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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     ExtractorError,
   9     float_or_none,
  10     int_or_none,
  11     parse_duration,
  12     parse_iso8601,
  13     unescapeHTML,
  14 )
  15 from ..compat import (
  16     compat_etree_fromstring,
  17     compat_HTTPError,
  18 )
  19
  20
  21 class BBCCoUkIE(InfoExtractor):
  22     IE_NAME = 'bbc.co.uk'
  23     IE_DESC = 'BBC iPlayer'
  24     _ID_REGEX = r'[pb][\da-z]{7}'
  25     _VALID_URL = r'''(?x)
  26                     https?://
  27                         (?:www\.)?bbc\.co\.uk/
  28                         (?:
  29                             programmes/(?!articles/)|
  30                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
  31                             music/clips[/#]|
  32                             radio/player/
  33                         )
  34                         (?P<id>%s)(?!/(?:episodes|broadcasts|clips))
  35                     ''' % _ID_REGEX
  36
  37     _MEDIASELECTOR_URLS = [
  38         # Provides HQ HLS streams with even better quality that pc mediaset but fails
  39         # with geolocation in some cases when it's even not geo restricted at all (e.g.
  40         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
  41         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
  42         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
  43     ]
  44
  45     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
  46     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
  47
  48     _NAMESPACES = (
  49         _MEDIASELECTION_NS,
  50         _EMP_PLAYLIST_NS,
  51     )
  52
  53     _TESTS = [
  54         {
  55             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  56             'info_dict': {
  57                 'id': 'b039d07m',
  58                 'ext': 'mp4',
  59                 'title': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
  60                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  61             },
  62             'params': {
  63                 'skip_download': True,
  64             }
  65         },
  66         {
  67             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  68             'info_dict': {
  69                 'id': 'b00yng1d',
  70                 'ext': 'flv',
  71                 'title': 'The Man in Black: Series 3: The Printed Name',
  72                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  73                 'duration': 1800,
  74             },
  75             'params': {
  76                 # rtmp download
  77                 'skip_download': True,
  78             },
  79             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  80         },
  81         {
  82             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  83             'info_dict': {
  84                 'id': 'b00yng1d',
  85                 'ext': 'flv',
  86                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
  87                 'description': 'Emma Willis and Marvin Humes present the fifth set of blind auditions in the singing competition, as the coaches continue to build their teams based on voice alone.',
  88                 'duration': 5100,
  89             },
  90             'params': {
  91                 # rtmp download
  92                 'skip_download': True,
  93             },
  94             'skip': 'this episode is not currently available',
  95         },
  96         {
  97             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
  98             'info_dict': {
  99                 'id': 'b03k3pb7',
 100                 'ext': 'flv',
 101                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
 102                 'description': '2. Invasion',
 103                 'duration': 3600,
 104             },
 105             'params': {
 106                 # rtmp download
 107                 'skip_download': True,
 108             },
 109             'skip': 'this episode is not currently available',
 110         }, {
 111             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 112             'info_dict': {
 113                 'id': 'b04v209v',
 114                 'ext': 'flv',
 115                 'title': 'Pete Tong, The Essential New Tune Special',
 116                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 117                 'duration': 10800,
 118             },
 119             'params': {
 120                 # rtmp download
 121                 'skip_download': True,
 122             },
 123             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 124         }, {
 125             'url': 'http://www.bbc.co.uk/music/clips/p022h44b',
 126             'note': 'Audio',
 127             'info_dict': {
 128                 'id': 'p022h44j',
 129                 'ext': 'mp4',
 130                 'title': 'BBC Proms Music Guides, Rachmaninov: Symphonic Dances',
 131                 'description': "In this Proms Music Guide, Andrew McGregor looks at Rachmaninov's Symphonic Dances.",
 132                 'duration': 227,
 133             },
 134             'params': {
 135                 'skip_download': True,
 136             }
 137         }, {
 138             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
 139             'note': 'Video',
 140             'info_dict': {
 141                 'id': 'p025c103',
 142                 'ext': 'mp4',
 143                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
 144                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
 145                 'duration': 226,
 146             },
 147             'params': {
 148                 'skip_download': True,
 149             }
 150         }, {
 151             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 152             'info_dict': {
 153                 'id': 'p02n76xf',
 154                 'ext': 'flv',
 155                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 156                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 157                 'duration': 3540,
 158             },
 159             'params': {
 160                 # rtmp download
 161                 'skip_download': True,
 162             },
 163             'skip': 'this episode is not currently available',
 164         }, {
 165             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 166             'info_dict': {
 167                 'id': 'b05zmgw1',
 168                 'ext': 'flv',
 169                 'description': 'Kirsty Wark and Morgan Quaintance visit the Royal Academy as it prepares for its annual artistic extravaganza, meeting people who have come together to make the show unique.',
 170                 'title': 'Royal Academy Summer Exhibition',
 171                 'duration': 3540,
 172             },
 173             'params': {
 174                 # rtmp download
 175                 'skip_download': True,
 176             },
 177             'skip': 'this episode is not currently available',
 178         }, {
 179             # iptv-all mediaset fails with geolocation however there is no geo restriction
 180             # for this programme at all
 181             'url': 'http://www.bbc.co.uk/programmes/b06rkn85',
 182             'info_dict': {
 183                 'id': 'b06rkms3',
 184                 'ext': 'flv',
 185                 'title': "Best of the Mini-Mixes 2015: Part 3, Annie Mac's Friday Night - BBC Radio 1",
 186                 'description': "Annie has part three in the Best of the Mini-Mixes 2015, plus the year's Most Played!",
 187             },
 188             'params': {
 189                 # rtmp download
 190                 'skip_download': True,
 191             },
 192             'skip': 'this episode is not currently available on BBC iPlayer Radio',
 193         }, {
 194             # compact player (https://github.com/rg3/youtube-dl/issues/8147)
 195             'url': 'http://www.bbc.co.uk/programmes/p028bfkf/player',
 196             'info_dict': {
 197                 'id': 'p028bfkj',
 198                 'ext': 'mp4',
 199                 'title': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 200                 'description': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 201             },
 202             'params': {
 203                 'skip_download': True,
 204             },
 205         }, {
 206             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 207             'only_matching': True,
 208         }, {
 209             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 210             'only_matching': True,
 211         }, {
 212             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 213             'only_matching': True,
 214         }, {
 215             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
 216             'only_matching': True,
 217         }
 218     ]
 219
 220     class MediaSelectionError(Exception):
 221         def __init__(self, id):
 222             self.id = id
 223
 224     def _extract_asx_playlist(self, connection, programme_id):
 225         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 226         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 227
 228     def _extract_connection(self, connection, programme_id):
 229         formats = []
 230         kind = connection.get('kind')
 231         protocol = connection.get('protocol')
 232         supplier = connection.get('supplier')
 233         if protocol == 'http':
 234             href = connection.get('href')
 235             transfer_format = connection.get('transferFormat')
 236             # ASX playlist
 237             if supplier == 'asx':
 238                 for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 239                     formats.append({
 240                         'url': ref,
 241                         'format_id': 'ref%s_%s' % (i, supplier),
 242                     })
 243             # Skip DASH until supported
 244             elif transfer_format == 'dash':
 245                 pass
 246             elif transfer_format == 'hls':
 247                 formats.extend(self._extract_m3u8_formats(
 248                     href, programme_id, 'mp4', 'm3u8_native',
 249                     m3u8_id=supplier, fatal=False))
 250             # Direct link
 251             else:
 252                 formats.append({
 253                     'url': href,
 254                     'format_id': supplier or kind or protocol,
 255                 })
 256         elif protocol == 'rtmp':
 257             application = connection.get('application', 'ondemand')
 258             auth_string = connection.get('authString')
 259             identifier = connection.get('identifier')
 260             server = connection.get('server')
 261             formats.append({
 262                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 263                 'play_path': identifier,
 264                 'app': '%s?%s' % (application, auth_string),
 265                 'page_url': 'http://www.bbc.co.uk',
 266                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 267                 'rtmp_live': False,
 268                 'ext': 'flv',
 269                 'format_id': supplier,
 270             })
 271         return formats
 272
 273     def _extract_items(self, playlist):
 274         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 275
 276     def _findall_ns(self, element, xpath):
 277         elements = []
 278         for ns in self._NAMESPACES:
 279             elements.extend(element.findall(xpath % ns))
 280         return elements
 281
 282     def _extract_medias(self, media_selection):
 283         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 284         if error is None:
 285             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 286         if error is not None:
 287             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 288         return self._findall_ns(media_selection, './{%s}media')
 289
 290     def _extract_connections(self, media):
 291         return self._findall_ns(media, './{%s}connection')
 292
 293     def _extract_video(self, media, programme_id):
 294         formats = []
 295         vbr = int_or_none(media.get('bitrate'))
 296         vcodec = media.get('encoding')
 297         service = media.get('service')
 298         width = int_or_none(media.get('width'))
 299         height = int_or_none(media.get('height'))
 300         file_size = int_or_none(media.get('media_file_size'))
 301         for connection in self._extract_connections(media):
 302             conn_formats = self._extract_connection(connection, programme_id)
 303             for format in conn_formats:
 304                 if format.get('protocol') != 'm3u8_native':
 305                     format.update({
 306                         'width': width,
 307                         'height': height,
 308                         'vbr': vbr,
 309                         'vcodec': vcodec,
 310                         'filesize': file_size,
 311                     })
 312                 if service:
 313                     format['format_id'] = '%s_%s' % (service, format['format_id'])
 314             formats.extend(conn_formats)
 315         return formats
 316
 317     def _extract_audio(self, media, programme_id):
 318         formats = []
 319         abr = int_or_none(media.get('bitrate'))
 320         acodec = media.get('encoding')
 321         service = media.get('service')
 322         for connection in self._extract_connections(media):
 323             conn_formats = self._extract_connection(connection, programme_id)
 324             for format in conn_formats:
 325                 format.update({
 326                     'format_id': '%s_%s' % (service, format['format_id']),
 327                     'abr': abr,
 328                     'acodec': acodec,
 329                     'vcodec': 'none',
 330                 })
 331             formats.extend(conn_formats)
 332         return formats
 333
 334     def _get_subtitles(self, media, programme_id):
 335         subtitles = {}
 336         for connection in self._extract_connections(media):
 337             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
 338             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 339             subtitles[lang] = [
 340                 {
 341                     'url': connection.get('href'),
 342                     'ext': 'ttml',
 343                 },
 344             ]
 345         return subtitles
 346
 347     def _raise_extractor_error(self, media_selection_error):
 348         raise ExtractorError(
 349             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 350             expected=True)
 351
 352     def _download_media_selector(self, programme_id):
 353         last_exception = None
 354         for mediaselector_url in self._MEDIASELECTOR_URLS:
 355             try:
 356                 return self._download_media_selector_url(
 357                     mediaselector_url % programme_id, programme_id)
 358             except BBCCoUkIE.MediaSelectionError as e:
 359                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 360                     last_exception = e
 361                     continue
 362                 self._raise_extractor_error(e)
 363         self._raise_extractor_error(last_exception)
 364
 365     def _download_media_selector_url(self, url, programme_id=None):
 366         try:
 367             media_selection = self._download_xml(
 368                 url, programme_id, 'Downloading media selection XML')
 369         except ExtractorError as ee:
 370             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
 371                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
 372             else:
 373                 raise
 374         return self._process_media_selector(media_selection, programme_id)
 375
 376     def _process_media_selector(self, media_selection, programme_id):
 377         formats = []
 378         subtitles = None
 379
 380         for media in self._extract_medias(media_selection):
 381             kind = media.get('kind')
 382             if kind == 'audio':
 383                 formats.extend(self._extract_audio(media, programme_id))
 384             elif kind == 'video':
 385                 formats.extend(self._extract_video(media, programme_id))
 386             elif kind == 'captions':
 387                 subtitles = self.extract_subtitles(media, programme_id)
 388         return formats, subtitles
 389
 390     def _download_playlist(self, playlist_id):
 391         try:
 392             playlist = self._download_json(
 393                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 394                 playlist_id, 'Downloading playlist JSON')
 395
 396             version = playlist.get('defaultAvailableVersion')
 397             if version:
 398                 smp_config = version['smpConfig']
 399                 title = smp_config['title']
 400                 description = smp_config['summary']
 401                 for item in smp_config['items']:
 402                     kind = item['kind']
 403                     if kind != 'programme' and kind != 'radioProgramme':
 404                         continue
 405                     programme_id = item.get('vpid')
 406                     duration = int_or_none(item.get('duration'))
 407                     formats, subtitles = self._download_media_selector(programme_id)
 408                 return programme_id, title, description, duration, formats, subtitles
 409         except ExtractorError as ee:
 410             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 411                 raise
 412
 413         # fallback to legacy playlist
 414         return self._process_legacy_playlist(playlist_id)
 415
 416     def _process_legacy_playlist_url(self, url, display_id):
 417         playlist = self._download_legacy_playlist_url(url, display_id)
 418         return self._extract_from_legacy_playlist(playlist, display_id)
 419
 420     def _process_legacy_playlist(self, playlist_id):
 421         return self._process_legacy_playlist_url(
 422             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 423
 424     def _download_legacy_playlist_url(self, url, playlist_id=None):
 425         return self._download_xml(
 426             url, playlist_id, 'Downloading legacy playlist XML')
 427
 428     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 429         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 430         if no_items is not None:
 431             reason = no_items.get('reason')
 432             if reason == 'preAvailability':
 433                 msg = 'Episode %s is not yet available' % playlist_id
 434             elif reason == 'postAvailability':
 435                 msg = 'Episode %s is no longer available' % playlist_id
 436             elif reason == 'noMedia':
 437                 msg = 'Episode %s is not currently available' % playlist_id
 438             else:
 439                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 440             raise ExtractorError(msg, expected=True)
 441
 442         for item in self._extract_items(playlist):
 443             kind = item.get('kind')
 444             if kind != 'programme' and kind != 'radioProgramme':
 445                 continue
 446             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 447             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 448             description = description_el.text if description_el is not None else None
 449
 450             def get_programme_id(item):
 451                 def get_from_attributes(item):
 452                     for p in('identifier', 'group'):
 453                         value = item.get(p)
 454                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 455                             return value
 456                 get_from_attributes(item)
 457                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 458                 if mediator is not None:
 459                     return get_from_attributes(mediator)
 460
 461             programme_id = get_programme_id(item)
 462             duration = int_or_none(item.get('duration'))
 463
 464             if programme_id:
 465                 formats, subtitles = self._download_media_selector(programme_id)
 466             else:
 467                 formats, subtitles = self._process_media_selector(item, playlist_id)
 468                 programme_id = playlist_id
 469
 470         return programme_id, title, description, duration, formats, subtitles
 471
 472     def _real_extract(self, url):
 473         group_id = self._match_id(url)
 474
 475         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 476
 477         programme_id = None
 478         duration = None
 479
 480         tviplayer = self._search_regex(
 481             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 482             webpage, 'player', default=None)
 483
 484         if tviplayer:
 485             player = self._parse_json(tviplayer, group_id).get('player', {})
 486             duration = int_or_none(player.get('duration'))
 487             programme_id = player.get('vpid')
 488
 489         if not programme_id:
 490             programme_id = self._search_regex(
 491                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 492
 493         if programme_id:
 494             formats, subtitles = self._download_media_selector(programme_id)
 495             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 496                 (r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>',
 497                  r'<div[^>]+class="info"[^>]*>\s*<h1>(.+?)</h1>'), webpage, 'title')
 498             description = self._search_regex(
 499                 (r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 500                  r'<div[^>]+class="info_+synopsis"[^>]*>([^<]+)</div>'),
 501                 webpage, 'description', default=None)
 502             if not description:
 503                 description = self._html_search_meta('description', webpage)
 504         else:
 505             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 506
 507         self._sort_formats(formats)
 508
 509         return {
 510             'id': programme_id,
 511             'title': title,
 512             'description': description,
 513             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 514             'duration': duration,
 515             'formats': formats,
 516             'subtitles': subtitles,
 517         }
 518
 519
 520 class BBCIE(BBCCoUkIE):
 521     IE_NAME = 'bbc'
 522     IE_DESC = 'BBC'
 523     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 524
 525     _MEDIASELECTOR_URLS = [
 526         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 527         # even not geo restricted at all
 528         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 529         # Provides more formats, namely direct mp4 links, but fails on some videos with
 530         # notukerror for non UK (?) users (e.g.
 531         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 532         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 533         # Provides fewer formats, but works everywhere for everybody (hopefully)
 534         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 535     ]
 536
 537     _TESTS = [{
 538         # article with multiple videos embedded with data-playable containing vpids
 539         'url': 'http://www.bbc.com/news/world-europe-32668511',
 540         'info_dict': {
 541             'id': 'world-europe-32668511',
 542             'title': 'Russia stages massive WW2 parade despite Western boycott',
 543             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 544         },
 545         'playlist_count': 2,
 546     }, {
 547         # article with multiple videos embedded with data-playable (more videos)
 548         'url': 'http://www.bbc.com/news/business-28299555',
 549         'info_dict': {
 550             'id': 'business-28299555',
 551             'title': 'Farnborough Airshow: Video highlights',
 552             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 553         },
 554         'playlist_count': 9,
 555         'skip': 'Save time',
 556     }, {
 557         # article with multiple videos embedded with `new SMP()`
 558         # broken
 559         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 560         'info_dict': {
 561             'id': '3662a707-0af9-3149-963f-47bea720b460',
 562             'title': 'BUGGER',
 563         },
 564         'playlist_count': 18,
 565     }, {
 566         # single video embedded with data-playable containing vpid
 567         'url': 'http://www.bbc.com/news/world-europe-32041533',
 568         'info_dict': {
 569             'id': 'p02mprgb',
 570             'ext': 'mp4',
 571             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 572             'description': 'md5:2868290467291b37feda7863f7a83f54',
 573             'duration': 47,
 574             'timestamp': 1427219242,
 575             'upload_date': '20150324',
 576         },
 577         'params': {
 578             # rtmp download
 579             'skip_download': True,
 580         }
 581     }, {
 582         # article with single video embedded with data-playable containing XML playlist
 583         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 584         # and playlist with f4m and m3u8 as streamingUrl
 585         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 586         'info_dict': {
 587             'id': '150615_telabyad_kentin_cogu',
 588             'ext': 'mp4',
 589             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 590             'timestamp': 1434397334,
 591             'upload_date': '20150615',
 592         },
 593         'params': {
 594             'skip_download': True,
 595         }
 596     }, {
 597         # single video embedded with data-playable containing XML playlists (regional section)
 598         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 599         'info_dict': {
 600             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 601             'ext': 'mp4',
 602             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 603             'timestamp': 1434713142,
 604             'upload_date': '20150619',
 605         },
 606         'params': {
 607             'skip_download': True,
 608         }
 609     }, {
 610         # single video from video playlist embedded with vxp-playlist-data JSON
 611         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 612         'info_dict': {
 613             'id': 'p02w6qjc',
 614             'ext': 'mp4',
 615             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 616             'duration': 56,
 617             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 618         },
 619         'params': {
 620             'skip_download': True,
 621         }
 622     }, {
 623         # single video story with digitalData
 624         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 625         'info_dict': {
 626             'id': 'p02q6gc4',
 627             'ext': 'flv',
 628             'title': 'Sri Lanka’s spicy secret',
 629             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 630             'timestamp': 1437674293,
 631             'upload_date': '20150723',
 632         },
 633         'params': {
 634             # rtmp download
 635             'skip_download': True,
 636         }
 637     }, {
 638         # single video story without digitalData
 639         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 640         'info_dict': {
 641             'id': 'p018zqqg',
 642             'ext': 'mp4',
 643             'title': 'Hyundai Santa Fe Sport: Rock star',
 644             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 645             'timestamp': 1415867444,
 646             'upload_date': '20141113',
 647         },
 648         'params': {
 649             # rtmp download
 650             'skip_download': True,
 651         }
 652     }, {
 653         # single video with playlist.sxml URL in playlist param
 654         'url': 'http://www.bbc.com/sport/0/football/33653409',
 655         'info_dict': {
 656             'id': 'p02xycnp',
 657             'ext': 'mp4',
 658             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 659             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 660             'duration': 140,
 661         },
 662         'params': {
 663             # rtmp download
 664             'skip_download': True,
 665         }
 666     }, {
 667         # article with multiple videos embedded with playlist.sxml in playlist param
 668         'url': 'http://www.bbc.com/sport/0/football/34475836',
 669         'info_dict': {
 670             'id': '34475836',
 671             'title': 'Jurgen Klopp: Furious football from a witty and winning coach',
 672             'description': 'Fast-paced football, wit, wisdom and a ready smile - why Liverpool fans should come to love new boss Jurgen Klopp.',
 673         },
 674         'playlist_count': 3,
 675     }, {
 676         # school report article with single video
 677         'url': 'http://www.bbc.co.uk/schoolreport/35744779',
 678         'info_dict': {
 679             'id': '35744779',
 680             'title': 'School which breaks down barriers in Jerusalem',
 681         },
 682         'playlist_count': 1,
 683     }, {
 684         # single video with playlist URL from weather section
 685         'url': 'http://www.bbc.com/weather/features/33601775',
 686         'only_matching': True,
 687     }, {
 688         # custom redirection to www.bbc.com
 689         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 690         'only_matching': True,
 691     }, {
 692         # single video article embedded with data-media-vpid
 693         'url': 'http://www.bbc.co.uk/sport/rowing/35908187',
 694         'only_matching': True,
 695     }]
 696
 697     @classmethod
 698     def suitable(cls, url):
 699         EXCLUDE_IE = (BBCCoUkIE, BBCCoUkArticleIE, BBCCoUkIPlayerPlaylistIE, BBCCoUkPlaylistIE)
 700         return (False if any(ie.suitable(url) for ie in EXCLUDE_IE)
 701                 else super(BBCIE, cls).suitable(url))
 702
 703     def _extract_from_media_meta(self, media_meta, video_id):
 704         # Direct links to media in media metadata (e.g.
 705         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 706         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 707         source_files = media_meta.get('sourceFiles')
 708         if source_files:
 709             return [{
 710                 'url': f['url'],
 711                 'format_id': format_id,
 712                 'ext': f.get('encoding'),
 713                 'tbr': float_or_none(f.get('bitrate'), 1000),
 714                 'filesize': int_or_none(f.get('filesize')),
 715             } for format_id, f in source_files.items() if f.get('url')], []
 716
 717         programme_id = media_meta.get('externalId')
 718         if programme_id:
 719             return self._download_media_selector(programme_id)
 720
 721         # Process playlist.sxml as legacy playlist
 722         href = media_meta.get('href')
 723         if href:
 724             playlist = self._download_legacy_playlist_url(href)
 725             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 726             return formats, subtitles
 727
 728         return [], []
 729
 730     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 731         programme_id, title, description, duration, formats, subtitles = \
 732             self._process_legacy_playlist_url(url, playlist_id)
 733         self._sort_formats(formats)
 734         return {
 735             'id': programme_id,
 736             'title': title,
 737             'description': description,
 738             'duration': duration,
 739             'timestamp': timestamp,
 740             'formats': formats,
 741             'subtitles': subtitles,
 742         }
 743
 744     def _real_extract(self, url):
 745         playlist_id = self._match_id(url)
 746
 747         webpage = self._download_webpage(url, playlist_id)
 748
 749         json_ld_info = self._search_json_ld(webpage, playlist_id, default=None)
 750         timestamp = json_ld_info.get('timestamp')
 751
 752         playlist_title = json_ld_info.get('title')
 753         if not playlist_title:
 754             playlist_title = self._og_search_title(
 755                 webpage, default=None) or self._html_search_regex(
 756                 r'<title>(.+?)</title>', webpage, 'playlist title', default=None)
 757             if playlist_title:
 758                 playlist_title = re.sub(r'(.+)\s*-\s*BBC.*?$', r'\1', playlist_title).strip()
 759
 760         playlist_description = json_ld_info.get(
 761             'description') or self._og_search_description(webpage, default=None)
 762
 763         if not timestamp:
 764             timestamp = parse_iso8601(self._search_regex(
 765                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 766                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 767                  r'"datePublished":\s*"([^"]+)'],
 768                 webpage, 'date', default=None))
 769
 770         entries = []
 771
 772         # article with multiple videos embedded with playlist.sxml (e.g.
 773         # http://www.bbc.com/sport/0/football/34475836)
 774         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 775         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 776         if playlists:
 777             entries = [
 778                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 779                 for playlist_url in playlists]
 780
 781         # news article with multiple videos embedded with data-playable
 782         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 783         if data_playables:
 784             for _, data_playable_json in data_playables:
 785                 data_playable = self._parse_json(
 786                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 787                 if not data_playable:
 788                     continue
 789                 settings = data_playable.get('settings', {})
 790                 if settings:
 791                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 792                     # http://www.bbc.com/news/world-us-canada-34473351)
 793                     playlist_object = settings.get('playlistObject', {})
 794                     if playlist_object:
 795                         items = playlist_object.get('items')
 796                         if items and isinstance(items, list):
 797                             title = playlist_object['title']
 798                             description = playlist_object.get('summary')
 799                             duration = int_or_none(items[0].get('duration'))
 800                             programme_id = items[0].get('vpid')
 801                             formats, subtitles = self._download_media_selector(programme_id)
 802                             self._sort_formats(formats)
 803                             entries.append({
 804                                 'id': programme_id,
 805                                 'title': title,
 806                                 'description': description,
 807                                 'timestamp': timestamp,
 808                                 'duration': duration,
 809                                 'formats': formats,
 810                                 'subtitles': subtitles,
 811                             })
 812                     else:
 813                         # data-playable without vpid but with a playlist.sxml URLs
 814                         # in otherSettings.playlist (e.g.
 815                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 816                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 817                         if playlist:
 818                             entries.append(self._extract_from_playlist_sxml(
 819                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
 820
 821         if entries:
 822             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 823
 824         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 825         programme_id = self._search_regex(
 826             [r'data-(?:video-player|media)-vpid="(%s)"' % self._ID_REGEX,
 827              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 828              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 829             webpage, 'vpid', default=None)
 830
 831         if programme_id:
 832             formats, subtitles = self._download_media_selector(programme_id)
 833             self._sort_formats(formats)
 834             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 835             digital_data = self._parse_json(
 836                 self._search_regex(
 837                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 838                 programme_id, fatal=False)
 839             page_info = digital_data.get('page', {}).get('pageInfo', {})
 840             title = page_info.get('pageName') or self._og_search_title(webpage)
 841             description = page_info.get('description') or self._og_search_description(webpage)
 842             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 843             return {
 844                 'id': programme_id,
 845                 'title': title,
 846                 'description': description,
 847                 'timestamp': timestamp,
 848                 'formats': formats,
 849                 'subtitles': subtitles,
 850             }
 851
 852         def extract_all(pattern):
 853             return list(filter(None, map(
 854                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 855                 re.findall(pattern, webpage))))
 856
 857         # Multiple video article (e.g.
 858         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 859         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 860         entries = []
 861         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 862             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 863             if embed_url and re.match(EMBED_URL, embed_url):
 864                 entries.append(embed_url)
 865         entries.extend(re.findall(
 866             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 867         if entries:
 868             return self.playlist_result(
 869                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 870                 playlist_id, playlist_title, playlist_description)
 871
 872         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 873         medias = extract_all(r"data-media-meta='({[^']+})'")
 874
 875         if not medias:
 876             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 877             media_asset = self._search_regex(
 878                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 879                 webpage, 'media asset', default=None)
 880             if media_asset:
 881                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 882                 medias = []
 883                 for video in media_asset_page.get('videos', {}).values():
 884                     medias.extend(video.values())
 885
 886         if not medias:
 887             # Multiple video playlist with single `now playing` entry (e.g.
 888             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 889             vxp_playlist = self._parse_json(
 890                 self._search_regex(
 891                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 892                     webpage, 'playlist data'),
 893                 playlist_id)
 894             playlist_medias = []
 895             for item in vxp_playlist:
 896                 media = item.get('media')
 897                 if not media:
 898                     continue
 899                 playlist_medias.append(media)
 900                 # Download single video if found media with asset id matching the video id from URL
 901                 if item.get('advert', {}).get('assetId') == playlist_id:
 902                     medias = [media]
 903                     break
 904             # Fallback to the whole playlist
 905             if not medias:
 906                 medias = playlist_medias
 907
 908         entries = []
 909         for num, media_meta in enumerate(medias, start=1):
 910             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 911             if not formats:
 912                 continue
 913             self._sort_formats(formats)
 914
 915             video_id = media_meta.get('externalId')
 916             if not video_id:
 917                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 918
 919             title = media_meta.get('caption')
 920             if not title:
 921                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 922
 923             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 924
 925             images = []
 926             for image in media_meta.get('images', {}).values():
 927                 images.extend(image.values())
 928             if 'image' in media_meta:
 929                 images.append(media_meta['image'])
 930
 931             thumbnails = [{
 932                 'url': image.get('href'),
 933                 'width': int_or_none(image.get('width')),
 934                 'height': int_or_none(image.get('height')),
 935             } for image in images]
 936
 937             entries.append({
 938                 'id': video_id,
 939                 'title': title,
 940                 'thumbnails': thumbnails,
 941                 'duration': duration,
 942                 'timestamp': timestamp,
 943                 'formats': formats,
 944                 'subtitles': subtitles,
 945             })
 946
 947         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 948
 949
 950 class BBCCoUkArticleIE(InfoExtractor):
 951     _VALID_URL = r'https?://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 952     IE_NAME = 'bbc.co.uk:article'
 953     IE_DESC = 'BBC articles'
 954
 955     _TEST = {
 956         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 957         'info_dict': {
 958             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 959             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 960             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 961         },
 962         'playlist_count': 4,
 963         'add_ie': ['BBCCoUk'],
 964     }
 965
 966     def _real_extract(self, url):
 967         playlist_id = self._match_id(url)
 968
 969         webpage = self._download_webpage(url, playlist_id)
 970
 971         title = self._og_search_title(webpage)
 972         description = self._og_search_description(webpage).strip()
 973
 974         entries = [self.url_result(programme_url) for programme_url in re.findall(
 975             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 976
 977         return self.playlist_result(entries, playlist_id, title, description)
 978
 979
 980 class BBCCoUkPlaylistBaseIE(InfoExtractor):
 981     def _real_extract(self, url):
 982         playlist_id = self._match_id(url)
 983
 984         webpage = self._download_webpage(url, playlist_id)
 985
 986         entries = [
 987             self.url_result(self._URL_TEMPLATE % video_id, BBCCoUkIE.ie_key())
 988             for video_id in re.findall(
 989                 self._VIDEO_ID_TEMPLATE % BBCCoUkIE._ID_REGEX, webpage)]
 990
 991         title, description = self._extract_title_and_description(webpage)
 992
 993         return self.playlist_result(entries, playlist_id, title, description)
 994
 995
 996 class BBCCoUkIPlayerPlaylistIE(BBCCoUkPlaylistBaseIE):
 997     IE_NAME = 'bbc.co.uk:iplayer:playlist'
 998     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/iplayer/episodes/(?P<id>%s)' % BBCCoUkIE._ID_REGEX
 999     _URL_TEMPLATE = 'http://www.bbc.co.uk/iplayer/episode/%s'
1000     _VIDEO_ID_TEMPLATE = r'data-ip-id=["\'](%s)'
1001     _TEST = {
1002         'url': 'http://www.bbc.co.uk/iplayer/episodes/b05rcz9v',
1003         'info_dict': {
1004             'id': 'b05rcz9v',
1005             'title': 'The Disappearance',
1006             'description': 'French thriller serial about a missing teenager.',
1007         },
1008         'playlist_mincount': 6,
1009     }
1010
1011     def _extract_title_and_description(self, webpage):
1012         title = self._search_regex(r'<h1>([^<]+)</h1>', webpage, 'title', fatal=False)
1013         description = self._search_regex(
1014             r'<p[^>]+class=(["\'])subtitle\1[^>]*>(?P<value>[^<]+)</p>',
1015             webpage, 'description', fatal=False, group='value')
1016         return title, description
1017
1018
1019 class BBCCoUkPlaylistIE(BBCCoUkPlaylistBaseIE):
1020     IE_NAME = 'bbc.co.uk:playlist'
1021     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/(?P<id>%s)/(?:episodes|broadcasts|clips)' % BBCCoUkIE._ID_REGEX
1022     _URL_TEMPLATE = 'http://www.bbc.co.uk/programmes/%s'
1023     _VIDEO_ID_TEMPLATE = r'data-pid=["\'](%s)'
1024     _TESTS = [{
1025         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1026         'info_dict': {
1027             'id': 'b05rcz9v',
1028             'title': 'The Disappearance - Clips - BBC Four',
1029             'description': 'French thriller serial about a missing teenager.',
1030         },
1031         'playlist_mincount': 7,
1032     }, {
1033         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/broadcasts/2016/06',
1034         'only_matching': True,
1035     }, {
1036         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1037         'only_matching': True,
1038     }, {
1039         'url': 'http://www.bbc.co.uk/programmes/b055jkys/episodes/player',
1040         'only_matching': True,
1041     }]
1042
1043     def _extract_title_and_description(self, webpage):
1044         title = self._og_search_title(webpage, fatal=False)
1045         description = self._og_search_description(webpage)
1046         return title, description