Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/youtube.py

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