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

   1 # coding: utf-8
   2
   3 import json
   4 import netrc
   5 import re
   6 import socket
   7 import itertools
   8
   9 from .common import InfoExtractor, SearchInfoExtractor
  10 from ..utils import (
  11     compat_http_client,
  12     compat_parse_qs,
  13     compat_urllib_error,
  14     compat_urllib_parse,
  15     compat_urllib_request,
  16     compat_str,
  17
  18     clean_html,
  19     get_element_by_id,
  20     ExtractorError,
  21     unescapeHTML,
  22     unified_strdate,
  23     orderedSet,
  24 )
  25
  26
  27 class YoutubeIE(InfoExtractor):
  28     IE_DESC = u'YouTube.com'
  29     _VALID_URL = r"""^
  30                      (
  31                          (?:https?://)?                                       # http(s):// (optional)
  32                          (?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/|
  33                             tube\.majestyc\.net/)                             # the various hostnames, with wildcard subdomains
  34                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
  35                          (?:                                                  # the various things that can precede the ID:
  36                              (?:(?:v|embed|e)/)                               # v/ or embed/ or e/
  37                              |(?:                                             # or the v= param in all its forms
  38                                  (?:watch|movie(?:_popup)?(?:\.php)?)?              # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
  39                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
  40                                  (?:.*?&)?                                    # any other preceding param (like /?s=tuff&v=xxxx)
  41                                  v=
  42                              )
  43                          )?                                                   # optional -> youtube.com/xxxx is OK
  44                      )?                                                       # all until now is optional -> you can pass the naked ID
  45                      ([0-9A-Za-z_-]+)                                         # here is it! the YouTube video ID
  46                      (?(1).+)?                                                # if we found the ID, everything can follow
  47                      $"""
  48     _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
  49     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  50     _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
  51     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
  52     _NETRC_MACHINE = 'youtube'
  53     # Listed in order of quality
  54     _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '17', '13']
  55     _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13']
  56     _video_extensions = {
  57         '13': '3gp',
  58         '17': 'mp4',
  59         '18': 'mp4',
  60         '22': 'mp4',
  61         '37': 'mp4',
  62         '38': 'mp4',
  63         '43': 'webm',
  64         '44': 'webm',
  65         '45': 'webm',
  66         '46': 'webm',
  67     }
  68     _video_dimensions = {
  69         '5': '240x400',
  70         '6': '???',
  71         '13': '???',
  72         '17': '144x176',
  73         '18': '360x640',
  74         '22': '720x1280',
  75         '34': '360x640',
  76         '35': '480x854',
  77         '37': '1080x1920',
  78         '38': '3072x4096',
  79         '43': '360x640',
  80         '44': '480x854',
  81         '45': '720x1280',
  82         '46': '1080x1920',
  83     }
  84     IE_NAME = u'youtube'
  85     _TESTS = [
  86         {
  87             u"url":  u"http://www.youtube.com/watch?v=BaW_jenozKc",
  88             u"file":  u"BaW_jenozKc.mp4",
  89             u"info_dict": {
  90                 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
  91                 u"uploader": u"Philipp Hagemeister",
  92                 u"uploader_id": u"phihag",
  93                 u"upload_date": u"20121002",
  94                 u"description": u"test chars:  \"'/\\ä↭𝕐\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
  95             }
  96         },
  97         {
  98             u"url":  u"http://www.youtube.com/watch?v=1ltcDfZMA3U",
  99             u"file":  u"1ltcDfZMA3U.flv",
 100             u"note": u"Test VEVO video (#897)",
 101             u"info_dict": {
 102                 u"upload_date": u"20070518",
 103                 u"title": u"Maps - It Will Find You",
 104                 u"description": u"Music video by Maps performing It Will Find You.",
 105                 u"uploader": u"MuteUSA",
 106                 u"uploader_id": u"MuteUSA"
 107             }
 108         },
 109         {
 110             u"url":  u"http://www.youtube.com/watch?v=UxxajLWwzqY",
 111             u"file":  u"UxxajLWwzqY.mp4",
 112             u"note": u"Test generic use_cipher_signature video (#897)",
 113             u"info_dict": {
 114                 u"upload_date": u"20120506",
 115                 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
 116                 u"description": u"md5:b085c9804f5ab69f4adea963a2dceb3c",
 117                 u"uploader": u"IconaPop",
 118                 u"uploader_id": u"IconaPop"
 119             }
 120         },
 121         {
 122             u"url":  u"https://www.youtube.com/watch?v=07FYdnEawAQ",
 123             u"file":  u"07FYdnEawAQ.mp4",
 124             u"note": u"Test VEVO video with age protection (#956)",
 125             u"info_dict": {
 126                 u"upload_date": u"20130703",
 127                 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
 128                 u"description": u"md5:64249768eec3bc4276236606ea996373",
 129                 u"uploader": u"justintimberlakeVEVO",
 130                 u"uploader_id": u"justintimberlakeVEVO"
 131             }
 132         },
 133     ]
 134
 135
 136     @classmethod
 137     def suitable(cls, url):
 138         """Receives a URL and returns True if suitable for this IE."""
 139         if YoutubePlaylistIE.suitable(url) or YoutubeSubscriptionsIE.suitable(url): return False
 140         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 141
 142     def report_lang(self):
 143         """Report attempt to set language."""
 144         self.to_screen(u'Setting language')
 145
 146     def report_video_webpage_download(self, video_id):
 147         """Report attempt to download video webpage."""
 148         self.to_screen(u'%s: Downloading video webpage' % video_id)
 149
 150     def report_video_info_webpage_download(self, video_id):
 151         """Report attempt to download video info webpage."""
 152         self.to_screen(u'%s: Downloading video info webpage' % video_id)
 153
 154     def report_video_subtitles_download(self, video_id):
 155         """Report attempt to download video info webpage."""
 156         self.to_screen(u'%s: Checking available subtitles' % video_id)
 157
 158     def report_video_subtitles_request(self, video_id, sub_lang, format):
 159         """Report attempt to download video info webpage."""
 160         self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format))
 161
 162     def report_video_subtitles_available(self, video_id, sub_lang_list):
 163         """Report available subtitles."""
 164         sub_lang = ",".join(list(sub_lang_list.keys()))
 165         self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang))
 166
 167     def report_information_extraction(self, video_id):
 168         """Report attempt to extract video information."""
 169         self.to_screen(u'%s: Extracting video information' % video_id)
 170
 171     def report_unavailable_format(self, video_id, format):
 172         """Report extracted video URL."""
 173         self.to_screen(u'%s: Format %s not available' % (video_id, format))
 174
 175     def report_rtmp_download(self):
 176         """Indicate the download will use the RTMP protocol."""
 177         self.to_screen(u'RTMP download detected')
 178
 179     def _decrypt_signature(self, s):
 180         """Turn the encrypted s field into a working signature"""
 181
 182         if len(s) == 88:
 183             return s[48] + s[81:67:-1] + s[82] + s[66:62:-1] + s[85] + s[61:48:-1] + s[67] + s[47:12:-1] + s[3] + s[11:3:-1] + s[2] + s[12]
 184         elif len(s) == 87:
 185             return s[62] + s[82:62:-1] + s[83] + s[61:52:-1] + s[0] + s[51:2:-1]
 186         elif len(s) == 86:
 187             return s[2:63] + s[82] + s[64:82] + s[63]
 188         elif len(s) == 85:
 189             return s[76] + s[82:76:-1] + s[83] + s[75:60:-1] + s[0] + s[59:50:-1] + s[1] + s[49:2:-1]
 190         elif len(s) == 84:
 191             return s[83:36:-1] + s[2] + s[35:26:-1] + s[3] + s[25:3:-1] + s[26]
 192         elif len(s) == 83:
 193             return s[:81]
 194         elif len(s) == 82:
 195             return s[36] + s[79:67:-1] + s[81] + s[66:40:-1] + s[33] + s[39:36:-1] + s[40] + s[35] + s[0] + s[67] + s[32:0:-1] + s[34]
 196
 197         else:
 198             raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
 199
 200     def _get_available_subtitles(self, video_id):
 201         self.report_video_subtitles_download(video_id)
 202         request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
 203         try:
 204             sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
 205         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 206             return (u'unable to download video subtitles: %s' % compat_str(err), None)
 207         sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
 208         sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
 209         if not sub_lang_list:
 210             return (u'video doesn\'t have subtitles', None)
 211         return sub_lang_list
 212
 213     def _list_available_subtitles(self, video_id):
 214         sub_lang_list = self._get_available_subtitles(video_id)
 215         self.report_video_subtitles_available(video_id, sub_lang_list)
 216
 217     def _request_subtitle(self, sub_lang, sub_name, video_id, format):
 218         """
 219         Return tuple:
 220         (error_message, sub_lang, sub)
 221         """
 222         self.report_video_subtitles_request(video_id, sub_lang, format)
 223         params = compat_urllib_parse.urlencode({
 224             'lang': sub_lang,
 225             'name': sub_name,
 226             'v': video_id,
 227             'fmt': format,
 228         })
 229         url = 'http://www.youtube.com/api/timedtext?' + params
 230         try:
 231             sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
 232         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 233             return (u'unable to download video subtitles: %s' % compat_str(err), None, None)
 234         if not sub:
 235             return (u'Did not fetch video subtitles', None, None)
 236         return (None, sub_lang, sub)
 237
 238     def _request_automatic_caption(self, video_id, webpage):
 239         """We need the webpage for getting the captions url, pass it as an
 240            argument to speed up the process."""
 241         sub_lang = self._downloader.params.get('subtitleslang') or 'en'
 242         sub_format = self._downloader.params.get('subtitlesformat')
 243         self.to_screen(u'%s: Looking for automatic captions' % video_id)
 244         mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
 245         err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
 246         if mobj is None:
 247             return [(err_msg, None, None)]
 248         player_config = json.loads(mobj.group(1))
 249         try:
 250             args = player_config[u'args']
 251             caption_url = args[u'ttsurl']
 252             timestamp = args[u'timestamp']
 253             params = compat_urllib_parse.urlencode({
 254                 'lang': 'en',
 255                 'tlang': sub_lang,
 256                 'fmt': sub_format,
 257                 'ts': timestamp,
 258                 'kind': 'asr',
 259             })
 260             subtitles_url = caption_url + '&' + params
 261             sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
 262             return [(None, sub_lang, sub)]
 263         except KeyError:
 264             return [(err_msg, None, None)]
 265
 266     def _extract_subtitle(self, video_id):
 267         """
 268         Return a list with a tuple:
 269         [(error_message, sub_lang, sub)]
 270         """
 271         sub_lang_list = self._get_available_subtitles(video_id)
 272         sub_format = self._downloader.params.get('subtitlesformat')
 273         if  isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
 274             return [(sub_lang_list[0], None, None)]
 275         if self._downloader.params.get('subtitleslang', False):
 276             sub_lang = self._downloader.params.get('subtitleslang')
 277         elif 'en' in sub_lang_list:
 278             sub_lang = 'en'
 279         else:
 280             sub_lang = list(sub_lang_list.keys())[0]
 281         if not sub_lang in sub_lang_list:
 282             return [(u'no closed captions found in the specified language "%s"' % sub_lang, None, None)]
 283
 284         subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 285         return [subtitle]
 286
 287     def _extract_all_subtitles(self, video_id):
 288         sub_lang_list = self._get_available_subtitles(video_id)
 289         sub_format = self._downloader.params.get('subtitlesformat')
 290         if  isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
 291             return [(sub_lang_list[0], None, None)]
 292         subtitles = []
 293         for sub_lang in sub_lang_list:
 294             subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 295             subtitles.append(subtitle)
 296         return subtitles
 297
 298     def _print_formats(self, formats):
 299         print('Available formats:')
 300         for x in formats:
 301             print('%s\t:\t%s\t[%s]' %(x, self._video_extensions.get(x, 'flv'), self._video_dimensions.get(x, '???')))
 302
 303     def _real_initialize(self):
 304         if self._downloader is None:
 305             return
 306
 307         # Set language
 308         request = compat_urllib_request.Request(self._LANG_URL)
 309         try:
 310             self.report_lang()
 311             compat_urllib_request.urlopen(request).read()
 312         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 313             self._downloader.report_warning(u'unable to set language: %s' % compat_str(err))
 314             return
 315
 316         (username, password) = self._get_login_info()
 317
 318         # No authentication to be performed
 319         if username is None:
 320             return
 321
 322         request = compat_urllib_request.Request(self._LOGIN_URL)
 323         try:
 324             login_page = compat_urllib_request.urlopen(request).read().decode('utf-8')
 325         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 326             self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err))
 327             return
 328
 329         galx = None
 330         dsh = None
 331         match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page)
 332         if match:
 333           galx = match.group(1)
 334
 335         match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page)
 336         if match:
 337           dsh = match.group(1)
 338
 339         # Log in
 340         login_form_strs = {
 341                 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
 342                 u'Email': username,
 343                 u'GALX': galx,
 344                 u'Passwd': password,
 345                 u'PersistentCookie': u'yes',
 346                 u'_utf8': u'霱',
 347                 u'bgresponse': u'js_disabled',
 348                 u'checkConnection': u'',
 349                 u'checkedDomains': u'youtube',
 350                 u'dnConn': u'',
 351                 u'dsh': dsh,
 352                 u'pstMsg': u'0',
 353                 u'rmShown': u'1',
 354                 u'secTok': u'',
 355                 u'signIn': u'Sign in',
 356                 u'timeStmp': u'',
 357                 u'service': u'youtube',
 358                 u'uilel': u'3',
 359                 u'hl': u'en_US',
 360         }
 361         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
 362         # chokes on unicode
 363         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
 364         login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
 365         request = compat_urllib_request.Request(self._LOGIN_URL, login_data)
 366         try:
 367             self.report_login()
 368             login_results = compat_urllib_request.urlopen(request).read().decode('utf-8')
 369             if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
 370                 self._downloader.report_warning(u'unable to log in: bad username or password')
 371                 return
 372         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 373             self._downloader.report_warning(u'unable to log in: %s' % compat_str(err))
 374             return
 375
 376         # Confirm age
 377         age_form = {
 378                 'next_url':     '/',
 379                 'action_confirm':   'Confirm',
 380                 }
 381         request = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
 382         try:
 383             self.report_age_confirmation()
 384             compat_urllib_request.urlopen(request).read().decode('utf-8')
 385         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 386             raise ExtractorError(u'Unable to confirm age: %s' % compat_str(err))
 387
 388     def _extract_id(self, url):
 389         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 390         if mobj is None:
 391             raise ExtractorError(u'Invalid URL: %s' % url)
 392         video_id = mobj.group(2)
 393         return video_id
 394
 395     def _real_extract(self, url):
 396         if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url):
 397             self._downloader.report_warning(u'Did you forget to quote the URL? Remember that & is a meta-character in most shells, so you want to put the URL in quotes, like  youtube-dl \'http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc\' (or simply  youtube-dl BaW_jenozKc  ).')
 398
 399         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
 400         mobj = re.search(self._NEXT_URL_RE, url)
 401         if mobj:
 402             url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
 403         video_id = self._extract_id(url)
 404
 405         # Get video webpage
 406         self.report_video_webpage_download(video_id)
 407         url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
 408         request = compat_urllib_request.Request(url)
 409         try:
 410             video_webpage_bytes = compat_urllib_request.urlopen(request).read()
 411         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 412             raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
 413
 414         video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
 415
 416         # Attempt to extract SWF player URL
 417         mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
 418         if mobj is not None:
 419             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
 420         else:
 421             player_url = None
 422
 423         # Get video info
 424         self.report_video_info_webpage_download(video_id)
 425         if re.search(r'player-age-gate-content">', video_webpage) is not None:
 426             self.report_age_confirmation()
 427             age_gate = True
 428             # We simulate the access to the video from www.youtube.com/v/{video_id}
 429             # this can be viewed without login into Youtube
 430             data = compat_urllib_parse.urlencode({'video_id': video_id,
 431                                                   'el': 'embedded',
 432                                                   'gl': 'US',
 433                                                   'hl': 'en',
 434                                                   'eurl': 'https://youtube.googleapis.com/v/' + video_id,
 435                                                   'asv': 3,
 436                                                   'sts':'1588',
 437                                                   })
 438             video_info_url = 'https://www.youtube.com/get_video_info?' + data
 439             video_info_webpage = self._download_webpage(video_info_url, video_id,
 440                                     note=False,
 441                                     errnote='unable to download video info webpage')
 442             video_info = compat_parse_qs(video_info_webpage)
 443         else:
 444             age_gate = False
 445             for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
 446                 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
 447                         % (video_id, el_type))
 448                 video_info_webpage = self._download_webpage(video_info_url, video_id,
 449                                         note=False,
 450                                         errnote='unable to download video info webpage')
 451                 video_info = compat_parse_qs(video_info_webpage)
 452                 if 'token' in video_info:
 453                     break
 454         if 'token' not in video_info:
 455             if 'reason' in video_info:
 456                 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
 457             else:
 458                 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
 459
 460         # Check for "rental" videos
 461         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 462             raise ExtractorError(u'"rental" videos not supported')
 463
 464         # Start extracting information
 465         self.report_information_extraction(video_id)
 466
 467         # uploader
 468         if 'author' not in video_info:
 469             raise ExtractorError(u'Unable to extract uploader name')
 470         video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
 471
 472         # uploader_id
 473         video_uploader_id = None
 474         mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
 475         if mobj is not None:
 476             video_uploader_id = mobj.group(1)
 477         else:
 478             self._downloader.report_warning(u'unable to extract uploader nickname')
 479
 480         # title
 481         if 'title' not in video_info:
 482             raise ExtractorError(u'Unable to extract video title')
 483         video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
 484
 485         # thumbnail image
 486         # We try first to get a high quality image:
 487         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
 488                             video_webpage, re.DOTALL)
 489         if m_thumb is not None:
 490             video_thumbnail = m_thumb.group(1)
 491         elif 'thumbnail_url' not in video_info:
 492             self._downloader.report_warning(u'unable to extract video thumbnail')
 493             video_thumbnail = ''
 494         else:   # don't panic if we can't find it
 495             video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
 496
 497         # upload date
 498         upload_date = None
 499         mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
 500         if mobj is not None:
 501             upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
 502             upload_date = unified_strdate(upload_date)
 503
 504         # description
 505         video_description = get_element_by_id("eow-description", video_webpage)
 506         if video_description:
 507             video_description = clean_html(video_description)
 508         else:
 509             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
 510             if fd_mobj:
 511                 video_description = unescapeHTML(fd_mobj.group(1))
 512             else:
 513                 video_description = u''
 514
 515         # subtitles
 516         video_subtitles = None
 517
 518         if self._downloader.params.get('writesubtitles', False):
 519             video_subtitles = self._extract_subtitle(video_id)
 520             if video_subtitles:
 521                 (sub_error, sub_lang, sub) = video_subtitles[0]
 522                 if sub_error:
 523                     self._downloader.report_warning(sub_error)
 524
 525         if self._downloader.params.get('writeautomaticsub', False):
 526             video_subtitles = self._request_automatic_caption(video_id, video_webpage)
 527             (sub_error, sub_lang, sub) = video_subtitles[0]
 528             if sub_error:
 529                 self._downloader.report_warning(sub_error)
 530
 531         if self._downloader.params.get('allsubtitles', False):
 532             video_subtitles = self._extract_all_subtitles(video_id)
 533             for video_subtitle in video_subtitles:
 534                 (sub_error, sub_lang, sub) = video_subtitle
 535                 if sub_error:
 536                     self._downloader.report_warning(sub_error)
 537
 538         if self._downloader.params.get('listsubtitles', False):
 539             self._list_available_subtitles(video_id)
 540             return
 541
 542         if 'length_seconds' not in video_info:
 543             self._downloader.report_warning(u'unable to extract video duration')
 544             video_duration = ''
 545         else:
 546             video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
 547
 548         # Decide which formats to download
 549         req_format = self._downloader.params.get('format', None)
 550
 551         try:
 552             mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
 553             if not mobj:
 554                 raise ValueError('Could not find vevo ID')
 555             info = json.loads(mobj.group(1))
 556             args = info['args']
 557             # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
 558             # this signatures are encrypted
 559             m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
 560             if m_s is not None:
 561                 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
 562                 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
 563         except ValueError:
 564             pass
 565
 566         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
 567             self.report_rtmp_download()
 568             video_url_list = [(None, video_info['conn'][0])]
 569         elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
 570             if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]:
 571                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
 572             url_map = {}
 573             for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
 574                 url_data = compat_parse_qs(url_data_str)
 575                 if 'itag' in url_data and 'url' in url_data:
 576                     url = url_data['url'][0]
 577                     if 'sig' in url_data:
 578                         url += '&signature=' + url_data['sig'][0]
 579                     elif 's' in url_data:
 580                         if self._downloader.params.get('verbose'):
 581                             s = url_data['s'][0]
 582                             if age_gate:
 583                                 player_version = self._search_regex(r'ad3-(.+?)\.swf',
 584                                     video_info['ad3_module'][0], 'flash player',
 585                                     fatal=False)
 586                                 player = 'flash player %s' % player_version
 587                             else:
 588                                 player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
 589                                     'html5 player', fatal=False)
 590                             self.to_screen('encrypted signature length %d (%d.%d), itag %s, %s' %
 591                                 (len(s), len(s.split('.')[0]), len(s.split('.')[1]), url_data['itag'][0], player))
 592                         signature = self._decrypt_signature(url_data['s'][0])
 593                         url += '&signature=' + signature
 594                     if 'ratebypass' not in url:
 595                         url += '&ratebypass=yes'
 596                     url_map[url_data['itag'][0]] = url
 597
 598             format_limit = self._downloader.params.get('format_limit', None)
 599             available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
 600             if format_limit is not None and format_limit in available_formats:
 601                 format_list = available_formats[available_formats.index(format_limit):]
 602             else:
 603                 format_list = available_formats
 604             existing_formats = [x for x in format_list if x in url_map]
 605             if len(existing_formats) == 0:
 606                 raise ExtractorError(u'no known formats available for video')
 607             if self._downloader.params.get('listformats', None):
 608                 self._print_formats(existing_formats)
 609                 return
 610             if req_format is None or req_format == 'best':
 611                 video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
 612             elif req_format == 'worst':
 613                 video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality
 614             elif req_format in ('-1', 'all'):
 615                 video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
 616             else:
 617                 # Specific formats. We pick the first in a slash-delimeted sequence.
 618                 # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
 619                 req_formats = req_format.split('/')
 620                 video_url_list = None
 621                 for rf in req_formats:
 622                     if rf in url_map:
 623                         video_url_list = [(rf, url_map[rf])]
 624                         break
 625                 if video_url_list is None:
 626                     raise ExtractorError(u'requested format not available')
 627         else:
 628             raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
 629
 630         results = []
 631         for format_param, video_real_url in video_url_list:
 632             # Extension
 633             video_extension = self._video_extensions.get(format_param, 'flv')
 634
 635             video_format = '{0} - {1}'.format(format_param if format_param else video_extension,
 636                                               self._video_dimensions.get(format_param, '???'))
 637
 638             results.append({
 639                 'id':       video_id,
 640                 'url':      video_real_url,
 641                 'uploader': video_uploader,
 642                 'uploader_id': video_uploader_id,
 643                 'upload_date':  upload_date,
 644                 'title':    video_title,
 645                 'ext':      video_extension,
 646                 'format':   video_format,
 647                 'thumbnail':    video_thumbnail,
 648                 'description':  video_description,
 649                 'player_url':   player_url,
 650                 'subtitles':    video_subtitles,
 651                 'duration':     video_duration
 652             })
 653         return results
 654
 655 class YoutubePlaylistIE(InfoExtractor):
 656     IE_DESC = u'YouTube.com playlists'
 657     _VALID_URL = r"""(?:
 658                         (?:https?://)?
 659                         (?:\w+\.)?
 660                         youtube\.com/
 661                         (?:
 662                            (?:course|view_play_list|my_playlists|artist|playlist|watch)
 663                            \? (?:.*?&)*? (?:p|a|list)=
 664                         |  p/
 665                         )
 666                         ((?:PL|EC|UU)?[0-9A-Za-z-_]{10,})
 667                         .*
 668                      |
 669                         ((?:PL|EC|UU)[0-9A-Za-z-_]{10,})
 670                      )"""
 671     _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
 672     _MAX_RESULTS = 50
 673     IE_NAME = u'youtube:playlist'
 674
 675     @classmethod
 676     def suitable(cls, url):
 677         """Receives a URL and returns True if suitable for this IE."""
 678         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 679
 680     def _real_extract(self, url):
 681         # Extract playlist id
 682         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 683         if mobj is None:
 684             raise ExtractorError(u'Invalid URL: %s' % url)
 685
 686         # Download playlist videos from API
 687         playlist_id = mobj.group(1) or mobj.group(2)
 688         page_num = 1
 689         videos = []
 690
 691         while True:
 692             url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, self._MAX_RESULTS * (page_num - 1) + 1)
 693             page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
 694
 695             try:
 696                 response = json.loads(page)
 697             except ValueError as err:
 698                 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
 699
 700             if 'feed' not in response:
 701                 raise ExtractorError(u'Got a malformed response from YouTube API')
 702             playlist_title = response['feed']['title']['$t']
 703             if 'entry' not in response['feed']:
 704                 # Number of videos is a multiple of self._MAX_RESULTS
 705                 break
 706
 707             for entry in response['feed']['entry']:
 708                 index = entry['yt$position']['$t']
 709                 if 'media$group' in entry and 'media$player' in entry['media$group']:
 710                     videos.append((index, entry['media$group']['media$player']['url']))
 711
 712             if len(response['feed']['entry']) < self._MAX_RESULTS:
 713                 break
 714             page_num += 1
 715
 716         videos = [v[1] for v in sorted(videos)]
 717
 718         url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
 719         return [self.playlist_result(url_results, playlist_id, playlist_title)]
 720
 721
 722 class YoutubeChannelIE(InfoExtractor):
 723     IE_DESC = u'YouTube.com channels'
 724     _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
 725     _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
 726     _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
 727     _MORE_PAGES_URL = 'http://www.youtube.com/channel_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
 728     IE_NAME = u'youtube:channel'
 729
 730     def extract_videos_from_page(self, page):
 731         ids_in_page = []
 732         for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
 733             if mobj.group(1) not in ids_in_page:
 734                 ids_in_page.append(mobj.group(1))
 735         return ids_in_page
 736
 737     def _real_extract(self, url):
 738         # Extract channel id
 739         mobj = re.match(self._VALID_URL, url)
 740         if mobj is None:
 741             raise ExtractorError(u'Invalid URL: %s' % url)
 742
 743         # Download channel page
 744         channel_id = mobj.group(1)
 745         video_ids = []
 746         pagenum = 1
 747
 748         url = self._TEMPLATE_URL % (channel_id, pagenum)
 749         page = self._download_webpage(url, channel_id,
 750                                       u'Downloading page #%s' % pagenum)
 751
 752         # Extract video identifiers
 753         ids_in_page = self.extract_videos_from_page(page)
 754         video_ids.extend(ids_in_page)
 755
 756         # Download any subsequent channel pages using the json-based channel_ajax query
 757         if self._MORE_PAGES_INDICATOR in page:
 758             while True:
 759                 pagenum = pagenum + 1
 760
 761                 url = self._MORE_PAGES_URL % (pagenum, channel_id)
 762                 page = self._download_webpage(url, channel_id,
 763                                               u'Downloading page #%s' % pagenum)
 764
 765                 page = json.loads(page)
 766
 767                 ids_in_page = self.extract_videos_from_page(page['content_html'])
 768                 video_ids.extend(ids_in_page)
 769
 770                 if self._MORE_PAGES_INDICATOR  not in page['load_more_widget_html']:
 771                     break
 772
 773         self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
 774
 775         urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
 776         url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls]
 777         return [self.playlist_result(url_entries, channel_id)]
 778
 779
 780 class YoutubeUserIE(InfoExtractor):
 781     IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
 782     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
 783     _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
 784     _GDATA_PAGE_SIZE = 50
 785     _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
 786     _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
 787     IE_NAME = u'youtube:user'
 788
 789     def _real_extract(self, url):
 790         # Extract username
 791         mobj = re.match(self._VALID_URL, url)
 792         if mobj is None:
 793             raise ExtractorError(u'Invalid URL: %s' % url)
 794
 795         username = mobj.group(1)
 796
 797         # Download video ids using YouTube Data API. Result size per
 798         # query is limited (currently to 50 videos) so we need to query
 799         # page by page until there are no video ids - it means we got
 800         # all of them.
 801
 802         video_ids = []
 803         pagenum = 0
 804
 805         while True:
 806             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
 807
 808             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
 809             page = self._download_webpage(gdata_url, username,
 810                                           u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
 811
 812             # Extract video identifiers
 813             ids_in_page = []
 814
 815             for mobj in re.finditer(self._VIDEO_INDICATOR, page):
 816                 if mobj.group(1) not in ids_in_page:
 817                     ids_in_page.append(mobj.group(1))
 818
 819             video_ids.extend(ids_in_page)
 820
 821             # A little optimization - if current page is not
 822             # "full", ie. does not contain PAGE_SIZE video ids then
 823             # we can assume that this page is the last one - there
 824             # are no more ids on further pages - no need to query
 825             # again.
 826
 827             if len(ids_in_page) < self._GDATA_PAGE_SIZE:
 828                 break
 829
 830             pagenum += 1
 831
 832         urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
 833         url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
 834         return [self.playlist_result(url_results, playlist_title = username)]
 835
 836 class YoutubeSearchIE(SearchInfoExtractor):
 837     IE_DESC = u'YouTube.com searches'
 838     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
 839     _MAX_RESULTS = 1000
 840     IE_NAME = u'youtube:search'
 841     _SEARCH_KEY = 'ytsearch'
 842
 843     def report_download_page(self, query, pagenum):
 844         """Report attempt to download search page with given number."""
 845         self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
 846
 847     def _get_n_results(self, query, n):
 848         """Get a specified number of results for a query"""
 849
 850         video_ids = []
 851         pagenum = 0
 852         limit = n
 853
 854         while (50 * pagenum) < limit:
 855             self.report_download_page(query, pagenum+1)
 856             result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
 857             request = compat_urllib_request.Request(result_url)
 858             try:
 859                 data = compat_urllib_request.urlopen(request).read().decode('utf-8')
 860             except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 861                 raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
 862             api_response = json.loads(data)['data']
 863
 864             if not 'items' in api_response:
 865                 raise ExtractorError(u'[youtube] No video results')
 866
 867             new_ids = list(video['id'] for video in api_response['items'])
 868             video_ids += new_ids
 869
 870             limit = min(n, api_response['totalItems'])
 871             pagenum += 1
 872
 873         if len(video_ids) > n:
 874             video_ids = video_ids[:n]
 875         videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
 876         return self.playlist_result(videos, query)
 877
 878
 879 class YoutubeShowIE(InfoExtractor):
 880     IE_DESC = u'YouTube.com (multi-season) shows'
 881     _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
 882     IE_NAME = u'youtube:show'
 883
 884     def _real_extract(self, url):
 885         mobj = re.match(self._VALID_URL, url)
 886         show_name = mobj.group(1)
 887         webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
 888         # There's one playlist for each season of the show
 889         m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
 890         self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
 891         return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
 892
 893
 894 class YoutubeSubscriptionsIE(YoutubeIE):
 895     """It's a subclass of YoutubeIE because we need to login"""
 896     IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
 897     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
 898     IE_NAME = u'youtube:subscriptions'
 899     _FEED_TEMPLATE = 'http://www.youtube.com/feed_ajax?action_load_system_feed=1&feed_name=subscriptions&paging=%s'
 900     _PAGING_STEP = 30
 901
 902     # Overwrite YoutubeIE properties we don't want
 903     _TESTS = []
 904     @classmethod
 905     def suitable(cls, url):
 906         return re.match(cls._VALID_URL, url) is not None
 907
 908     def _real_initialize(self):
 909         (username, password) = self._get_login_info()
 910         if username is None:
 911             raise ExtractorError(u'No login info available, needed for downloading the Youtube subscriptions.', expected=True)
 912         super(YoutubeSubscriptionsIE, self)._real_initialize()
 913
 914     def _real_extract(self, url):
 915         feed_entries = []
 916         # The step argument is available only in 2.7 or higher
 917         for i in itertools.count(0):
 918             paging = i*self._PAGING_STEP
 919             info = self._download_webpage(self._FEED_TEMPLATE % paging, 'feed',
 920                                           u'Downloading page %s' % i)
 921             info = json.loads(info)
 922             feed_html = info['feed_html']
 923             m_ids = re.finditer(r'"/watch\?v=(.*?)"', feed_html)
 924             ids = orderedSet(m.group(1) for m in m_ids)
 925             feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
 926             if info['paging'] is None:
 927                 break
 928         return self.playlist_result(feed_entries, playlist_title='Youtube Subscriptions')