 return {}
 724             original_lang = original_lang_node.attrib['lang_code']
 725             caption_kind = original_lang_node.attrib.get('kind', '')
 726
 727             sub_lang_list = {}
 728             for lang_node in caption_list.findall('target'):
 729                 sub_lang = lang_node.attrib['lang_code']
 730                 sub_formats = []
 731                 for ext in ['sbv', 'vtt', 'srt']:
 732                     params = compat_urllib_parse.urlencode({
 733                         'lang': original_lang,
 734                         'tlang': sub_lang,
 735                         'fmt': ext,
 736                         'ts': timestamp,
 737                         'kind': caption_kind,
 738                     })
 739                     sub_formats.append({
 740                         'url': caption_url + '&' + params,
 741                         'ext': ext,
 742                     })
 743                 sub_lang_list[sub_lang] = sub_formats
 744             return sub_lang_list
 745         # An extractor error can be raise by the download process if there are
 746         # no automatic captions but there are subtitles
 747         except (KeyError, ExtractorError):
 748             self._downloader.report_warning(err_msg)
 749             return {}
 750
 751     @classmethod
 752     def extract_id(cls, url):
 753         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
 754         if mobj is None:
 755             raise ExtractorError('Invalid URL: %s' % url)
 756         video_id = mobj.group(2)
 757         return video_id
 758
 759     def _extract_from_m3u8(self, manifest_url, video_id):
 760         url_map = {}
 761
 762         def _get_urls(_manifest):
 763             lines = _manifest.split('\n')
 764             urls = filter(lambda l: l and not l.startswith('#'),
 765                           lines)
 766             return urls
 767         manifest = self._download_webpage(manifest_url, video_id, 'Downloading formats manifest')
 768         formats_urls = _get_urls(manifest)
 769         for format_url in formats_urls:
 770             itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
 771             url_map[itag] = format_url
 772         return url_map
 773
 774     def _extract_annotations(self, video_id):
 775         url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
 776         return self._download_webpage(url, video_id, note='Searching for annotations.', errnote='Unable to download video annotations.')
 777
 778     def _parse_dash_manifest(
 779             self, video_id, dash_manifest_url, player_url, age_gate):
 780         def decrypt_sig(mobj):
 781             s = mobj.group(1)
 782             dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
 783             return '/signature/%s' % dec_s
 784         dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
 785         dash_doc = self._download_xml(
 786             dash_manifest_url, video_id,
 787             note='Downloading DASH manifest',
 788             errnote='Could not download DASH manifest')
 789
 790         formats = []
 791         for r in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
 792             url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
 793             if url_el is None:
 794                 continue
 795             format_id = r.attrib['id']
 796             video_url = url_el.text
 797             filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
 798             f = {
 799                 'format_id': format_id,
 800                 'url': video_url,
 801                 'width': int_or_none(r.attrib.get('width')),
 802                 'height': int_or_none(r.attrib.get('height')),
 803                 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
 804                 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
 805                 'filesize': filesize,
 806                 'fps': int_or_none(r.attrib.get('frameRate')),
 807             }
 808             try:
 809                 existing_format = next(
 810                     fo for fo in formats
 811                     if fo['format_id'] == format_id)
 812             except StopIteration:
 813                 full_info = self._formats.get(format_id, {}).copy()
 814                 full_info.update(f)
 815                 formats.append(full_info)
 816             else:
 817                 existing_format.update(f)
 818         return formats
 819
 820     def _real_extract(self, url):
 821         proto = (
 822             'http' if self._downloader.params.get('prefer_insecure', False)
 823             else 'https')
 824
 825         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
 826         mobj = re.search(self._NEXT_URL_RE, url)
 827         if mobj:
 828             url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
 829         video_id = self.extract_id(url)
 830
 831         # Get video webpage
 832         url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
 833         video_webpage = self._download_webpage(url, video_id)
 834
 835         # Attempt to extract SWF player URL
 836         mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
 837         if mobj is not None:
 838             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
 839         else:
 840             player_url = None
 841
 842         # Get video info
 843         embed_webpage = None
 844         if re.search(r'player-age-gate-content">', video_webpage) is not None:
 845             age_gate = True
 846             # We simulate the access to the video from www.youtube.com/v/{video_id}
 847             # this can be viewed without login into Youtube
 848             url = proto + '://www.youtube.com/embed/%s' % video_id
 849             embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage')
 850             data = compat_urllib_parse.urlencode({
 851                 'video_id': video_id,
 852                 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
 853                 'sts': self._search_regex(
 854                     r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
 855             })
 856             video_info_url = proto + '://www.youtube.com/get_video_info?' + data
 857             video_info_webpage = self._download_webpage(
 858                 video_info_url, video_id,
 859                 note='Refetching age-gated info webpage',
 860                 errnote='unable to download video info webpage')
 861             video_info = compat_parse_qs(video_info_webpage)
 862         else:
 863             age_gate = False
 864             try:
 865                 # Try looking directly into the video webpage
 866                 mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
 867                 if not mobj:
 868                     raise ValueError('Could not find ytplayer.config')  # caught below
 869                 json_code = uppercase_escape(mobj.group(1))
 870                 ytplayer_config = json.loads(json_code)
 871                 args = ytplayer_config['args']
 872                 # Convert to the same format returned by compat_parse_qs
 873                 video_info = dict((k, [v]) for k, v in args.items())
 874                 if ('url_encoded_fmt_stream_map' not in args or
 875                         args['url_encoded_fmt_stream_map'] == ''):
 876                     raise ValueError('No stream_map present')  # caught below
 877             except ValueError:
 878                 # We fallback to the get_video_info pages (used by the embed page)
 879                 self.report_video_info_webpage_download(video_id)
 880                 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
 881                     video_info_url = (
 882                         '%s://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
 883                         % (proto, video_id, el_type))
 884                     video_info_webpage = self._download_webpage(
 885                         video_info_url,
 886                         video_id, note=False,
 887                         errnote='unable to download video info webpage')
 888                     video_info = compat_parse_qs(video_info_webpage)
 889                     if 'token' in video_info:
 890                         break
 891         if 'token' not in video_info:
 892             if 'reason' in video_info:
 893                 raise ExtractorError(
 894                     'YouTube said: %s' % video_info['reason'][0],
 895                     expected=True, video_id=video_id)
 896             else:
 897                 raise ExtractorError(
 898                     '"token" parameter not in video info for unknown reason',
 899                     video_id=video_id)
 900
 901         if 'view_count' in video_info:
 902             view_count = int(video_info['view_count'][0])
 903         else:
 904             view_count = None
 905
 906         # Check for "rental" videos
 907         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 908             raise ExtractorError('"rental" videos not supported')
 909
 910         # Start extracting information
 911         self.report_information_extraction(video_id)
 912
 913         # uploader
 914         if 'author' not in video_info:
 915             raise ExtractorError('Unable to extract uploader name')
 916         video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
 917
 918         # uploader_id
 919         video_uploader_id = None
 920         mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
 921         if mobj is not None:
 922             video_uploader_id = mobj.group(1)
 923         else:
 924             self._downloader.report_warning('unable to extract uploader nickname')
 925
 926         # title
 927         if 'title' in video_info:
 928             video_title = video_info['title'][0]
 929         else:
 930             self._downloader.report_warning('Unable to extract video title')
 931             video_title = '_'
 932
 933         # thumbnail image
 934         # We try first to get a high quality image:
 935         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
 936                             video_webpage, re.DOTALL)
 937         if m_thumb is not None:
 938             video_thumbnail = m_thumb.group(1)
 939         elif 'thumbnail_url' not in video_info:
 940             self._downloader.report_warning('unable to extract video thumbnail')
 941             video_thumbnail = None
 942         else:   # don't panic if we can't find it
 943             video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
 944
 945         # upload date
 946         upload_date = None
 947         mobj = re.search(r'(?s)id="eow-date.*?>(.*?)</span>', video_webpage)
 948         if mobj is None:
 949             mobj = re.search(
 950                 r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
 951                 video_webpage)
 952         if mobj is not None:
 953             upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
 954             upload_date = unified_strdate(upload_date)
 955
 956         m_cat_container = self._search_regex(
 957             r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
 958             video_webpage, 'categories', default=None)
 959         if m_cat_container:
 960             category = self._html_search_regex(
 961                 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
 962                 default=None)
 963             video_categories = None if category is None else [category]
 964         else:
 965             video_categories = None
 966
 967         # description
 968         video_description = get_element_by_id("eow-description", video_webpage)
 969         if video_description:
 970             video_description = re.sub(r'''(?x)
 971                 <a\s+
 972                     (?:[a-zA-Z-]+="[^"]+"\s+)*?
 973                     title="([^"]+)"\s+
 974                     (?:[a-zA-Z-]+="[^"]+"\s+)*?
 975                     class="yt-uix-redirect-link"\s*>
 976                 [^<]+
 977                 </a>
 978             ''', r'\1', video_description)
 979             video_description = clean_html(video_description)
 980         else:
 981             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
 982             if fd_mobj:
 983                 video_description = unescapeHTML(fd_mobj.group(1))
 984             else:
 985                 video_description = ''
 986
 987         def _extract_count(count_name):
 988             count = self._search_regex(
 989                 r'id="watch-%s"[^>]*>.*?([\d,]+)\s*</span>' % re.escape(count_name),
 990                 video_webpage, count_name, default=None)
 991             if count is not None:
 992                 return int(count.replace(',', ''))
 993             return None
 994         like_count = _extract_count('like')
 995         dislike_count = _extract_count('dislike')
 996
 997         # subtitles
 998         video_subtitles = self.extract_subtitles(video_id, video_webpage)
 999         automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
1000
1001         if 'length_seconds' not in video_info:
1002             self._downloader.report_warning('unable to extract video duration')
1003             video_duration = None
1004         else:
1005             video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
1006
1007         # annotations
1008         video_annotations = None
1009         if self._downloader.params.get('writeannotations', False):
1010             video_annotations = self._extract_annotations(video_id)
1011
1012         def _map_to_format_list(urlmap):
1013             formats = []
1014             for itag, video_real_url in urlmap.items():
1015                 dct = {
1016                     'format_id': itag,
1017                     'url': video_real_url,
1018                     'player_url': player_url,
1019                 }
1020                 if itag in self._formats:
1021                     dct.update(self._formats[itag])
1022                 formats.append(dct)
1023             return formats
1024
1025         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1026             self.report_rtmp_download()
1027             formats = [{
1028                 'format_id': '_rtmp',
1029                 'protocol': 'rtmp',
1030                 'url': video_info['conn'][0],
1031                 'player_url': player_url,
1032             }]
1033         elif len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1:
1034             encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
1035             if 'rtmpe%3Dyes' in encoded_url_map:
1036                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
1037             url_map = {}
1038             for url_data_str in encoded_url_map.split(','):
1039                 url_data = compat_parse_qs(url_data_str)
1040                 if 'itag' not in url_data or 'url' not in url_data:
1041                     continue
1042                 format_id = url_data['itag'][0]
1043                 url = url_data['url'][0]
1044
1045                 if 'sig' in url_data:
1046                     url += '&signature=' + url_data['sig'][0]
1047                 elif 's' in url_data:
1048                     encrypted_sig = url_data['s'][0]
1049                     ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
1050
1051                     jsplayer_url_json = self._search_regex(
1052                         ASSETS_RE,
1053                         embed_webpage if age_gate else video_webpage,
1054                         'JS player URL (1)', default=None)
1055                     if not jsplayer_url_json and not age_gate:
1056                         # We need the embed website after all
1057                         if embed_webpage is None:
1058                             embed_url = proto + '://www.youtube.com/embed/%s' % video_id
1059                             embed_webpage = self._download_webpage(
1060                                 embed_url, video_id, 'Downloading embed webpage')
1061                         jsplayer_url_json = self._search_regex(
1062                             ASSETS_RE, embed_webpage, 'JS player URL')
1063
1064                     player_url = json.loads(jsplayer_url_json)
1065                     if player_url is None:
1066                         player_url_json = self._search_regex(
1067                             r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
1068                             video_webpage, 'age gate player URL')
1069                         player_url = json.loads(player_url_json)
1070
1071                     if self._downloader.params.get('verbose'):
1072                         if player_url is None:
1073                             player_version = 'unknown'
1074                             player_desc = 'unknown'
1075                         else:
1076                             if player_url.endswith('swf'):
1077                                 player_version = self._search_regex(
1078                                     r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
1079                                     'flash player', fatal=False)
1080                                 player_desc = 'flash player %s' % player_version
1081                             else:
1082                                 player_version = self._search_regex(
1083                                     r'html5player-([^/]+?)(?:/html5player)?\.js',
1084                                     player_url,
1085                                     'html5 player', fatal=False)
1086                                 player_desc = 'html5 player %s' % player_version
1087
1088                         parts_sizes = self._signature_cache_id(encrypted_sig)
1089                         self.to_screen('{%s} signature length %s, %s' %
1090                                        (format_id, parts_sizes, player_desc))
1091
1092                     signature = self._decrypt_signature(
1093                         encrypted_sig, video_id, player_url, age_gate)
1094                     url += '&signature=' + signature
1095                 if 'ratebypass' not in url:
1096                     url += '&ratebypass=yes'
1097                 url_map[format_id] = url
1098             formats = _map_to_format_list(url_map)
1099         elif video_info.get('hlsvp'):
1100             manifest_url = video_info['hlsvp'][0]
1101             url_map = self._extract_from_m3u8(manifest_url, video_id)
1102             formats = _map_to_format_list(url_map)
1103         else:
1104             raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
1105
1106         # Look for the DASH manifest
1107         if self._downloader.params.get('youtube_include_dash_manifest', True):
1108             dash_mpd = video_info.get('dashmpd')
1109             if dash_mpd:
1110                 dash_manifest_url = dash_mpd[0]
1111                 try:
1112                     dash_formats = self._parse_dash_manifest(
1113                         video_id, dash_manifest_url, player_url, age_gate)
1114                 except (ExtractorError, KeyError) as e:
1115                     self.report_warning(
1116                         'Skipping DASH manifest: %r' % e, video_id)
1117                 else:
1118                     # Hide the formats we found through non-DASH
1119                     dash_keys = set(df['format_id'] for df in dash_formats)
1120                     for f in formats:
1121                         if f['format_id'] in dash_keys:
1122                             f['format_id'] = 'nondash-%s' % f['format_id']
1123                             f['preference'] = f.get('preference', 0) - 10000
1124                     formats.extend(dash_formats)
1125
1126         # Check for malformed aspect ratio
1127         stretched_m = re.search(
1128             r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">',
1129             video_webpage)
1130         if stretched_m:
1131             ratio = float(stretched_m.group('w')) / float(stretched_m.group('h'))
1132             for f in formats:
1133                 if f.get('vcodec') != 'none':
1134                     f['stretched_ratio'] = ratio
1135
1136         self._sort_formats(formats)
1137
1138         return {
1139             'id': video_id,
1140             'uploader': video_uploader,
1141             'uploader_id': video_uploader_id,
1142             'upload_date': upload_date,
1143             'title': video_title,
1144             'thumbnail': video_thumbnail,
1145             'description': video_description,
1146             'categories': video_categories,
1147             'subtitles': video_subtitles,
1148             'automatic_captions': automatic_captions,
1149             'duration': video_duration,
1150             'age_limit': 18 if age_gate else 0,
1151             'annotations': video_annotations,
1152             'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
1153             'view_count': view_count,
1154             'like_count': like_count,
1155             'dislike_count': dislike_count,
1156             'average_rating': float_or_none(video_info.get('avg_rating', [None])[0]),
1157             'formats': formats,
1158         }
1159
1160
1161 class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
1162     IE_DESC = 'YouTube.com playlists'
1163     _VALID_URL = r"""(?x)(?:
1164                         (?:https?://)?
1165                         (?:\w+\.)?
1166                         youtube\.com/
1167                         (?:
1168                            (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/videoseries)
1169                            \? (?:.*?&)*? (?:p|a|list)=
1170                         |  p/
1171                         )
1172                         (
1173                             (?:PL|LL|EC|UU|FL|RD|UL)?[0-9A-Za-z-_]{10,}
1174                             # Top tracks, they can also include dots
1175                             |(?:MC)[\w\.]*
1176                         )
1177                         .*
1178                      |
1179                         ((?:PL|LL|EC|UU|FL|RD|UL)[0-9A-Za-z-_]{10,})
1180                      )"""
1181     _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
1182     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)'
1183     IE_NAME = 'youtube:playlist'
1184     _TESTS = [{
1185         'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1186         'info_dict': {
1187             'title': 'ytdl test PL',
1188             'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
1189         },
1190         'playlist_count': 3,
1191     }, {
1192         'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1193         'info_dict': {
1194             'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
1195             'title': 'YDL_Empty_List',
1196         },
1197         'playlist_count': 0,
1198     }, {
1199         'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
1200         'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1201         'info_dict': {
1202             'title': '29C3: Not my department',
1203             'id': 'PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
1204         },
1205         'playlist_count': 95,
1206     }, {
1207         'note': 'issue #673',
1208         'url': 'PLBB231211A4F62143',
1209         'info_dict': {
1210             'title': '[OLD]Team Fortress 2 (Class-based LP)',
1211             'id': 'PLBB231211A4F62143',
1212         },
1213         'playlist_mincount': 26,
1214     }, {
1215         'note': 'Large playlist',
1216         'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
1217         'info_dict': {
1218             'title': 'Uploads from Cauchemar',
1219             'id': 'UUBABnxM4Ar9ten8Mdjj1j0Q',
1220         },
1221         'playlist_mincount': 799,
1222     }, {
1223         'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1224         'info_dict': {
1225             'title': 'YDL_safe_search',
1226             'id': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
1227         },
1228         'playlist_count': 2,
1229     }, {
1230         'note': 'embedded',
1231         'url': 'http://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1232         'playlist_count': 4,
1233         'info_dict': {
1234             'title': 'JODA15',
1235             'id': 'PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
1236         }
1237     }, {
1238         'note': 'Embedded SWF player',
1239         'url': 'http://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
1240         'playlist_count': 4,
1241         'info_dict': {
1242             'title': 'JODA7',
1243             'id': 'YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ',
1244         }
1245     }, {
1246         'note': 'Buggy playlist: the webpage has a "Load more" button but it doesn\'t have more videos',
1247         'url': 'https://www.youtube.com/playlist?list=UUXw-G3eDE9trcvY2sBMM_aA',
1248         'info_dict': {
1249             'title': 'Uploads from Interstellar Movie',
1250             'id': 'UUXw-G3eDE9trcvY2sBMM_aA',
1251         },
1252         'playlist_mincout': 21,
1253     }]
1254
1255     def _real_initialize(self):
1256         self._login()
1257
1258     def _ids_to_results(self, ids):
1259         return [
1260             self.url_result(vid_id, 'Youtube', video_id=vid_id)
1261             for vid_id in ids]
1262
1263     def _extract_mix(self, playlist_id):
1264         # The mixes are generated from a single video
1265         # the id of the playlist is just 'RD' + video_id
1266         url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id)
1267         webpage = self._download_webpage(
1268             url, playlist_id, 'Downloading Youtube mix')
1269         search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
1270         title_span = (
1271             search_title('playlist-title') or
1272             search_title('title long-title') or
1273             search_title('title'))
1274         title = clean_html(title_span)
1275         ids = orderedSet(re.findall(
1276             r'''(?xs)data-video-username=".*?".*?
1277                        href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
1278             webpage))
1279         url_results = self._ids_to_results(ids)
1280
1281         return self.playlist_result(url_results, playlist_id, title)
1282
1283     def _extract_playlist(self, playlist_id):
1284         url = self._TEMPLATE_URL % playlist_id
1285         page = self._download_webpage(url, playlist_id)
1286         more_widget_html = content_html = page
1287
1288         # Check if the playlist exists or is private
1289         if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None:
1290             raise ExtractorError(
1291                 'The playlist doesn\'t exist or is private, use --username or '
1292                 '--netrc to access it.',
1293                 expected=True)
1294
1295         # Extract the video ids from the playlist pages
1296         ids = []
1297
1298         for page_num in itertools.count(1):
1299             matches = re.finditer(self._VIDEO_RE, content_html)
1300             # We remove the duplicates and the link with index 0
1301             # (it's not the first video of the playlist)
1302             new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
1303             ids.extend(new_ids)
1304
1305             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1306             if not mobj:
1307                 break
1308
1309             more = self._download_json(
1310                 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
1311                 'Downloading page #%s' % page_num,
1312                 transform_source=uppercase_escape)
1313             content_html = more['content_html']
1314             if not content_html.strip():
1315                 # Some webpages show a "Load more" button but they don't
1316                 # have more videos
1317                 break
1318             more_widget_html = more['load_more_widget_html']
1319
1320         playlist_title = self._html_search_regex(
1321             r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
1322             page, 'title')
1323
1324         url_results = self._ids_to_results(ids)
1325         return self.playlist_result(url_results, playlist_id, playlist_title)
1326
1327     def _real_extract(self, url):
1328         # Extract playlist id
1329         mobj = re.match(self._VALID_URL, url)
1330         if mobj is None:
1331             raise ExtractorError('Invalid URL: %s' % url)
1332         playlist_id = mobj.group(1) or mobj.group(2)
1333
1334         # Check if it's a video-specific URL
1335         query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
1336         if 'v' in query_dict:
1337             video_id = query_dict['v'][0]
1338             if self._downloader.params.get('noplaylist'):
1339                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1340                 return self.url_result(video_id, 'Youtube', video_id=video_id)
1341             else:
1342                 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
1343
1344         if playlist_id.startswith('RD') or playlist_id.startswith('UL'):
1345             # Mixes require a custom extraction process
1346             return self._extract_mix(playlist_id)
1347
1348         return self._extract_playlist(playlist_id)
1349
1350
1351 class YoutubeChannelIE(InfoExtractor):
1352     IE_DESC = 'YouTube.com channels'
1353     _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/(?P<id>[0-9A-Za-z_-]+)'
1354     IE_NAME = 'youtube:channel'
1355     _TESTS = [{
1356         'note': 'paginated channel',
1357         'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
1358         'playlist_mincount': 91,
1359         'info_dict': {
1360             'id': 'UCKfVa3S1e4PHvxWcwyMMg8w',
1361         }
1362     }]
1363
1364     def extract_videos_from_page(self, page):
1365         ids_in_page = []
1366         for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
1367             if mobj.group(1) not in ids_in_page:
1368                 ids_in_page.append(mobj.group(1))
1369         return ids_in_page
1370
1371     def _real_extract(self, url):
1372         channel_id = self._match_id(url)
1373
1374         video_ids = []
1375         url = 'https://www.youtube.com/channel/%s/videos' % channel_id
1376         channel_page = self._download_webpage(url, channel_id)
1377         autogenerated = re.search(r'''(?x)
1378                 class="[^"]*?(?:
1379                     channel-header-autogenerated-label|
1380                     yt-channel-title-autogenerated
1381                 )[^"]*"''', channel_page) is not None
1382
1383         if autogenerated:
1384             # The videos are contained in a single page
1385             # the ajax pages can't be used, they are empty
1386             video_ids = self.extract_videos_from_page(channel_page)
1387             entries = [
1388                 self.url_result(video_id, 'Youtube', video_id=video_id)
1389                 for video_id in video_ids]
1390             return self.playlist_result(entries, channel_id)
1391
1392         def _entries():
1393             more_widget_html = content_html = channel_page
1394             for pagenum in itertools.count(1):
1395
1396                 ids_in_page = self.extract_videos_from_page(content_html)
1397                 for video_id in ids_in_page:
1398                     yield self.url_result(
1399                         video_id, 'Youtube', video_id=video_id)
1400
1401                 mobj = re.search(
1402                     r'data-uix-load-more-href="/?(?P<more>[^"]+)"',
1403                     more_widget_html)
1404                 if not mobj:
1405                     break
1406
1407                 more = self._download_json(
1408                     'https://youtube.com/%s' % mobj.group('more'), channel_id,
1409                     'Downloading page #%s' % (pagenum + 1),
1410                     transform_source=uppercase_escape)
1411                 content_html = more['content_html']
1412                 more_widget_html = more['load_more_widget_html']
1413
1414         return self.playlist_result(_entries(), channel_id)
1415
1416
1417 class YoutubeUserIE(InfoExtractor):
1418     IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
1419     _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_-]+)'
1420     _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
1421     _GDATA_PAGE_SIZE = 50
1422     _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
1423     IE_NAME = 'youtube:user'
1424
1425     _TESTS = [{
1426         'url': 'https://www.youtube.com/user/TheLinuxFoundation',
1427         'playlist_mincount': 320,
1428         'info_dict': {
1429             'title': 'TheLinuxFoundation',
1430         }
1431     }, {
1432         'url': 'ytuser:phihag',
1433         'only_matching': True,
1434     }]
1435
1436     @classmethod
1437     def suitable(cls, url):
1438         # Don't return True if the url can be extracted with other youtube
1439         # extractor, the regex would is too permissive and it would match.
1440         other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
1441         if any(ie.suitable(url) for ie in other_ies):
1442             return False
1443         else:
1444             return super(YoutubeUserIE, cls).suitable(url)
1445
1446     def _real_extract(self, url):
1447         username = self._match_id(url)
1448
1449         # Download video ids using YouTube Data API. Result size per
1450         # query is limited (currently to 50 videos) so we need to query
1451         # page by page until there are no video ids - it means we got
1452         # all of them.
1453
1454         def download_page(pagenum):
1455             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1456
1457             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1458             page = self._download_webpage(
1459                 gdata_url, username,
1460                 'Downloading video ids from %d to %d' % (
1461                     start_index, start_index + self._GDATA_PAGE_SIZE))
1462
1463             try:
1464                 response = json.loads(page)
1465             except ValueError as err:
1466                 raise ExtractorError('Invalid JSON in API response: ' + compat_str(err))
1467             if 'entry' not in response['feed']:
1468                 return
1469
1470             # Extract video identifiers
1471             entries = response['feed']['entry']
1472             for entry in entries:
1473                 title = entry['title']['$t']
1474                 video_id = entry['id']['$t'].split('/')[-1]
1475                 yield {
1476                     '_type': 'url',
1477                     'url': video_id,
1478                     'ie_key': 'Youtube',
1479                     'id': video_id,
1480                     'title': title,
1481                 }
1482         url_results = OnDemandPagedList(download_page, self._GDATA_PAGE_SIZE)
1483
1484         return self.playlist_result(url_results, playlist_title=username)
1485
1486
1487 class YoutubeSearchIE(SearchInfoExtractor):
1488     IE_DESC = 'YouTube.com searches'
1489     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1490     _MAX_RESULTS = 1000
1491     IE_NAME = 'youtube:search'
1492     _SEARCH_KEY = 'ytsearch'
1493
1494     def _get_n_results(self, query, n):
1495         """Get a specified number of results for a query"""
1496
1497         video_ids = []
1498         pagenum = 0
1499         limit = n
1500         PAGE_SIZE = 50
1501
1502         while (PAGE_SIZE * pagenum) < limit:
1503             result_url = self._API_URL % (
1504                 compat_urllib_parse.quote_plus(query.encode('utf-8')),
1505                 (PAGE_SIZE * pagenum) + 1)
1506             data_json = self._download_webpage(
1507                 result_url, video_id='query "%s"' % query,
1508                 note='Downloading page %s' % (pagenum + 1),
1509                 errnote='Unable to download API page')
1510             data = json.loads(data_json)
1511             api_response = data['data']
1512
1513             if 'items' not in api_response:
1514                 raise ExtractorError(
1515                     '[youtube] No video results', expected=True)
1516
1517             new_ids = list(video['id'] for video in api_response['items'])
1518             video_ids += new_ids
1519
1520             limit = min(n, api_response['totalItems'])
1521             pagenum += 1
1522
1523         if len(video_ids) > n:
1524             video_ids = video_ids[:n]
1525         videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
1526                   for video_id in video_ids]
1527         return self.playlist_result(videos, query)
1528
1529
1530 class YoutubeSearchDateIE(YoutubeSearchIE):
1531     IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
1532     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
1533     _SEARCH_KEY = 'ytsearchdate'
1534     IE_DESC = 'YouTube.com searches, newest videos first'
1535
1536
1537 class YoutubeSearchURLIE(InfoExtractor):
1538     IE_DESC = 'YouTube.com search URLs'
1539     IE_NAME = 'youtube:search_url'
1540     _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
1541     _TESTS = [{
1542         'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
1543         'playlist_mincount': 5,
1544         'info_dict': {
1545             'title': 'youtube-dl test video',
1546         }
1547     }]
1548
1549     def _real_extract(self, url):
1550         mobj = re.match(self._VALID_URL, url)
1551         query = compat_urllib_parse.unquote_plus(mobj.group('query'))
1552
1553         webpage = self._download_webpage(url, query)
1554         result_code = self._search_regex(
1555             r'(?s)<ol[^>]+class="item-section"(.*?)</ol>', webpage, 'result HTML')
1556
1557         part_codes = re.findall(
1558             r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
1559         entries = []
1560         for part_code in part_codes:
1561             part_title = self._html_search_regex(
1562                 [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
1563             part_url_snippet = self._html_search_regex(
1564                 r'(?s)href="([^"]+)"', part_code, 'item URL')
1565             part_url = compat_urlparse.urljoin(
1566                 'https://www.youtube.com/', part_url_snippet)
1567             entries.append({
1568                 '_type': 'url',
1569                 'url': part_url,
1570                 'title': part_title,
1571             })
1572
1573         return {
1574             '_type': 'playlist',
1575             'entries': entries,
1576             'title': query,
1577         }
1578
1579
1580 class YoutubeShowIE(InfoExtractor):
1581     IE_DESC = 'YouTube.com (multi-season) shows'
1582     _VALID_URL = r'https?://www\.youtube\.com/show/(?P<id>[^?#]*)'
1583     IE_NAME = 'youtube:show'
1584     _TESTS = [{
1585         'url': 'http://www.youtube.com/show/airdisasters',
1586         'playlist_mincount': 3,
1587         'info_dict': {
1588             'id': 'airdisasters',
1589             'title': 'Air Disasters',
1590         }
1591     }]
1592
1593     def _real_extract(self, url):
1594         mobj = re.match(self._VALID_URL, url)
1595         playlist_id = mobj.group('id')
1596         webpage = self._download_webpage(
1597             url, playlist_id, 'Downloading show webpage')
1598         # There's one playlist for each season of the show
1599         m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1600         self.to_screen('%s: Found %s seasons' % (playlist_id, len(m_seasons)))
1601         entries = [
1602             self.url_result(
1603                 'https://www.youtube.com' + season.group(1), 'YoutubePlaylist')
1604             for season in m_seasons
1605         ]
1606         title = self._og_search_title(webpage, fatal=False)
1607
1608         return {
1609             '_type': 'playlist',
1610             'id': playlist_id,
1611             'title': title,
1612             'entries': entries,
1613         }
1614
1615
1616 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1617     """
1618     Base class for extractors that fetch info from
1619     http://www.youtube.com/feed_ajax
1620     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1621     """
1622     _LOGIN_REQUIRED = True
1623     # use action_load_personal_feed instead of action_load_system_feed
1624     _PERSONAL_FEED = False
1625
1626     @property
1627     def _FEED_TEMPLATE(self):
1628         action = 'action_load_system_feed'
1629         if self._PERSONAL_FEED:
1630             action = 'action_load_personal_feed'
1631         return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
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         feed_entries = []
1642         paging = 0
1643         for i in itertools.count(1):
1644             info = self._download_json(
1645                 self._FEED_TEMPLATE % paging,
1646                 '%s feed' % self._FEED_NAME,
1647                 'Downloading page %s' % i,
1648                 transform_source=uppercase_escape)
1649             feed_html = info.get('feed_html') or info.get('content_html')
1650             load_more_widget_html = info.get('load_more_widget_html') or feed_html
1651             m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1652             ids = orderedSet(m.group(1) for m in m_ids)
1653             feed_entries.extend(
1654                 self.url_result(video_id, 'Youtube', video_id=video_id)
1655                 for video_id in ids)
1656             mobj = re.search(
1657                 r'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)',
1658                 load_more_widget_html)
1659             if mobj is None:
1660                 break
1661             paging = mobj.group('paging')
1662         return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1663
1664
1665 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1666     IE_NAME = 'youtube:recommended'
1667     IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
1668     _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1669     _FEED_NAME = 'recommended'
1670     _PLAYLIST_TITLE = 'Youtube Recommended videos'
1671
1672
1673 class YoutubeWatchLaterIE(YoutubePlaylistIE):
1674     IE_NAME = 'youtube:watchlater'
1675     IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
1676     _VALID_URL = r'https?://www\.youtube\.com/(?:feed/watch_later|playlist\?list=WL)|:ytwatchlater'
1677
1678     _TESTS = []  # override PlaylistIE tests
1679
1680     def _real_extract(self, url):
1681         return self._extract_playlist('WL')
1682
1683
1684 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
1685     IE_NAME = 'youtube:history'
1686     IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
1687     _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory'
1688     _FEED_NAME = 'history'
1689     _PERSONAL_FEED = True
1690     _PLAYLIST_TITLE = 'Youtube Watch History'
1691
1692
1693 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1694     IE_NAME = 'youtube:favorites'
1695     IE_DESC = 'YouTube.com favourite videos, ":ytfav" for short (requires authentication)'
1696     _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
1697     _LOGIN_REQUIRED = True
1698
1699     def _real_extract(self, url):
1700         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1701         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
1702         return self.url_result(playlist_id, 'YoutubePlaylist')
1703
1704
1705 class YoutubeSubscriptionsIE(YoutubePlaylistIE):
1706     IE_NAME = 'youtube:subscriptions'
1707     IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
1708     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1709     _TESTS = []
1710
1711     def _real_extract(self, url):
1712         title = 'Youtube Subscriptions'
1713         page = self._download_webpage('https://www.youtube.com/feed/subscriptions', title)
1714
1715         # The extraction process is the same as for playlists, but the regex
1716         # for the video ids doesn't contain an index
1717         ids = []
1718         more_widget_html = content_html = page
1719
1720         for page_num in itertools.count(1):
1721             matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
1722             new_ids = orderedSet(matches)
1723             ids.extend(new_ids)
1724
1725             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
1726             if not mobj:
1727                 break
1728
1729             more = self._download_json(
1730                 'https://youtube.com/%s' % mobj.group('more'), title,
1731                 'Downloading page #%s' % page_num,
1732                 transform_source=uppercase_escape)
1733             content_html = more['content_html']
1734             more_widget_html = more['load_more_widget_html']
1735
1736         return {
1737             '_type': 'playlist',
1738             'title': title,
1739             'entries': self._ids_to_results(ids),
1740         }
1741
1742
1743 class YoutubeTruncatedURLIE(InfoExtractor):
1744     IE_NAME = 'youtube:truncated_url'
1745     IE_DESC = False  # Do not list
1746     _VALID_URL = r'''(?x)
1747         (?:https?://)?
1748         (?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/
1749         (?:watch\?(?:
1750             feature=[a-z_]+|
1751             annotation_id=annotation_[^&]+|
1752             x-yt-cl=[0-9]+|
1753             hl=[^&]*|
1754         )?
1755         |
1756             attribution_link\?a=[^&]+
1757         )
1758         $
1759     '''
1760
1761     _TESTS = [{
1762         'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
1763         'only_matching': True,
1764     }, {
1765         'url': 'http://www.youtube.com/watch?',
1766         'only_matching': True,
1767     }, {
1768         'url': 'https://www.youtube.com/watch?x-yt-cl=84503534',
1769         'only_matching': True,
1770     }, {
1771         'url': 'https://www.youtube.com/watch?feature=foo',
1772         'only_matching': True,
1773     }, {
1774         'url': 'https://www.youtube.com/watch?hl=en-GB',
1775         'only_matching': True,
1776     }]
1777
1778     def _real_extract(self, url):
1779         raise ExtractorError(
1780             'Did you forget to quote the URL? Remember that & is a meta '
1781             'character in most shells, so you want to put the URL in quotes, '
1782             'like  youtube-dl '
1783             '"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
1784             ' or simply  youtube-dl BaW_jenozKc  .',
1785             expected=True)
1786
1787
1788 class YoutubeTruncatedIDIE(InfoExtractor):
1789     IE_NAME = 'youtube:truncated_id'
1790     IE_DESC = False  # Do not list
1791     _VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
1792
1793     _TESTS = [{
1794         'url': 'https://www.youtube.com/watch?v=N_708QY7Ob',
1795         'only_matching': True,
1796     }]
1797
1798     def _real_extract(self, url):
1799         video_id = self._match_id(url)
1800         raise ExtractorError(
1801             'Incomplete YouTube ID %s. URL %s looks truncated.' % (video_id, url),
1802             expected=True)