_ 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     # Unified Streaming Platform
  48     _USP_RE = r'/([^/]+)\.ism(?:\.hlsv2\.ism)?/[^/]+\.m3u8'
  49
  50     _NAMESPACES = (
  51         _MEDIASELECTION_NS,
  52         _EMP_PLAYLIST_NS,
  53     )
  54
  55     _TESTS = [
  56         {
  57             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  58             'info_dict': {
  59                 'id': 'b039d07m',
  60                 'ext': 'mp4',
  61                 'title': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
  62                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  63             },
  64             'params': {
  65                 'skip_download': True,
  66             }
  67         },
  68         {
  69             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  70             'info_dict': {
  71                 'id': 'b00yng1d',
  72                 'ext': 'flv',
  73                 'title': 'The Man in Black: Series 3: The Printed Name',
  74                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  75                 'duration': 1800,
  76             },
  77             'params': {
  78                 # rtmp download
  79                 'skip_download': True,
  80             },
  81             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  82         },
  83         {
  84             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  85             'info_dict': {
  86                 'id': 'b00yng1d',
  87                 'ext': 'flv',
  88                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
  89                 '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.',
  90                 'duration': 5100,
  91             },
  92             'params': {
  93                 # rtmp download
  94                 'skip_download': True,
  95             },
  96             'skip': 'this episode is not currently available',
  97         },
  98         {
  99             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
 100             'info_dict': {
 101                 'id': 'b03k3pb7',
 102                 'ext': 'flv',
 103                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
 104                 'description': '2. Invasion',
 105                 'duration': 3600,
 106             },
 107             'params': {
 108                 # rtmp download
 109                 'skip_download': True,
 110             },
 111             'skip': 'this episode is not currently available',
 112         }, {
 113             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 114             'info_dict': {
 115                 'id': 'b04v209v',
 116                 'ext': 'flv',
 117                 'title': 'Pete Tong, The Essential New Tune Special',
 118                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 119                 'duration': 10800,
 120             },
 121             'params': {
 122                 # rtmp download
 123                 'skip_download': True,
 124             },
 125             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 126         }, {
 127             'url': 'http://www.bbc.co.uk/music/clips/p022h44b',
 128             'note': 'Audio',
 129             'info_dict': {
 130                 'id': 'p022h44j',
 131                 'ext': 'mp4',
 132                 'title': 'BBC Proms Music Guides, Rachmaninov: Symphonic Dances',
 133                 'description': "In this Proms Music Guide, Andrew McGregor looks at Rachmaninov's Symphonic Dances.",
 134                 'duration': 227,
 135             },
 136             'params': {
 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': 'mp4',
 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                 'skip_download': True,
 151             }
 152         }, {
 153             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 154             'info_dict': {
 155                 'id': 'p02n76xf',
 156                 'ext': 'flv',
 157                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 158                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 159                 'duration': 3540,
 160             },
 161             'params': {
 162                 # rtmp download
 163                 'skip_download': True,
 164             },
 165             'skip': 'this episode is not currently available',
 166         }, {
 167             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 168             'info_dict': {
 169                 'id': 'b05zmgw1',
 170                 'ext': 'flv',
 171                 '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.',
 172                 'title': 'Royal Academy Summer Exhibition',
 173                 'duration': 3540,
 174             },
 175             'params': {
 176                 # rtmp download
 177                 'skip_download': True,
 178             },
 179             'skip': 'this episode is not currently available',
 180         }, {
 181             # iptv-all mediaset fails with geolocation however there is no geo restriction
 182             # for this programme at all
 183             'url': 'http://www.bbc.co.uk/programmes/b06rkn85',
 184             'info_dict': {
 185                 'id': 'b06rkms3',
 186                 'ext': 'flv',
 187                 'title': "Best of the Mini-Mixes 2015: Part 3, Annie Mac's Friday Night - BBC Radio 1",
 188                 'description': "Annie has part three in the Best of the Mini-Mixes 2015, plus the year's Most Played!",
 189             },
 190             'params': {
 191                 # rtmp download
 192                 'skip_download': True,
 193             },
 194             'skip': 'this episode is not currently available on BBC iPlayer Radio',
 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': 'mp4',
 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                 'skip_download': True,
 206             },
 207         }, {
 208             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 209             'only_matching': True,
 210         }, {
 211             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 212             'only_matching': True,
 213         }, {
 214             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 215             'only_matching': True,
 216         }, {
 217             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
 218             'only_matching': True,
 219         }
 220     ]
 221
 222     class MediaSelectionError(Exception):
 223         def __init__(self, id):
 224             self.id = id
 225
 226     def _extract_asx_playlist(self, connection, programme_id):
 227         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 228         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 229
 230     def _extract_connection(self, connection, programme_id):
 231         formats = []
 232         kind = connection.get('kind')
 233         protocol = connection.get('protocol')
 234         supplier = connection.get('supplier')
 235         if protocol == 'http':
 236             href = connection.get('href')
 237             transfer_format = connection.get('transferFormat')
 238             # ASX playlist
 239             if supplier == 'asx':
 240                 for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 241                     formats.append({
 242                         'url': ref,
 243                         'format_id': 'ref%s_%s' % (i, supplier),
 244                     })
 245             # Skip DASH until supported
 246             elif transfer_format == 'dash':
 247                 pass
 248             elif transfer_format == 'hls':
 249                 is_unified_streaming = re.search(self._USP_RE, href)
 250                 if is_unified_streaming:
 251                     href = re.sub(self._USP_RE, r'/\1.ism/\1.m3u8', href)
 252                 m3u8_formats = self._extract_m3u8_formats(
 253                     href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 254                     m3u8_id=supplier, fatal=False)
 255                 if is_unified_streaming:
 256                     self._check_formats(m3u8_formats, programme_id)
 257                 formats.extend(m3u8_formats)
 258             # Direct link
 259             else:
 260                 formats.append({
 261                     'url': href,
 262                     'format_id': supplier or kind or protocol,
 263                 })
 264         elif protocol == 'rtmp':
 265             application = connection.get('application', 'ondemand')
 266             auth_string = connection.get('authString')
 267             identifier = connection.get('identifier')
 268             server = connection.get('server')
 269             formats.append({
 270                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 271                 'play_path': identifier,
 272                 'app': '%s?%s' % (application, auth_string),
 273                 'page_url': 'http://www.bbc.co.uk',
 274                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 275                 'rtmp_live': False,
 276                 'ext': 'flv',
 277                 'format_id': supplier,
 278             })
 279         return formats
 280
 281     def _extract_items(self, playlist):
 282         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 283
 284     def _findall_ns(self, element, xpath):
 285         elements = []
 286         for ns in self._NAMESPACES:
 287             elements.extend(element.findall(xpath % ns))
 288         return elements
 289
 290     def _extract_medias(self, media_selection):
 291         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 292         if error is None:
 293             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 294         if error is not None:
 295             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 296         return self._findall_ns(media_selection, './{%s}media')
 297
 298     def _extract_connections(self, media):
 299         return self._findall_ns(media, './{%s}connection')
 300
 301     def _extract_video(self, media, programme_id):
 302         formats = []
 303         vbr = int_or_none(media.get('bitrate'))
 304         vcodec = media.get('encoding')
 305         service = media.get('service')
 306         width = int_or_none(media.get('width'))
 307         height = int_or_none(media.get('height'))
 308         file_size = int_or_none(media.get('media_file_size'))
 309         for connection in self._extract_connections(media):
 310             conn_formats = self._extract_connection(connection, programme_id)
 311             for format in conn_formats:
 312                 if format.get('protocol') != 'm3u8_native':
 313                     format.update({
 314                         'width': width,
 315                         'height': height,
 316                         'vbr': vbr,
 317                         'vcodec': vcodec,
 318                         'filesize': file_size,
 319                     })
 320                 if service:
 321                     format['format_id'] = '%s_%s' % (service, format['format_id'])
 322             formats.extend(conn_formats)
 323         return formats
 324
 325     def _extract_audio(self, media, programme_id):
 326         formats = []
 327         abr = int_or_none(media.get('bitrate'))
 328         acodec = media.get('encoding')
 329         service = media.get('service')
 330         for connection in self._extract_connections(media):
 331             conn_formats = self._extract_connection(connection, programme_id)
 332             for format in conn_formats:
 333                 format.update({
 334                     'format_id': '%s_%s' % (service, format['format_id']),
 335                     'abr': abr,
 336                     'acodec': acodec,
 337                     'vcodec': 'none',
 338                 })
 339             formats.extend(conn_formats)
 340         return formats
 341
 342     def _get_subtitles(self, media, programme_id):
 343         subtitles = {}
 344         for connection in self._extract_connections(media):
 345             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
 346             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 347             subtitles[lang] = [
 348                 {
 349                     'url': connection.get('href'),
 350                     'ext': 'ttml',
 351                 },
 352             ]
 353         return subtitles
 354
 355     def _raise_extractor_error(self, media_selection_error):
 356         raise ExtractorError(
 357             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 358             expected=True)
 359
 360     def _download_media_selector(self, programme_id):
 361         last_exception = None
 362         for mediaselector_url in self._MEDIASELECTOR_URLS:
 363             try:
 364                 return self._download_media_selector_url(
 365                     mediaselector_url % programme_id, programme_id)
 366             except BBCCoUkIE.MediaSelectionError as e:
 367                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 368                     last_exception = e
 369                     continue
 370                 self._raise_extractor_error(e)
 371         self._raise_extractor_error(last_exception)
 372
 373     def _download_media_selector_url(self, url, programme_id=None):
 374         try:
 375             media_selection = self._download_xml(
 376                 url, programme_id, 'Downloading media selection XML')
 377         except ExtractorError as ee:
 378             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
 379                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
 380             else:
 381                 raise
 382         return self._process_media_selector(media_selection, programme_id)
 383
 384     def _process_media_selector(self, media_selection, programme_id):
 385         formats = []
 386         subtitles = None
 387
 388         for media in self._extract_medias(media_selection):
 389             kind = media.get('kind')
 390             if kind == 'audio':
 391                 formats.extend(self._extract_audio(media, programme_id))
 392             elif kind == 'video':
 393                 formats.extend(self._extract_video(media, programme_id))
 394             elif kind == 'captions':
 395                 subtitles = self.extract_subtitles(media, programme_id)
 396         return formats, subtitles
 397
 398     def _download_playlist(self, playlist_id):
 399         try:
 400             playlist = self._download_json(
 401                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 402                 playlist_id, 'Downloading playlist JSON')
 403
 404             version = playlist.get('defaultAvailableVersion')
 405             if version:
 406                 smp_config = version['smpConfig']
 407                 title = smp_config['title']
 408                 description = smp_config['summary']
 409                 for item in smp_config['items']:
 410                     kind = item['kind']
 411                     if kind != 'programme' and kind != 'radioProgramme':
 412                         continue
 413                     programme_id = item.get('vpid')
 414                     duration = int_or_none(item.get('duration'))
 415                     formats, subtitles = self._download_media_selector(programme_id)
 416                 return programme_id, title, description, duration, formats, subtitles
 417         except ExtractorError as ee:
 418             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 419                 raise
 420
 421         # fallback to legacy playlist
 422         return self._process_legacy_playlist(playlist_id)
 423
 424     def _process_legacy_playlist_url(self, url, display_id):
 425         playlist = self._download_legacy_playlist_url(url, display_id)
 426         return self._extract_from_legacy_playlist(playlist, display_id)
 427
 428     def _process_legacy_playlist(self, playlist_id):
 429         return self._process_legacy_playlist_url(
 430             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 431
 432     def _download_legacy_playlist_url(self, url, playlist_id=None):
 433         return self._download_xml(
 434             url, playlist_id, 'Downloading legacy playlist XML')
 435
 436     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 437         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 438         if no_items is not None:
 439             reason = no_items.get('reason')
 440             if reason == 'preAvailability':
 441                 msg = 'Episode %s is not yet available' % playlist_id
 442             elif reason == 'postAvailability':
 443                 msg = 'Episode %s is no longer available' % playlist_id
 444             elif reason == 'noMedia':
 445                 msg = 'Episode %s is not currently available' % playlist_id
 446             else:
 447                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 448             raise ExtractorError(msg, expected=True)
 449
 450         for item in self._extract_items(playlist):
 451             kind = item.get('kind')
 452             if kind != 'programme' and kind != 'radioProgramme':
 453                 continue
 454             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 455             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 456             description = description_el.text if description_el is not None else None
 457
 458             def get_programme_id(item):
 459                 def get_from_attributes(item):
 460                     for p in('identifier', 'group'):
 461                         value = item.get(p)
 462                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 463                             return value
 464                 get_from_attributes(item)
 465                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 466                 if mediator is not None:
 467                     return get_from_attributes(mediator)
 468
 469             programme_id = get_programme_id(item)
 470             duration = int_or_none(item.get('duration'))
 471
 472             if programme_id:
 473                 formats, subtitles = self._download_media_selector(programme_id)
 474             else:
 475                 formats, subtitles = self._process_media_selector(item, playlist_id)
 476                 programme_id = playlist_id
 477
 478         return programme_id, title, description, duration, formats, subtitles
 479
 480     def _real_extract(self, url):
 481         group_id = self._match_id(url)
 482
 483         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 484
 485         programme_id = None
 486         duration = None
 487
 488         tviplayer = self._search_regex(
 489             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 490             webpage, 'player', default=None)
 491
 492         if tviplayer:
 493             player = self._parse_json(tviplayer, group_id).get('player', {})
 494             duration = int_or_none(player.get('duration'))
 495             programme_id = player.get('vpid')
 496
 497         if not programme_id:
 498             programme_id = self._search_regex(
 499                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 500
 501         if programme_id:
 502             formats, subtitles = self._download_media_selector(programme_id)
 503             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 504                 (r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>',
 505                  r'<div[^>]+class="info"[^>]*>\s*<h1>(.+?)</h1>'), webpage, 'title')
 506             description = self._search_regex(
 507                 (r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 508                  r'<div[^>]+class="info_+synopsis"[^>]*>([^<]+)</div>'),
 509                 webpage, 'description', default=None)
 510             if not description:
 511                 description = self._html_search_meta('description', webpage)
 512         else:
 513             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 514
 515         self._sort_formats(formats)
 516
 517         return {
 518             'id': programme_id,
 519             'title': title,
 520             'description': description,
 521             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 522             'duration': duration,
 523             'formats': formats,
 524             'subtitles': subtitles,
 525         }
 526
 527
 528 class BBCIE(BBCCoUkIE):
 529     IE_NAME = 'bbc'
 530     IE_DESC = 'BBC'
 531     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 532
 533     _MEDIASELECTOR_URLS = [
 534         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 535         # even not geo restricted at all
 536         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 537         # Provides more formats, namely direct mp4 links, but fails on some videos with
 538         # notukerror for non UK (?) users (e.g.
 539         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 540         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 541         # Provides fewer formats, but works everywhere for everybody (hopefully)
 542         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 543     ]
 544
 545     _TESTS = [{
 546         # article with multiple videos embedded with data-playable containing vpids
 547         'url': 'http://www.bbc.com/news/world-europe-32668511',
 548         'info_dict': {
 549             'id': 'world-europe-32668511',
 550             'title': 'Russia stages massive WW2 parade despite Western boycott',
 551             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 552         },
 553         'playlist_count': 2,
 554     }, {
 555         # article with multiple videos embedded with data-playable (more videos)
 556         'url': 'http://www.bbc.com/news/business-28299555',
 557         'info_dict': {
 558             'id': 'business-28299555',
 559             'title': 'Farnborough Airshow: Video highlights',
 560             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 561         },
 562         'playlist_count': 9,
 563         'skip': 'Save time',
 564     }, {
 565         # article with multiple videos embedded with `new SMP()`
 566         # broken
 567         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 568         'info_dict': {
 569             'id': '3662a707-0af9-3149-963f-47bea720b460',
 570             'title': 'BUGGER',
 571         },
 572         'playlist_count': 18,
 573     }, {
 574         # single video embedded with data-playable containing vpid
 575         'url': 'http://www.bbc.com/news/world-europe-32041533',
 576         'info_dict': {
 577             'id': 'p02mprgb',
 578             'ext': 'mp4',
 579             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 580             'description': 'md5:2868290467291b37feda7863f7a83f54',
 581             'duration': 47,
 582             'timestamp': 1427219242,
 583             'upload_date': '20150324',
 584         },
 585         'params': {
 586             # rtmp download
 587             'skip_download': True,
 588         }
 589     }, {
 590         # article with single video embedded with data-playable containing XML playlist
 591         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 592         # and playlist with f4m and m3u8 as streamingUrl
 593         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 594         'info_dict': {
 595             'id': '150615_telabyad_kentin_cogu',
 596             'ext': 'mp4',
 597             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 598             'timestamp': 1434397334,
 599             'upload_date': '20150615',
 600         },
 601         'params': {
 602             'skip_download': True,
 603         }
 604     }, {
 605         # single video embedded with data-playable containing XML playlists (regional section)
 606         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 607         'info_dict': {
 608             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 609             'ext': 'mp4',
 610             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 611             'timestamp': 1434713142,
 612             'upload_date': '20150619',
 613         },
 614         'params': {
 615             'skip_download': True,
 616         }
 617     }, {
 618         # single video from video playlist embedded with vxp-playlist-data JSON
 619         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 620         'info_dict': {
 621             'id': 'p02w6qjc',
 622             'ext': 'mp4',
 623             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 624             'duration': 56,
 625             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 626         },
 627         'params': {
 628             'skip_download': True,
 629         }
 630     }, {
 631         # single video story with digitalData
 632         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 633         'info_dict': {
 634             'id': 'p02q6gc4',
 635             'ext': 'flv',
 636             'title': 'Sri Lanka’s spicy secret',
 637             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 638             'timestamp': 1437674293,
 639             'upload_date': '20150723',
 640         },
 641         'params': {
 642             # rtmp download
 643             'skip_download': True,
 644         }
 645     }, {
 646         # single video story without digitalData
 647         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 648         'info_dict': {
 649             'id': 'p018zqqg',
 650             'ext': 'mp4',
 651             'title': 'Hyundai Santa Fe Sport: Rock star',
 652             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 653             'timestamp': 1415867444,
 654             'upload_date': '20141113',
 655         },
 656         'params': {
 657             # rtmp download
 658             'skip_download': True,
 659         }
 660     }, {
 661         # single video with playlist.sxml URL in playlist param
 662         'url': 'http://www.bbc.com/sport/0/football/33653409',
 663         'info_dict': {
 664             'id': 'p02xycnp',
 665             'ext': 'mp4',
 666             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 667             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 668             'duration': 140,
 669         },
 670         'params': {
 671             # rtmp download
 672             'skip_download': True,
 673         }
 674     }, {
 675         # article with multiple videos embedded with playlist.sxml in playlist param
 676         'url': 'http://www.bbc.com/sport/0/football/34475836',
 677         'info_dict': {
 678             'id': '34475836',
 679             'title': 'Jurgen Klopp: Furious football from a witty and winning coach',
 680             'description': 'Fast-paced football, wit, wisdom and a ready smile - why Liverpool fans should come to love new boss Jurgen Klopp.',
 681         },
 682         'playlist_count': 3,
 683     }, {
 684         # school report article with single video
 685         'url': 'http://www.bbc.co.uk/schoolreport/35744779',
 686         'info_dict': {
 687             'id': '35744779',
 688             'title': 'School which breaks down barriers in Jerusalem',
 689         },
 690         'playlist_count': 1,
 691     }, {
 692         # single video with playlist URL from weather section
 693         'url': 'http://www.bbc.com/weather/features/33601775',
 694         'only_matching': True,
 695     }, {
 696         # custom redirection to www.bbc.com
 697         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 698         'only_matching': True,
 699     }, {
 700         # single video article embedded with data-media-vpid
 701         'url': 'http://www.bbc.co.uk/sport/rowing/35908187',
 702         'only_matching': True,
 703     }]
 704
 705     @classmethod
 706     def suitable(cls, url):
 707         EXCLUDE_IE = (BBCCoUkIE, BBCCoUkArticleIE, BBCCoUkIPlayerPlaylistIE, BBCCoUkPlaylistIE)
 708         return (False if any(ie.suitable(url) for ie in EXCLUDE_IE)
 709                 else super(BBCIE, cls).suitable(url))
 710
 711     def _extract_from_media_meta(self, media_meta, video_id):
 712         # Direct links to media in media metadata (e.g.
 713         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 714         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 715         source_files = media_meta.get('sourceFiles')
 716         if source_files:
 717             return [{
 718                 'url': f['url'],
 719                 'format_id': format_id,
 720                 'ext': f.get('encoding'),
 721                 'tbr': float_or_none(f.get('bitrate'), 1000),
 722                 'filesize': int_or_none(f.get('filesize')),
 723             } for format_id, f in source_files.items() if f.get('url')], []
 724
 725         programme_id = media_meta.get('externalId')
 726         if programme_id:
 727             return self._download_media_selector(programme_id)
 728
 729         # Process playlist.sxml as legacy playlist
 730         href = media_meta.get('href')
 731         if href:
 732             playlist = self._download_legacy_playlist_url(href)
 733             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 734             return formats, subtitles
 735
 736         return [], []
 737
 738     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 739         programme_id, title, description, duration, formats, subtitles = \
 740             self._process_legacy_playlist_url(url, playlist_id)
 741         self._sort_formats(formats)
 742         return {
 743             'id': programme_id,
 744             'title': title,
 745             'description': description,
 746             'duration': duration,
 747             'timestamp': timestamp,
 748             'formats': formats,
 749             'subtitles': subtitles,
 750         }
 751
 752     def _real_extract(self, url):
 753         playlist_id = self._match_id(url)
 754
 755         webpage = self._download_webpage(url, playlist_id)
 756
 757         json_ld_info = self._search_json_ld(webpage, playlist_id, default=None)
 758         timestamp = json_ld_info.get('timestamp')
 759
 760         playlist_title = json_ld_info.get('title')
 761         if not playlist_title:
 762             playlist_title = self._og_search_title(
 763                 webpage, default=None) or self._html_search_regex(
 764                 r'<title>(.+?)</title>', webpage, 'playlist title', default=None)
 765             if playlist_title:
 766                 playlist_title = re.sub(r'(.+)\s*-\s*BBC.*?$', r'\1', playlist_title).strip()
 767
 768         playlist_description = json_ld_info.get(
 769             'description') or self._og_search_description(webpage, default=None)
 770
 771         if not timestamp:
 772             timestamp = parse_iso8601(self._search_regex(
 773                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 774                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 775                  r'"datePublished":\s*"([^"]+)'],
 776                 webpage, 'date', default=None))
 777
 778         entries = []
 779
 780         # article with multiple videos embedded with playlist.sxml (e.g.
 781         # http://www.bbc.com/sport/0/football/34475836)
 782         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 783         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 784         if playlists:
 785             entries = [
 786                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 787                 for playlist_url in playlists]
 788
 789         # news article with multiple videos embedded with data-playable
 790         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 791         if data_playables:
 792             for _, data_playable_json in data_playables:
 793                 data_playable = self._parse_json(
 794                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 795                 if not data_playable:
 796                     continue
 797                 settings = data_playable.get('settings', {})
 798                 if settings:
 799                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 800                     # http://www.bbc.com/news/world-us-canada-34473351)
 801                     playlist_object = settings.get('playlistObject', {})
 802                     if playlist_object:
 803                         items = playlist_object.get('items')
 804                         if items and isinstance(items, list):
 805                             title = playlist_object['title']
 806                             description = playlist_object.get('summary')
 807                             duration = int_or_none(items[0].get('duration'))
 808                             programme_id = items[0].get('vpid')
 809                             formats, subtitles = self._download_media_selector(programme_id)
 810                             self._sort_formats(formats)
 811                             entries.append({
 812                                 'id': programme_id,
 813                                 'title': title,
 814                                 'description': description,
 815                                 'timestamp': timestamp,
 816                                 'duration': duration,
 817                                 'formats': formats,
 818                                 'subtitles': subtitles,
 819                             })
 820                     else:
 821                         # data-playable without vpid but with a playlist.sxml URLs
 822                         # in otherSettings.playlist (e.g.
 823                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 824                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 825                         if playlist:
 826                             entries.append(self._extract_from_playlist_sxml(
 827                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
 828
 829         if entries:
 830             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 831
 832         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 833         programme_id = self._search_regex(
 834             [r'data-(?:video-player|media)-vpid="(%s)"' % self._ID_REGEX,
 835              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 836              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 837             webpage, 'vpid', default=None)
 838
 839         if programme_id:
 840             formats, subtitles = self._download_media_selector(programme_id)
 841             self._sort_formats(formats)
 842             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 843             digital_data = self._parse_json(
 844                 self._search_regex(
 845                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 846                 programme_id, fatal=False)
 847             page_info = digital_data.get('page', {}).get('pageInfo', {})
 848             title = page_info.get('pageName') or self._og_search_title(webpage)
 849             description = page_info.get('description') or self._og_search_description(webpage)
 850             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 851             return {
 852                 'id': programme_id,
 853                 'title': title,
 854                 'description': description,
 855                 'timestamp': timestamp,
 856                 'formats': formats,
 857                 'subtitles': subtitles,
 858             }
 859
 860         def extract_all(pattern):
 861             return list(filter(None, map(
 862                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 863                 re.findall(pattern, webpage))))
 864
 865         # Multiple video article (e.g.
 866         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 867         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 868         entries = []
 869         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 870             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 871             if embed_url and re.match(EMBED_URL, embed_url):
 872                 entries.append(embed_url)
 873         entries.extend(re.findall(
 874             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 875         if entries:
 876             return self.playlist_result(
 877                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 878                 playlist_id, playlist_title, playlist_description)
 879
 880         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 881         medias = extract_all(r"data-media-meta='({[^']+})'")
 882
 883         if not medias:
 884             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 885             media_asset = self._search_regex(
 886                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 887                 webpage, 'media asset', default=None)
 888             if media_asset:
 889                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 890                 medias = []
 891                 for video in media_asset_page.get('videos', {}).values():
 892                     medias.extend(video.values())
 893
 894         if not medias:
 895             # Multiple video playlist with single `now playing` entry (e.g.
 896             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 897             vxp_playlist = self._parse_json(
 898                 self._search_regex(
 899                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 900                     webpage, 'playlist data'),
 901                 playlist_id)
 902             playlist_medias = []
 903             for item in vxp_playlist:
 904                 media = item.get('media')
 905                 if not media:
 906                     continue
 907                 playlist_medias.append(media)
 908                 # Download single video if found media with asset id matching the video id from URL
 909                 if item.get('advert', {}).get('assetId') == playlist_id:
 910                     medias = [media]
 911                     break
 912             # Fallback to the whole playlist
 913             if not medias:
 914                 medias = playlist_medias
 915
 916         entries = []
 917         for num, media_meta in enumerate(medias, start=1):
 918             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 919             if not formats:
 920                 continue
 921             self._sort_formats(formats)
 922
 923             video_id = media_meta.get('externalId')
 924             if not video_id:
 925                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 926
 927             title = media_meta.get('caption')
 928             if not title:
 929                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 930
 931             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 932
 933             images = []
 934             for image in media_meta.get('images', {}).values():
 935                 images.extend(image.values())
 936             if 'image' in media_meta:
 937                 images.append(media_meta['image'])
 938
 939             thumbnails = [{
 940                 'url': image.get('href'),
 941                 'width': int_or_none(image.get('width')),
 942                 'height': int_or_none(image.get('height')),
 943             } for image in images]
 944
 945             entries.append({
 946                 'id': video_id,
 947                 'title': title,
 948                 'thumbnails': thumbnails,
 949                 'duration': duration,
 950                 'timestamp': timestamp,
 951                 'formats': formats,
 952                 'subtitles': subtitles,
 953             })
 954
 955         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 956
 957
 958 class BBCCoUkArticleIE(InfoExtractor):
 959     _VALID_URL = r'https?://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 960     IE_NAME = 'bbc.co.uk:article'
 961     IE_DESC = 'BBC articles'
 962
 963     _TEST = {
 964         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 965         'info_dict': {
 966             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 967             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 968             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 969         },
 970         'playlist_count': 4,
 971         'add_ie': ['BBCCoUk'],
 972     }
 973
 974     def _real_extract(self, url):
 975         playlist_id = self._match_id(url)
 976
 977         webpage = self._download_webpage(url, playlist_id)
 978
 979         title = self._og_search_title(webpage)
 980         description = self._og_search_description(webpage).strip()
 981
 982         entries = [self.url_result(programme_url) for programme_url in re.findall(
 983             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 984
 985         return self.playlist_result(entries, playlist_id, title, description)
 986
 987
 988 class BBCCoUkPlaylistBaseIE(InfoExtractor):
 989     def _real_extract(self, url):
 990         playlist_id = self._match_id(url)
 991
 992         webpage = self._download_webpage(url, playlist_id)
 993
 994         entries = [
 995             self.url_result(self._URL_TEMPLATE % video_id, BBCCoUkIE.ie_key())
 996             for video_id in re.findall(
 997                 self._VIDEO_ID_TEMPLATE % BBCCoUkIE._ID_REGEX, webpage)]
 998
 999         title, description = self._extract_title_and_description(webpage)
1000
1001         return self.playlist_result(entries, playlist_id, title, description)
1002
1003
1004 class BBCCoUkIPlayerPlaylistIE(BBCCoUkPlaylistBaseIE):
1005     IE_NAME = 'bbc.co.uk:iplayer:playlist'
1006     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/iplayer/episodes/(?P<id>%s)' % BBCCoUkIE._ID_REGEX
1007     _URL_TEMPLATE = 'http://www.bbc.co.uk/iplayer/episode/%s'
1008     _VIDEO_ID_TEMPLATE = r'data-ip-id=["\'](%s)'
1009     _TEST = {
1010         'url': 'http://www.bbc.co.uk/iplayer/episodes/b05rcz9v',
1011         'info_dict': {
1012             'id': 'b05rcz9v',
1013             'title': 'The Disappearance',
1014             'description': 'French thriller serial about a missing teenager.',
1015         },
1016         'playlist_mincount': 6,
1017     }
1018
1019     def _extract_title_and_description(self, webpage):
1020         title = self._search_regex(r'<h1>([^<]+)</h1>', webpage, 'title', fatal=False)
1021         description = self._search_regex(
1022             r'<p[^>]+class=(["\'])subtitle\1[^>]*>(?P<value>[^<]+)</p>',
1023             webpage, 'description', fatal=False, group='value')
1024         return title, description
1025
1026
1027 class BBCCoUkPlaylistIE(BBCCoUkPlaylistBaseIE):
1028     IE_NAME = 'bbc.co.uk:playlist'
1029     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/(?P<id>%s)/(?:episodes|broadcasts|clips)' % BBCCoUkIE._ID_REGEX
1030     _URL_TEMPLATE = 'http://www.bbc.co.uk/programmes/%s'
1031     _VIDEO_ID_TEMPLATE = r'data-pid=["\'](%s)'
1032     _TESTS = [{
1033         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1034         'info_dict': {
1035             'id': 'b05rcz9v',
1036             'title': 'The Disappearance - Clips - BBC Four',
1037             'description': 'French thriller serial about a missing teenager.',
1038         },
1039         'playlist_mincount': 7,
1040     }, {
1041         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/broadcasts/2016/06',
1042         'only_matching': True,
1043     }, {
1044         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1045         'only_matching': True,
1046     }, {
1047         'url': 'http://www.bbc.co.uk/programmes/b055jkys/episodes/player',
1048         'only_matching': True,
1049     }]
1050
1051     def _extract_title_and_description(self, webpage):
1052         title = self._og_search_title(webpage, fatal=False)
1053         description = self._og_search_description(webpage)
1054         return title, description