_ 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     remove_end,
  14     unescapeHTML,
  15 )
  16 from ..compat import (
  17     compat_etree_fromstring,
  18     compat_HTTPError,
  19 )
  20
  21
  22 class BBCCoUkIE(InfoExtractor):
  23     IE_NAME = 'bbc.co.uk'
  24     IE_DESC = 'BBC iPlayer'
  25     _ID_REGEX = r'[pb][\da-z]{7}'
  26     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:(?:programmes/(?!articles/)|iplayer(?:/[^/]+)?/(?:episode/|playlist/))|music/clips[/#])(?P<id>%s)' % _ID_REGEX
  27
  28     _MEDIASELECTOR_URLS = [
  29         # Provides HQ HLS streams with even better quality that pc mediaset but fails
  30         # with geolocation in some cases when it's even not geo restricted at all (e.g.
  31         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
  32         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
  33         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
  34     ]
  35
  36     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
  37     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
  38
  39     _NAMESPACES = (
  40         _MEDIASELECTION_NS,
  41         _EMP_PLAYLIST_NS,
  42     )
  43
  44     _TESTS = [
  45         {
  46             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  47             'info_dict': {
  48                 'id': 'b039d07m',
  49                 'ext': 'flv',
  50                 'title': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
  51                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  52             },
  53             'params': {
  54                 # rtmp download
  55                 'skip_download': True,
  56             }
  57         },
  58         {
  59             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  60             'info_dict': {
  61                 'id': 'b00yng1d',
  62                 'ext': 'flv',
  63                 'title': 'The Man in Black: Series 3: The Printed Name',
  64                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  65                 'duration': 1800,
  66             },
  67             'params': {
  68                 # rtmp download
  69                 'skip_download': True,
  70             },
  71             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  72         },
  73         {
  74             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  75             'info_dict': {
  76                 'id': 'b00yng1d',
  77                 'ext': 'flv',
  78                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
  79                 '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.",
  80                 'duration': 5100,
  81             },
  82             'params': {
  83                 # rtmp download
  84                 'skip_download': True,
  85             },
  86             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
  87         },
  88         {
  89             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
  90             'info_dict': {
  91                 'id': 'b03k3pb7',
  92                 'ext': 'flv',
  93                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
  94                 'description': '2. Invasion',
  95                 'duration': 3600,
  96             },
  97             'params': {
  98                 # rtmp download
  99                 'skip_download': True,
 100             },
 101             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 102         }, {
 103             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 104             'info_dict': {
 105                 'id': 'b04v209v',
 106                 'ext': 'flv',
 107                 'title': 'Pete Tong, The Essential New Tune Special',
 108                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 109                 'duration': 10800,
 110             },
 111             'params': {
 112                 # rtmp download
 113                 'skip_download': True,
 114             },
 115             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 116         }, {
 117             'url': 'http://www.bbc.co.uk/music/clips/p02frcc3',
 118             'note': 'Audio',
 119             'info_dict': {
 120                 'id': 'p02frcch',
 121                 'ext': 'flv',
 122                 'title': 'Pete Tong, Past, Present and Future Special, Madeon - After Hours mix',
 123                 'description': 'French house superstar Madeon takes us out of the club and onto the after party.',
 124                 'duration': 3507,
 125             },
 126             'params': {
 127                 # rtmp download
 128                 'skip_download': True,
 129             }
 130         }, {
 131             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
 132             'note': 'Video',
 133             'info_dict': {
 134                 'id': 'p025c103',
 135                 'ext': 'flv',
 136                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
 137                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
 138                 'duration': 226,
 139             },
 140             'params': {
 141                 # rtmp download
 142                 'skip_download': True,
 143             }
 144         }, {
 145             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 146             'info_dict': {
 147                 'id': 'p02n76xf',
 148                 'ext': 'flv',
 149                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 150                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 151                 'duration': 3540,
 152             },
 153             'params': {
 154                 # rtmp download
 155                 'skip_download': True,
 156             },
 157             'skip': 'geolocation',
 158         }, {
 159             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 160             'info_dict': {
 161                 'id': 'b05zmgw1',
 162                 'ext': 'flv',
 163                 '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.',
 164                 'title': 'Royal Academy Summer Exhibition',
 165                 'duration': 3540,
 166             },
 167             'params': {
 168                 # rtmp download
 169                 'skip_download': True,
 170             },
 171             'skip': 'geolocation',
 172         }, {
 173             # iptv-all mediaset fails with geolocation however there is no geo restriction
 174             # for this programme at all
 175             'url': 'http://www.bbc.co.uk/programmes/b06bp7lf',
 176             'info_dict': {
 177                 'id': 'b06bp7kf',
 178                 'ext': 'flv',
 179                 'title': "Annie Mac's Friday Night, B.Traits sits in for Annie",
 180                 'description': 'B.Traits sits in for Annie Mac with a Mini-Mix from Disclosure.',
 181                 'duration': 10800,
 182             },
 183             'params': {
 184                 # rtmp download
 185                 'skip_download': True,
 186             },
 187         }, {
 188             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 189             'only_matching': True,
 190         }, {
 191             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 192             'only_matching': True,
 193         }, {
 194             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 195             'only_matching': True,
 196         }
 197     ]
 198
 199     class MediaSelectionError(Exception):
 200         def __init__(self, id):
 201             self.id = id
 202
 203     def _extract_asx_playlist(self, connection, programme_id):
 204         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 205         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 206
 207     def _extract_connection(self, connection, programme_id):
 208         formats = []
 209         kind = connection.get('kind')
 210         protocol = connection.get('protocol')
 211         supplier = connection.get('supplier')
 212         if protocol == 'http':
 213             href = connection.get('href')
 214             transfer_format = connection.get('transferFormat')
 215             # ASX playlist
 216             if supplier == 'asx':
 217                 for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 218                     formats.append({
 219                         'url': ref,
 220                         'format_id': 'ref%s_%s' % (i, supplier),
 221                     })
 222             # Skip DASH until supported
 223             elif transfer_format == 'dash':
 224                 pass
 225             elif transfer_format == 'hls':
 226                 formats.extend(self._extract_m3u8_formats(
 227                     href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 228                     m3u8_id=supplier, fatal=False))
 229             # Direct link
 230             else:
 231                 formats.append({
 232                     'url': href,
 233                     'format_id': supplier or kind or protocol,
 234                 })
 235         elif protocol == 'rtmp':
 236             application = connection.get('application', 'ondemand')
 237             auth_string = connection.get('authString')
 238             identifier = connection.get('identifier')
 239             server = connection.get('server')
 240             formats.append({
 241                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 242                 'play_path': identifier,
 243                 'app': '%s?%s' % (application, auth_string),
 244                 'page_url': 'http://www.bbc.co.uk',
 245                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 246                 'rtmp_live': False,
 247                 'ext': 'flv',
 248                 'format_id': supplier,
 249             })
 250         return formats
 251
 252     def _extract_items(self, playlist):
 253         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 254
 255     def _findall_ns(self, element, xpath):
 256         elements = []
 257         for ns in self._NAMESPACES:
 258             elements.extend(element.findall(xpath % ns))
 259         return elements
 260
 261     def _extract_medias(self, media_selection):
 262         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 263         if error is None:
 264             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 265         if error is not None:
 266             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 267         return self._findall_ns(media_selection, './{%s}media')
 268
 269     def _extract_connections(self, media):
 270         return self._findall_ns(media, './{%s}connection')
 271
 272     def _extract_video(self, media, programme_id):
 273         formats = []
 274         vbr = int_or_none(media.get('bitrate'))
 275         vcodec = media.get('encoding')
 276         service = media.get('service')
 277         width = int_or_none(media.get('width'))
 278         height = int_or_none(media.get('height'))
 279         file_size = int_or_none(media.get('media_file_size'))
 280         for connection in self._extract_connections(media):
 281             conn_formats = self._extract_connection(connection, programme_id)
 282             for format in conn_formats:
 283                 format.update({
 284                     'width': width,
 285                     'height': height,
 286                     'vbr': vbr,
 287                     'vcodec': vcodec,
 288                     'filesize': file_size,
 289                 })
 290                 if service:
 291                     format['format_id'] = '%s_%s' % (service, format['format_id'])
 292             formats.extend(conn_formats)
 293         return formats
 294
 295     def _extract_audio(self, media, programme_id):
 296         formats = []
 297         abr = int_or_none(media.get('bitrate'))
 298         acodec = media.get('encoding')
 299         service = media.get('service')
 300         for connection in self._extract_connections(media):
 301             conn_formats = self._extract_connection(connection, programme_id)
 302             for format in conn_formats:
 303                 format.update({
 304                     'format_id': '%s_%s' % (service, format['format_id']),
 305                     'abr': abr,
 306                     'acodec': acodec,
 307                 })
 308             formats.extend(conn_formats)
 309         return formats
 310
 311     def _get_subtitles(self, media, programme_id):
 312         subtitles = {}
 313         for connection in self._extract_connections(media):
 314             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
 315             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 316             subtitles[lang] = [
 317                 {
 318                     'url': connection.get('href'),
 319                     'ext': 'ttml',
 320                 },
 321             ]
 322         return subtitles
 323
 324     def _raise_extractor_error(self, media_selection_error):
 325         raise ExtractorError(
 326             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 327             expected=True)
 328
 329     def _download_media_selector(self, programme_id):
 330         last_exception = None
 331         for mediaselector_url in self._MEDIASELECTOR_URLS:
 332             try:
 333                 return self._download_media_selector_url(
 334                     mediaselector_url % programme_id, programme_id)
 335             except BBCCoUkIE.MediaSelectionError as e:
 336                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 337                     last_exception = e
 338                     continue
 339                 self._raise_extractor_error(e)
 340         self._raise_extractor_error(last_exception)
 341
 342     def _download_media_selector_url(self, url, programme_id=None):
 343         try:
 344             media_selection = self._download_xml(
 345                 url, programme_id, 'Downloading media selection XML')
 346         except ExtractorError as ee:
 347             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
 348                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
 349             else:
 350                 raise
 351         return self._process_media_selector(media_selection, programme_id)
 352
 353     def _process_media_selector(self, media_selection, programme_id):
 354         formats = []
 355         subtitles = None
 356
 357         for media in self._extract_medias(media_selection):
 358             kind = media.get('kind')
 359             if kind == 'audio':
 360                 formats.extend(self._extract_audio(media, programme_id))
 361             elif kind == 'video':
 362                 formats.extend(self._extract_video(media, programme_id))
 363             elif kind == 'captions':
 364                 subtitles = self.extract_subtitles(media, programme_id)
 365         return formats, subtitles
 366
 367     def _download_playlist(self, playlist_id):
 368         try:
 369             playlist = self._download_json(
 370                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 371                 playlist_id, 'Downloading playlist JSON')
 372
 373             version = playlist.get('defaultAvailableVersion')
 374             if version:
 375                 smp_config = version['smpConfig']
 376                 title = smp_config['title']
 377                 description = smp_config['summary']
 378                 for item in smp_config['items']:
 379                     kind = item['kind']
 380                     if kind != 'programme' and kind != 'radioProgramme':
 381                         continue
 382                     programme_id = item.get('vpid')
 383                     duration = int_or_none(item.get('duration'))
 384                     formats, subtitles = self._download_media_selector(programme_id)
 385                 return programme_id, title, description, duration, formats, subtitles
 386         except ExtractorError as ee:
 387             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 388                 raise
 389
 390         # fallback to legacy playlist
 391         return self._process_legacy_playlist(playlist_id)
 392
 393     def _process_legacy_playlist_url(self, url, display_id):
 394         playlist = self._download_legacy_playlist_url(url, display_id)
 395         return self._extract_from_legacy_playlist(playlist, display_id)
 396
 397     def _process_legacy_playlist(self, playlist_id):
 398         return self._process_legacy_playlist_url(
 399             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 400
 401     def _download_legacy_playlist_url(self, url, playlist_id=None):
 402         return self._download_xml(
 403             url, playlist_id, 'Downloading legacy playlist XML')
 404
 405     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 406         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 407         if no_items is not None:
 408             reason = no_items.get('reason')
 409             if reason == 'preAvailability':
 410                 msg = 'Episode %s is not yet available' % playlist_id
 411             elif reason == 'postAvailability':
 412                 msg = 'Episode %s is no longer available' % playlist_id
 413             elif reason == 'noMedia':
 414                 msg = 'Episode %s is not currently available' % playlist_id
 415             else:
 416                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 417             raise ExtractorError(msg, expected=True)
 418
 419         for item in self._extract_items(playlist):
 420             kind = item.get('kind')
 421             if kind != 'programme' and kind != 'radioProgramme':
 422                 continue
 423             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 424             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 425             description = description_el.text if description_el is not None else None
 426
 427             def get_programme_id(item):
 428                 def get_from_attributes(item):
 429                     for p in('identifier', 'group'):
 430                         value = item.get(p)
 431                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 432                             return value
 433                 get_from_attributes(item)
 434                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 435                 if mediator is not None:
 436                     return get_from_attributes(mediator)
 437
 438             programme_id = get_programme_id(item)
 439             duration = int_or_none(item.get('duration'))
 440
 441             if programme_id:
 442                 formats, subtitles = self._download_media_selector(programme_id)
 443             else:
 444                 formats, subtitles = self._process_media_selector(item, playlist_id)
 445                 programme_id = playlist_id
 446
 447         return programme_id, title, description, duration, formats, subtitles
 448
 449     def _real_extract(self, url):
 450         group_id = self._match_id(url)
 451
 452         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 453
 454         programme_id = None
 455         duration = None
 456
 457         tviplayer = self._search_regex(
 458             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 459             webpage, 'player', default=None)
 460
 461         if tviplayer:
 462             player = self._parse_json(tviplayer, group_id).get('player', {})
 463             duration = int_or_none(player.get('duration'))
 464             programme_id = player.get('vpid')
 465
 466         if not programme_id:
 467             programme_id = self._search_regex(
 468                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 469
 470         if programme_id:
 471             formats, subtitles = self._download_media_selector(programme_id)
 472             title = self._og_search_title(webpage)
 473             description = self._search_regex(
 474                 r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 475                 webpage, 'description', default=None)
 476             if not description:
 477                 description = self._html_search_meta('description', webpage)
 478         else:
 479             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 480
 481         self._sort_formats(formats)
 482
 483         return {
 484             'id': programme_id,
 485             'title': title,
 486             'description': description,
 487             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 488             'duration': duration,
 489             'formats': formats,
 490             'subtitles': subtitles,
 491         }
 492
 493
 494 class BBCIE(BBCCoUkIE):
 495     IE_NAME = 'bbc'
 496     IE_DESC = 'BBC'
 497     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 498
 499     _MEDIASELECTOR_URLS = [
 500         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 501         # even not geo restricted at all
 502         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 503         # Provides more formats, namely direct mp4 links, but fails on some videos with
 504         # notukerror for non UK (?) users (e.g.
 505         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 506         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 507         # Provides fewer formats, but works everywhere for everybody (hopefully)
 508         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 509     ]
 510
 511     _TESTS = [{
 512         # article with multiple videos embedded with data-playable containing vpids
 513         'url': 'http://www.bbc.com/news/world-europe-32668511',
 514         'info_dict': {
 515             'id': 'world-europe-32668511',
 516             'title': 'Russia stages massive WW2 parade despite Western boycott',
 517             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 518         },
 519         'playlist_count': 2,
 520     }, {
 521         # article with multiple videos embedded with data-playable (more videos)
 522         'url': 'http://www.bbc.com/news/business-28299555',
 523         'info_dict': {
 524             'id': 'business-28299555',
 525             'title': 'Farnborough Airshow: Video highlights',
 526             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 527         },
 528         'playlist_count': 9,
 529         'skip': 'Save time',
 530     }, {
 531         # article with multiple videos embedded with `new SMP()`
 532         # broken
 533         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 534         'info_dict': {
 535             'id': '3662a707-0af9-3149-963f-47bea720b460',
 536             'title': 'BBC Blogs - Adam Curtis - BUGGER',
 537         },
 538         'playlist_count': 18,
 539     }, {
 540         # single video embedded with data-playable containing vpid
 541         'url': 'http://www.bbc.com/news/world-europe-32041533',
 542         'info_dict': {
 543             'id': 'p02mprgb',
 544             'ext': 'mp4',
 545             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 546             'description': 'md5:2868290467291b37feda7863f7a83f54',
 547             'duration': 47,
 548             'timestamp': 1427219242,
 549             'upload_date': '20150324',
 550         },
 551         'params': {
 552             # rtmp download
 553             'skip_download': True,
 554         }
 555     }, {
 556         # article with single video embedded with data-playable containing XML playlist
 557         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 558         # and playlist with f4m and m3u8 as streamingUrl
 559         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 560         'info_dict': {
 561             'id': '150615_telabyad_kentin_cogu',
 562             'ext': 'mp4',
 563             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 564             'timestamp': 1434397334,
 565             'upload_date': '20150615',
 566         },
 567         'params': {
 568             'skip_download': True,
 569         }
 570     }, {
 571         # single video embedded with data-playable containing XML playlists (regional section)
 572         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 573         'info_dict': {
 574             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 575             'ext': 'mp4',
 576             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 577             'timestamp': 1434713142,
 578             'upload_date': '20150619',
 579         },
 580         'params': {
 581             'skip_download': True,
 582         }
 583     }, {
 584         # single video from video playlist embedded with vxp-playlist-data JSON
 585         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 586         'info_dict': {
 587             'id': 'p02w6qjc',
 588             'ext': 'mp4',
 589             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 590             'duration': 56,
 591             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 592         },
 593         'params': {
 594             'skip_download': True,
 595         }
 596     }, {
 597         # single video story with digitalData
 598         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 599         'info_dict': {
 600             'id': 'p02q6gc4',
 601             'ext': 'flv',
 602             'title': 'Sri Lanka’s spicy secret',
 603             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 604             'timestamp': 1437674293,
 605             'upload_date': '20150723',
 606         },
 607         'params': {
 608             # rtmp download
 609             'skip_download': True,
 610         }
 611     }, {
 612         # single video story without digitalData
 613         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 614         'info_dict': {
 615             'id': 'p018zqqg',
 616             'ext': 'mp4',
 617             'title': 'Hyundai Santa Fe Sport: Rock star',
 618             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 619             'timestamp': 1415867444,
 620             'upload_date': '20141113',
 621         },
 622         'params': {
 623             # rtmp download
 624             'skip_download': True,
 625         }
 626     }, {
 627         # single video with playlist.sxml URL in playlist param
 628         'url': 'http://www.bbc.com/sport/0/football/33653409',
 629         'info_dict': {
 630             'id': 'p02xycnp',
 631             'ext': 'mp4',
 632             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 633             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 634             'duration': 140,
 635         },
 636         'params': {
 637             # rtmp download
 638             'skip_download': True,
 639         }
 640     }, {
 641         # article with multiple videos embedded with playlist.sxml in playlist param
 642         'url': 'http://www.bbc.com/sport/0/football/34475836',
 643         'info_dict': {
 644             'id': '34475836',
 645             'title': 'What Liverpool can expect from Klopp',
 646         },
 647         'playlist_count': 3,
 648     }, {
 649         # single video with playlist URL from weather section
 650         'url': 'http://www.bbc.com/weather/features/33601775',
 651         'only_matching': True,
 652     }, {
 653         # custom redirection to www.bbc.com
 654         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 655         'only_matching': True,
 656     }]
 657
 658     @classmethod
 659     def suitable(cls, url):
 660         return False if BBCCoUkIE.suitable(url) or BBCCoUkArticleIE.suitable(url) else super(BBCIE, cls).suitable(url)
 661
 662     def _extract_from_media_meta(self, media_meta, video_id):
 663         # Direct links to media in media metadata (e.g.
 664         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 665         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 666         source_files = media_meta.get('sourceFiles')
 667         if source_files:
 668             return [{
 669                 'url': f['url'],
 670                 'format_id': format_id,
 671                 'ext': f.get('encoding'),
 672                 'tbr': float_or_none(f.get('bitrate'), 1000),
 673                 'filesize': int_or_none(f.get('filesize')),
 674             } for format_id, f in source_files.items() if f.get('url')], []
 675
 676         programme_id = media_meta.get('externalId')
 677         if programme_id:
 678             return self._download_media_selector(programme_id)
 679
 680         # Process playlist.sxml as legacy playlist
 681         href = media_meta.get('href')
 682         if href:
 683             playlist = self._download_legacy_playlist_url(href)
 684             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 685             return formats, subtitles
 686
 687         return [], []
 688
 689     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 690         programme_id, title, description, duration, formats, subtitles = \
 691             self._process_legacy_playlist_url(url, playlist_id)
 692         self._sort_formats(formats)
 693         return {
 694             'id': programme_id,
 695             'title': title,
 696             'description': description,
 697             'duration': duration,
 698             'timestamp': timestamp,
 699             'formats': formats,
 700             'subtitles': subtitles,
 701         }
 702
 703     def _real_extract(self, url):
 704         playlist_id = self._match_id(url)
 705
 706         webpage = self._download_webpage(url, playlist_id)
 707
 708         timestamp = None
 709         playlist_title = None
 710         playlist_description = None
 711
 712         ld = self._parse_json(
 713             self._search_regex(
 714                 r'(?s)<script type="application/ld\+json">(.+?)</script>',
 715                 webpage, 'ld json', default='{}'),
 716             playlist_id, fatal=False)
 717         if ld:
 718             timestamp = parse_iso8601(ld.get('datePublished'))
 719             playlist_title = ld.get('headline')
 720             playlist_description = ld.get('articleBody')
 721
 722         if not timestamp:
 723             timestamp = parse_iso8601(self._search_regex(
 724                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 725                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 726                  r'"datePublished":\s*"([^"]+)'],
 727                 webpage, 'date', default=None))
 728
 729         entries = []
 730
 731         # article with multiple videos embedded with playlist.sxml (e.g.
 732         # http://www.bbc.com/sport/0/football/34475836)
 733         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 734         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 735         if playlists:
 736             entries = [
 737                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 738                 for playlist_url in playlists]
 739
 740         # news article with multiple videos embedded with data-playable
 741         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 742         if data_playables:
 743             for _, data_playable_json in data_playables:
 744                 data_playable = self._parse_json(
 745                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 746                 if not data_playable:
 747                     continue
 748                 settings = data_playable.get('settings', {})
 749                 if settings:
 750                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 751                     # http://www.bbc.com/news/world-us-canada-34473351)
 752                     playlist_object = settings.get('playlistObject', {})
 753                     if playlist_object:
 754                         items = playlist_object.get('items')
 755                         if items and isinstance(items, list):
 756                             title = playlist_object['title']
 757                             description = playlist_object.get('summary')
 758                             duration = int_or_none(items[0].get('duration'))
 759                             programme_id = items[0].get('vpid')
 760                             formats, subtitles = self._download_media_selector(programme_id)
 761                             self._sort_formats(formats)
 762                             entries.append({
 763                                 'id': programme_id,
 764                                 'title': title,
 765                                 'description': description,
 766                                 'timestamp': timestamp,
 767                                 'duration': duration,
 768                                 'formats': formats,
 769                                 'subtitles': subtitles,
 770                             })
 771                     else:
 772                         # data-playable without vpid but with a playlist.sxml URLs
 773                         # in otherSettings.playlist (e.g.
 774                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 775                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 776                         if playlist:
 777                             entries.append(self._extract_from_playlist_sxml(
 778                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
 779
 780         if entries:
 781             playlist_title = playlist_title or remove_end(self._og_search_title(webpage), ' - BBC News')
 782             playlist_description = playlist_description or self._og_search_description(webpage, default=None)
 783             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 784
 785         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 786         programme_id = self._search_regex(
 787             [r'data-video-player-vpid="(%s)"' % self._ID_REGEX,
 788              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 789              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 790             webpage, 'vpid', default=None)
 791
 792         if programme_id:
 793             formats, subtitles = self._download_media_selector(programme_id)
 794             self._sort_formats(formats)
 795             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 796             digital_data = self._parse_json(
 797                 self._search_regex(
 798                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 799                 programme_id, fatal=False)
 800             page_info = digital_data.get('page', {}).get('pageInfo', {})
 801             title = page_info.get('pageName') or self._og_search_title(webpage)
 802             description = page_info.get('description') or self._og_search_description(webpage)
 803             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 804             return {
 805                 'id': programme_id,
 806                 'title': title,
 807                 'description': description,
 808                 'timestamp': timestamp,
 809                 'formats': formats,
 810                 'subtitles': subtitles,
 811             }
 812
 813         playlist_title = self._html_search_regex(
 814             r'<title>(.*?)(?:\s*-\s*BBC [^ ]+)?</title>', webpage, 'playlist title')
 815         playlist_description = self._og_search_description(webpage, default=None)
 816
 817         def extract_all(pattern):
 818             return list(filter(None, map(
 819                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 820                 re.findall(pattern, webpage))))
 821
 822         # Multiple video article (e.g.
 823         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 824         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 825         entries = []
 826         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 827             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 828             if embed_url and re.match(EMBED_URL, embed_url):
 829                 entries.append(embed_url)
 830         entries.extend(re.findall(
 831             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 832         if entries:
 833             return self.playlist_result(
 834                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 835                 playlist_id, playlist_title, playlist_description)
 836
 837         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 838         medias = extract_all(r"data-media-meta='({[^']+})'")
 839
 840         if not medias:
 841             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 842             media_asset = self._search_regex(
 843                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 844                 webpage, 'media asset', default=None)
 845             if media_asset:
 846                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 847                 medias = []
 848                 for video in media_asset_page.get('videos', {}).values():
 849                     medias.extend(video.values())
 850
 851         if not medias:
 852             # Multiple video playlist with single `now playing` entry (e.g.
 853             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 854             vxp_playlist = self._parse_json(
 855                 self._search_regex(
 856                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 857                     webpage, 'playlist data'),
 858                 playlist_id)
 859             playlist_medias = []
 860             for item in vxp_playlist:
 861                 media = item.get('media')
 862                 if not media:
 863                     continue
 864                 playlist_medias.append(media)
 865                 # Download single video if found media with asset id matching the video id from URL
 866                 if item.get('advert', {}).get('assetId') == playlist_id:
 867                     medias = [media]
 868                     break
 869             # Fallback to the whole playlist
 870             if not medias:
 871                 medias = playlist_medias
 872
 873         entries = []
 874         for num, media_meta in enumerate(medias, start=1):
 875             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 876             if not formats:
 877                 continue
 878             self._sort_formats(formats)
 879
 880             video_id = media_meta.get('externalId')
 881             if not video_id:
 882                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 883
 884             title = media_meta.get('caption')
 885             if not title:
 886                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 887
 888             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 889
 890             images = []
 891             for image in media_meta.get('images', {}).values():
 892                 images.extend(image.values())
 893             if 'image' in media_meta:
 894                 images.append(media_meta['image'])
 895
 896             thumbnails = [{
 897                 'url': image.get('href'),
 898                 'width': int_or_none(image.get('width')),
 899                 'height': int_or_none(image.get('height')),
 900             } for image in images]
 901
 902             entries.append({
 903                 'id': video_id,
 904                 'title': title,
 905                 'thumbnails': thumbnails,
 906                 'duration': duration,
 907                 'timestamp': timestamp,
 908                 'formats': formats,
 909                 'subtitles': subtitles,
 910             })
 911
 912         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 913
 914
 915 class BBCCoUkArticleIE(InfoExtractor):
 916     _VALID_URL = 'http://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 917     IE_NAME = 'bbc.co.uk:article'
 918     IE_DESC = 'BBC articles'
 919
 920     _TEST = {
 921         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 922         'info_dict': {
 923             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 924             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 925             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 926         },
 927         'playlist_count': 4,
 928         'add_ie': ['BBCCoUk'],
 929     }
 930
 931     def _real_extract(self, url):
 932         playlist_id = self._match_id(url)
 933
 934         webpage = self._download_webpage(url, playlist_id)
 935
 936         title = self._og_search_title(webpage)
 937         description = self._og_search_description(webpage).strip()
 938
 939         entries = [self.url_result(programme_url) for programme_url in re.findall(
 940             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 941
 942         return self.playlist_result(entries, playlist_id, title, description)