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