]> git.bitcoin.ninja Git - youtube-dl/blob - youtube_dl/extractor/bbc.py
[README.md] Make download URLs consistent
[youtube-dl] / youtube_dl / extractor / bbc.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .common import InfoExtractor
7 from ..utils import (
8     ExtractorError,
9     float_or_none,
10     int_or_none,
11     parse_duration,
12     parse_iso8601,
13     unescapeHTML,
14 )
15 from ..compat import (
16     compat_etree_fromstring,
17     compat_HTTPError,
18 )
19
20
21 class BBCCoUkIE(InfoExtractor):
22     IE_NAME = 'bbc.co.uk'
23     IE_DESC = 'BBC iPlayer'
24     _ID_REGEX = r'[pb][\da-z]{7}'
25     _VALID_URL = r'''(?x)
26                     https?://
27                         (?:www\.)?bbc\.co\.uk/
28                         (?:
29                             programmes/(?!articles/)|
30                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
31                             music/clips[/#]|
32                             radio/player/
33                         )
34                         (?P<id>%s)(?!/(?:episodes|broadcasts|clips))
35                     ''' % _ID_REGEX
36
37     _MEDIASELECTOR_URLS = [
38         # Provides HQ HLS streams with even better quality that pc mediaset but fails
39         # with geolocation in some cases when it's even not geo restricted at all (e.g.
40         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
41         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
42         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
43     ]
44
45     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
46     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
47
48     _NAMESPACES = (
49         _MEDIASELECTION_NS,
50         _EMP_PLAYLIST_NS,
51     )
52
53     _TESTS = [
54         {
55             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
56             'info_dict': {
57                 'id': 'b039d07m',
58                 'ext': '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             'skip': 'Now it\'s really geo-restricted',
196         }, {
197             # compact player (https://github.com/rg3/youtube-dl/issues/8147)
198             'url': 'http://www.bbc.co.uk/programmes/p028bfkf/player',
199             'info_dict': {
200                 'id': 'p028bfkj',
201                 'ext': 'flv',
202                 'title': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
203                 'description': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
204             },
205             'params': {
206                 # rtmp download
207                 'skip_download': True,
208             },
209         }, {
210             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
211             'only_matching': True,
212         }, {
213             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
214             'only_matching': True,
215         }, {
216             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
217             'only_matching': True,
218         }, {
219             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
220             'only_matching': True,
221         }
222     ]
223
224     class MediaSelectionError(Exception):
225         def __init__(self, id):
226             self.id = id
227
228     def _extract_asx_playlist(self, connection, programme_id):
229         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
230         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
231
232     def _extract_connection(self, connection, programme_id):
233         formats = []
234         kind = connection.get('kind')
235         protocol = connection.get('protocol')
236         supplier = connection.get('supplier')
237         if protocol == 'http':
238             href = connection.get('href')
239             transfer_format = connection.get('transferFormat')
240             # ASX playlist
241             if supplier == 'asx':
242                 for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
243                     formats.append({
244                         'url': ref,
245                         'format_id': 'ref%s_%s' % (i, supplier),
246                     })
247             # Skip DASH until supported
248             elif transfer_format == 'dash':
249                 pass
250             elif transfer_format == 'hls':
251                 formats.extend(self._extract_m3u8_formats(
252                     href, programme_id, ext='mp4', entry_protocol='m3u8_native',
253                     m3u8_id=supplier, fatal=False))
254             # Direct link
255             else:
256                 formats.append({
257                     'url': href,
258                     'format_id': supplier or kind or protocol,
259                 })
260         elif protocol == 'rtmp':
261             application = connection.get('application', 'ondemand')
262             auth_string = connection.get('authString')
263             identifier = connection.get('identifier')
264             server = connection.get('server')
265             formats.append({
266                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
267                 'play_path': identifier,
268                 'app': '%s?%s' % (application, auth_string),
269                 'page_url': 'http://www.bbc.co.uk',
270                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
271                 'rtmp_live': False,
272                 'ext': 'flv',
273                 'format_id': supplier,
274             })
275         return formats
276
277     def _extract_items(self, playlist):
278         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
279
280     def _findall_ns(self, element, xpath):
281         elements = []
282         for ns in self._NAMESPACES:
283             elements.extend(element.findall(xpath % ns))
284         return elements
285
286     def _extract_medias(self, media_selection):
287         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
288         if error is None:
289             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
290         if error is not None:
291             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
292         return self._findall_ns(media_selection, './{%s}media')
293
294     def _extract_connections(self, media):
295         return self._findall_ns(media, './{%s}connection')
296
297     def _extract_video(self, media, programme_id):
298         formats = []
299         vbr = int_or_none(media.get('bitrate'))
300         vcodec = media.get('encoding')
301         service = media.get('service')
302         width = int_or_none(media.get('width'))
303         height = int_or_none(media.get('height'))
304         file_size = int_or_none(media.get('media_file_size'))
305         for connection in self._extract_connections(media):
306             conn_formats = self._extract_connection(connection, programme_id)
307             for format in conn_formats:
308                 format.update({
309                     'width': width,
310                     'height': height,
311                     'vbr': vbr,
312                     'vcodec': vcodec,
313                     'filesize': file_size,
314                 })
315                 if service:
316                     format['format_id'] = '%s_%s' % (service, format['format_id'])
317             formats.extend(conn_formats)
318         return formats
319
320     def _extract_audio(self, media, programme_id):
321         formats = []
322         abr = int_or_none(media.get('bitrate'))
323         acodec = media.get('encoding')
324         service = media.get('service')
325         for connection in self._extract_connections(media):
326             conn_formats = self._extract_connection(connection, programme_id)
327             for format in conn_formats:
328                 format.update({
329                     'format_id': '%s_%s' % (service, format['format_id']),
330                     'abr': abr,
331                     'acodec': acodec,
332                     'vcodec': 'none',
333                 })
334             formats.extend(conn_formats)
335         return formats
336
337     def _get_subtitles(self, media, programme_id):
338         subtitles = {}
339         for connection in self._extract_connections(media):
340             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
341             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
342             subtitles[lang] = [
343                 {
344                     'url': connection.get('href'),
345                     'ext': 'ttml',
346                 },
347             ]
348         return subtitles
349
350     def _raise_extractor_error(self, media_selection_error):
351         raise ExtractorError(
352             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
353             expected=True)
354
355     def _download_media_selector(self, programme_id):
356         last_exception = None
357         for mediaselector_url in self._MEDIASELECTOR_URLS:
358             try:
359                 return self._download_media_selector_url(
360                     mediaselector_url % programme_id, programme_id)
361             except BBCCoUkIE.MediaSelectionError as e:
362                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
363                     last_exception = e
364                     continue
365                 self._raise_extractor_error(e)
366         self._raise_extractor_error(last_exception)
367
368     def _download_media_selector_url(self, url, programme_id=None):
369         try:
370             media_selection = self._download_xml(
371                 url, programme_id, 'Downloading media selection XML')
372         except ExtractorError as ee:
373             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
374                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
375             else:
376                 raise
377         return self._process_media_selector(media_selection, programme_id)
378
379     def _process_media_selector(self, media_selection, programme_id):
380         formats = []
381         subtitles = None
382
383         for media in self._extract_medias(media_selection):
384             kind = media.get('kind')
385             if kind == 'audio':
386                 formats.extend(self._extract_audio(media, programme_id))
387             elif kind == 'video':
388                 formats.extend(self._extract_video(media, programme_id))
389             elif kind == 'captions':
390                 subtitles = self.extract_subtitles(media, programme_id)
391         return formats, subtitles
392
393     def _download_playlist(self, playlist_id):
394         try:
395             playlist = self._download_json(
396                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
397                 playlist_id, 'Downloading playlist JSON')
398
399             version = playlist.get('defaultAvailableVersion')
400             if version:
401                 smp_config = version['smpConfig']
402                 title = smp_config['title']
403                 description = smp_config['summary']
404                 for item in smp_config['items']:
405                     kind = item['kind']
406                     if kind != 'programme' and kind != 'radioProgramme':
407                         continue
408                     programme_id = item.get('vpid')
409                     duration = int_or_none(item.get('duration'))
410                     formats, subtitles = self._download_media_selector(programme_id)
411                 return programme_id, title, description, duration, formats, subtitles
412         except ExtractorError as ee:
413             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
414                 raise
415
416         # fallback to legacy playlist
417         return self._process_legacy_playlist(playlist_id)
418
419     def _process_legacy_playlist_url(self, url, display_id):
420         playlist = self._download_legacy_playlist_url(url, display_id)
421         return self._extract_from_legacy_playlist(playlist, display_id)
422
423     def _process_legacy_playlist(self, playlist_id):
424         return self._process_legacy_playlist_url(
425             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
426
427     def _download_legacy_playlist_url(self, url, playlist_id=None):
428         return self._download_xml(
429             url, playlist_id, 'Downloading legacy playlist XML')
430
431     def _extract_from_legacy_playlist(self, playlist, playlist_id):
432         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
433         if no_items is not None:
434             reason = no_items.get('reason')
435             if reason == 'preAvailability':
436                 msg = 'Episode %s is not yet available' % playlist_id
437             elif reason == 'postAvailability':
438                 msg = 'Episode %s is no longer available' % playlist_id
439             elif reason == 'noMedia':
440                 msg = 'Episode %s is not currently available' % playlist_id
441             else:
442                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
443             raise ExtractorError(msg, expected=True)
444
445         for item in self._extract_items(playlist):
446             kind = item.get('kind')
447             if kind != 'programme' and kind != 'radioProgramme':
448                 continue
449             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
450             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
451             description = description_el.text if description_el is not None else None
452
453             def get_programme_id(item):
454                 def get_from_attributes(item):
455                     for p in('identifier', 'group'):
456                         value = item.get(p)
457                         if value and re.match(r'^[pb][\da-z]{7}$', value):
458                             return value
459                 get_from_attributes(item)
460                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
461                 if mediator is not None:
462                     return get_from_attributes(mediator)
463
464             programme_id = get_programme_id(item)
465             duration = int_or_none(item.get('duration'))
466
467             if programme_id:
468                 formats, subtitles = self._download_media_selector(programme_id)
469             else:
470                 formats, subtitles = self._process_media_selector(item, playlist_id)
471                 programme_id = playlist_id
472
473         return programme_id, title, description, duration, formats, subtitles
474
475     def _real_extract(self, url):
476         group_id = self._match_id(url)
477
478         webpage = self._download_webpage(url, group_id, 'Downloading video page')
479
480         programme_id = None
481         duration = None
482
483         tviplayer = self._search_regex(
484             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
485             webpage, 'player', default=None)
486
487         if tviplayer:
488             player = self._parse_json(tviplayer, group_id).get('player', {})
489             duration = int_or_none(player.get('duration'))
490             programme_id = player.get('vpid')
491
492         if not programme_id:
493             programme_id = self._search_regex(
494                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
495
496         if programme_id:
497             formats, subtitles = self._download_media_selector(programme_id)
498             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
499                 (r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>',
500                  r'<div[^>]+class="info"[^>]*>\s*<h1>(.+?)</h1>'), webpage, 'title')
501             description = self._search_regex(
502                 (r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
503                  r'<div[^>]+class="info_+synopsis"[^>]*>([^<]+)</div>'),
504                 webpage, 'description', default=None)
505             if not description:
506                 description = self._html_search_meta('description', webpage)
507         else:
508             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
509
510         self._sort_formats(formats)
511
512         return {
513             'id': programme_id,
514             'title': title,
515             'description': description,
516             'thumbnail': self._og_search_thumbnail(webpage, default=None),
517             'duration': duration,
518             'formats': formats,
519             'subtitles': subtitles,
520         }
521
522
523 class BBCIE(BBCCoUkIE):
524     IE_NAME = 'bbc'
525     IE_DESC = 'BBC'
526     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
527
528     _MEDIASELECTOR_URLS = [
529         # Provides HQ HLS streams but fails with geolocation in some cases when it's
530         # even not geo restricted at all
531         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
532         # Provides more formats, namely direct mp4 links, but fails on some videos with
533         # notukerror for non UK (?) users (e.g.
534         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
535         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
536         # Provides fewer formats, but works everywhere for everybody (hopefully)
537         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
538     ]
539
540     _TESTS = [{
541         # article with multiple videos embedded with data-playable containing vpids
542         'url': 'http://www.bbc.com/news/world-europe-32668511',
543         'info_dict': {
544             'id': 'world-europe-32668511',
545             'title': 'Russia stages massive WW2 parade despite Western boycott',
546             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
547         },
548         'playlist_count': 2,
549     }, {
550         # article with multiple videos embedded with data-playable (more videos)
551         'url': 'http://www.bbc.com/news/business-28299555',
552         'info_dict': {
553             'id': 'business-28299555',
554             'title': 'Farnborough Airshow: Video highlights',
555             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
556         },
557         'playlist_count': 9,
558         'skip': 'Save time',
559     }, {
560         # article with multiple videos embedded with `new SMP()`
561         # broken
562         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
563         'info_dict': {
564             'id': '3662a707-0af9-3149-963f-47bea720b460',
565             'title': 'BUGGER',
566         },
567         'playlist_count': 18,
568     }, {
569         # single video embedded with data-playable containing vpid
570         'url': 'http://www.bbc.com/news/world-europe-32041533',
571         'info_dict': {
572             'id': 'p02mprgb',
573             'ext': 'mp4',
574             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
575             'description': 'md5:2868290467291b37feda7863f7a83f54',
576             'duration': 47,
577             'timestamp': 1427219242,
578             'upload_date': '20150324',
579         },
580         'params': {
581             # rtmp download
582             'skip_download': True,
583         }
584     }, {
585         # article with single video embedded with data-playable containing XML playlist
586         # with direct video links as progressiveDownloadUrl (for now these are extracted)
587         # and playlist with f4m and m3u8 as streamingUrl
588         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
589         'info_dict': {
590             'id': '150615_telabyad_kentin_cogu',
591             'ext': 'mp4',
592             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
593             'timestamp': 1434397334,
594             'upload_date': '20150615',
595         },
596         'params': {
597             'skip_download': True,
598         }
599     }, {
600         # single video embedded with data-playable containing XML playlists (regional section)
601         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
602         'info_dict': {
603             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
604             'ext': 'mp4',
605             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
606             'timestamp': 1434713142,
607             'upload_date': '20150619',
608         },
609         'params': {
610             'skip_download': True,
611         }
612     }, {
613         # single video from video playlist embedded with vxp-playlist-data JSON
614         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
615         'info_dict': {
616             'id': 'p02w6qjc',
617             'ext': 'mp4',
618             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
619             'duration': 56,
620             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
621         },
622         'params': {
623             'skip_download': True,
624         }
625     }, {
626         # single video story with digitalData
627         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
628         'info_dict': {
629             'id': 'p02q6gc4',
630             'ext': 'flv',
631             'title': 'Sri Lanka’s spicy secret',
632             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
633             'timestamp': 1437674293,
634             'upload_date': '20150723',
635         },
636         'params': {
637             # rtmp download
638             'skip_download': True,
639         }
640     }, {
641         # single video story without digitalData
642         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
643         'info_dict': {
644             'id': 'p018zqqg',
645             'ext': 'mp4',
646             'title': 'Hyundai Santa Fe Sport: Rock star',
647             'description': 'md5:b042a26142c4154a6e472933cf20793d',
648             'timestamp': 1415867444,
649             'upload_date': '20141113',
650         },
651         'params': {
652             # rtmp download
653             'skip_download': True,
654         }
655     }, {
656         # single video with playlist.sxml URL in playlist param
657         'url': 'http://www.bbc.com/sport/0/football/33653409',
658         'info_dict': {
659             'id': 'p02xycnp',
660             'ext': 'mp4',
661             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
662             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
663             'duration': 140,
664         },
665         'params': {
666             # rtmp download
667             'skip_download': True,
668         }
669     }, {
670         # article with multiple videos embedded with playlist.sxml in playlist param
671         'url': 'http://www.bbc.com/sport/0/football/34475836',
672         'info_dict': {
673             'id': '34475836',
674             'title': 'Jurgen Klopp: Furious football from a witty and winning coach',
675             'description': 'Fast-paced football, wit, wisdom and a ready smile - why Liverpool fans should come to love new boss Jurgen Klopp.',
676         },
677         'playlist_count': 3,
678     }, {
679         # school report article with single video
680         'url': 'http://www.bbc.co.uk/schoolreport/35744779',
681         'info_dict': {
682             'id': '35744779',
683             'title': 'School which breaks down barriers in Jerusalem',
684         },
685         'playlist_count': 1,
686     }, {
687         # single video with playlist URL from weather section
688         'url': 'http://www.bbc.com/weather/features/33601775',
689         'only_matching': True,
690     }, {
691         # custom redirection to www.bbc.com
692         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
693         'only_matching': True,
694     }, {
695         # single video article embedded with data-media-vpid
696         'url': 'http://www.bbc.co.uk/sport/rowing/35908187',
697         'only_matching': True,
698     }]
699
700     @classmethod
701     def suitable(cls, url):
702         EXCLUDE_IE = (BBCCoUkIE, BBCCoUkArticleIE, BBCCoUkIPlayerPlaylistIE, BBCCoUkPlaylistIE)
703         return (False if any(ie.suitable(url) for ie in EXCLUDE_IE)
704                 else super(BBCIE, cls).suitable(url))
705
706     def _extract_from_media_meta(self, media_meta, video_id):
707         # Direct links to media in media metadata (e.g.
708         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
709         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
710         source_files = media_meta.get('sourceFiles')
711         if source_files:
712             return [{
713                 'url': f['url'],
714                 'format_id': format_id,
715                 'ext': f.get('encoding'),
716                 'tbr': float_or_none(f.get('bitrate'), 1000),
717                 'filesize': int_or_none(f.get('filesize')),
718             } for format_id, f in source_files.items() if f.get('url')], []
719
720         programme_id = media_meta.get('externalId')
721         if programme_id:
722             return self._download_media_selector(programme_id)
723
724         # Process playlist.sxml as legacy playlist
725         href = media_meta.get('href')
726         if href:
727             playlist = self._download_legacy_playlist_url(href)
728             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
729             return formats, subtitles
730
731         return [], []
732
733     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
734         programme_id, title, description, duration, formats, subtitles = \
735             self._process_legacy_playlist_url(url, playlist_id)
736         self._sort_formats(formats)
737         return {
738             'id': programme_id,
739             'title': title,
740             'description': description,
741             'duration': duration,
742             'timestamp': timestamp,
743             'formats': formats,
744             'subtitles': subtitles,
745         }
746
747     def _real_extract(self, url):
748         playlist_id = self._match_id(url)
749
750         webpage = self._download_webpage(url, playlist_id)
751
752         json_ld_info = self._search_json_ld(webpage, playlist_id, default=None)
753         timestamp = json_ld_info.get('timestamp')
754
755         playlist_title = json_ld_info.get('title')
756         if not playlist_title:
757             playlist_title = self._og_search_title(
758                 webpage, default=None) or self._html_search_regex(
759                 r'<title>(.+?)</title>', webpage, 'playlist title', default=None)
760             if playlist_title:
761                 playlist_title = re.sub(r'(.+)\s*-\s*BBC.*?$', r'\1', playlist_title).strip()
762
763         playlist_description = json_ld_info.get(
764             'description') or self._og_search_description(webpage, default=None)
765
766         if not timestamp:
767             timestamp = parse_iso8601(self._search_regex(
768                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
769                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
770                  r'"datePublished":\s*"([^"]+)'],
771                 webpage, 'date', default=None))
772
773         entries = []
774
775         # article with multiple videos embedded with playlist.sxml (e.g.
776         # http://www.bbc.com/sport/0/football/34475836)
777         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
778         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
779         if playlists:
780             entries = [
781                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
782                 for playlist_url in playlists]
783
784         # news article with multiple videos embedded with data-playable
785         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
786         if data_playables:
787             for _, data_playable_json in data_playables:
788                 data_playable = self._parse_json(
789                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
790                 if not data_playable:
791                     continue
792                 settings = data_playable.get('settings', {})
793                 if settings:
794                     # data-playable with video vpid in settings.playlistObject.items (e.g.
795                     # http://www.bbc.com/news/world-us-canada-34473351)
796                     playlist_object = settings.get('playlistObject', {})
797                     if playlist_object:
798                         items = playlist_object.get('items')
799                         if items and isinstance(items, list):
800                             title = playlist_object['title']
801                             description = playlist_object.get('summary')
802                             duration = int_or_none(items[0].get('duration'))
803                             programme_id = items[0].get('vpid')
804                             formats, subtitles = self._download_media_selector(programme_id)
805                             self._sort_formats(formats)
806                             entries.append({
807                                 'id': programme_id,
808                                 'title': title,
809                                 'description': description,
810                                 'timestamp': timestamp,
811                                 'duration': duration,
812                                 'formats': formats,
813                                 'subtitles': subtitles,
814                             })
815                     else:
816                         # data-playable without vpid but with a playlist.sxml URLs
817                         # in otherSettings.playlist (e.g.
818                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
819                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
820                         if playlist:
821                             entries.append(self._extract_from_playlist_sxml(
822                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
823
824         if entries:
825             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
826
827         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
828         programme_id = self._search_regex(
829             [r'data-(?:video-player|media)-vpid="(%s)"' % self._ID_REGEX,
830              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
831              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
832             webpage, 'vpid', default=None)
833
834         if programme_id:
835             formats, subtitles = self._download_media_selector(programme_id)
836             self._sort_formats(formats)
837             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
838             digital_data = self._parse_json(
839                 self._search_regex(
840                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
841                 programme_id, fatal=False)
842             page_info = digital_data.get('page', {}).get('pageInfo', {})
843             title = page_info.get('pageName') or self._og_search_title(webpage)
844             description = page_info.get('description') or self._og_search_description(webpage)
845             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
846             return {
847                 'id': programme_id,
848                 'title': title,
849                 'description': description,
850                 'timestamp': timestamp,
851                 'formats': formats,
852                 'subtitles': subtitles,
853             }
854
855         def extract_all(pattern):
856             return list(filter(None, map(
857                 lambda s: self._parse_json(s, playlist_id, fatal=False),
858                 re.findall(pattern, webpage))))
859
860         # Multiple video article (e.g.
861         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
862         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
863         entries = []
864         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
865             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
866             if embed_url and re.match(EMBED_URL, embed_url):
867                 entries.append(embed_url)
868         entries.extend(re.findall(
869             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
870         if entries:
871             return self.playlist_result(
872                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
873                 playlist_id, playlist_title, playlist_description)
874
875         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
876         medias = extract_all(r"data-media-meta='({[^']+})'")
877
878         if not medias:
879             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
880             media_asset = self._search_regex(
881                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
882                 webpage, 'media asset', default=None)
883             if media_asset:
884                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
885                 medias = []
886                 for video in media_asset_page.get('videos', {}).values():
887                     medias.extend(video.values())
888
889         if not medias:
890             # Multiple video playlist with single `now playing` entry (e.g.
891             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
892             vxp_playlist = self._parse_json(
893                 self._search_regex(
894                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
895                     webpage, 'playlist data'),
896                 playlist_id)
897             playlist_medias = []
898             for item in vxp_playlist:
899                 media = item.get('media')
900                 if not media:
901                     continue
902                 playlist_medias.append(media)
903                 # Download single video if found media with asset id matching the video id from URL
904                 if item.get('advert', {}).get('assetId') == playlist_id:
905                     medias = [media]
906                     break
907             # Fallback to the whole playlist
908             if not medias:
909                 medias = playlist_medias
910
911         entries = []
912         for num, media_meta in enumerate(medias, start=1):
913             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
914             if not formats:
915                 continue
916             self._sort_formats(formats)
917
918             video_id = media_meta.get('externalId')
919             if not video_id:
920                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
921
922             title = media_meta.get('caption')
923             if not title:
924                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
925
926             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
927
928             images = []
929             for image in media_meta.get('images', {}).values():
930                 images.extend(image.values())
931             if 'image' in media_meta:
932                 images.append(media_meta['image'])
933
934             thumbnails = [{
935                 'url': image.get('href'),
936                 'width': int_or_none(image.get('width')),
937                 'height': int_or_none(image.get('height')),
938             } for image in images]
939
940             entries.append({
941                 'id': video_id,
942                 'title': title,
943                 'thumbnails': thumbnails,
944                 'duration': duration,
945                 'timestamp': timestamp,
946                 'formats': formats,
947                 'subtitles': subtitles,
948             })
949
950         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
951
952
953 class BBCCoUkArticleIE(InfoExtractor):
954     _VALID_URL = r'https?://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
955     IE_NAME = 'bbc.co.uk:article'
956     IE_DESC = 'BBC articles'
957
958     _TEST = {
959         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
960         'info_dict': {
961             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
962             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
963             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
964         },
965         'playlist_count': 4,
966         'add_ie': ['BBCCoUk'],
967     }
968
969     def _real_extract(self, url):
970         playlist_id = self._match_id(url)
971
972         webpage = self._download_webpage(url, playlist_id)
973
974         title = self._og_search_title(webpage)
975         description = self._og_search_description(webpage).strip()
976
977         entries = [self.url_result(programme_url) for programme_url in re.findall(
978             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
979
980         return self.playlist_result(entries, playlist_id, title, description)
981
982
983 class BBCCoUkPlaylistBaseIE(InfoExtractor):
984     def _real_extract(self, url):
985         playlist_id = self._match_id(url)
986
987         webpage = self._download_webpage(url, playlist_id)
988
989         entries = [
990             self.url_result(self._URL_TEMPLATE % video_id, BBCCoUkIE.ie_key())
991             for video_id in re.findall(
992                 self._VIDEO_ID_TEMPLATE % BBCCoUkIE._ID_REGEX, webpage)]
993
994         title, description = self._extract_title_and_description(webpage)
995
996         return self.playlist_result(entries, playlist_id, title, description)
997
998
999 class BBCCoUkIPlayerPlaylistIE(BBCCoUkPlaylistBaseIE):
1000     IE_NAME = 'bbc.co.uk:iplayer:playlist'
1001     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/iplayer/episodes/(?P<id>%s)' % BBCCoUkIE._ID_REGEX
1002     _URL_TEMPLATE = 'http://www.bbc.co.uk/iplayer/episode/%s'
1003     _VIDEO_ID_TEMPLATE = r'data-ip-id=["\'](%s)'
1004     _TEST = {
1005         'url': 'http://www.bbc.co.uk/iplayer/episodes/b05rcz9v',
1006         'info_dict': {
1007             'id': 'b05rcz9v',
1008             'title': 'The Disappearance',
1009             'description': 'French thriller serial about a missing teenager.',
1010         },
1011         'playlist_mincount': 6,
1012     }
1013
1014     def _extract_title_and_description(self, webpage):
1015         title = self._search_regex(r'<h1>([^<]+)</h1>', webpage, 'title', fatal=False)
1016         description = self._search_regex(
1017             r'<p[^>]+class=(["\'])subtitle\1[^>]*>(?P<value>[^<]+)</p>',
1018             webpage, 'description', fatal=False, group='value')
1019         return title, description
1020
1021
1022 class BBCCoUkPlaylistIE(BBCCoUkPlaylistBaseIE):
1023     IE_NAME = 'bbc.co.uk:playlist'
1024     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/(?P<id>%s)/(?:episodes|broadcasts|clips)' % BBCCoUkIE._ID_REGEX
1025     _URL_TEMPLATE = 'http://www.bbc.co.uk/programmes/%s'
1026     _VIDEO_ID_TEMPLATE = r'data-pid=["\'](%s)'
1027     _TESTS = [{
1028         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1029         'info_dict': {
1030             'id': 'b05rcz9v',
1031             'title': 'The Disappearance - Clips - BBC Four',
1032             'description': 'French thriller serial about a missing teenager.',
1033         },
1034         'playlist_mincount': 7,
1035     }, {
1036         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/broadcasts/2016/06',
1037         'only_matching': True,
1038     }, {
1039         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1040         'only_matching': True,
1041     }, {
1042         'url': 'http://www.bbc.co.uk/programmes/b055jkys/episodes/player',
1043         'only_matching': True,
1044     }]
1045
1046     def _extract_title_and_description(self, webpage):
1047         title = self._og_search_title(webpage, fatal=False)
1048         description = self._og_search_description(webpage)
1049         return title, description