2 from __future__ import unicode_literals
7 from .common import InfoExtractor
9 compat_etree_fromstring,
12 compat_urllib_parse_urlparse,
14 compat_xml_parse_error,
32 class BrightcoveLegacyIE(InfoExtractor):
33 IE_NAME = 'brightcove:legacy'
34 _VALID_URL = r'(?:https?://.*brightcove\.com/(services|viewer).*?\?|brightcove:)(?P<query>.*)'
35 _FEDERATED_URL = 'http://c.brightcove.com/services/viewer/htmlFederated'
39 # From http://www.8tv.cat/8aldia/videos/xavier-sala-i-martin-aquesta-tarda-a-8-al-dia/
40 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1654948606001&flashID=myExperience&%40videoPlayer=2371591881001',
41 'md5': '5423e113865d26e40624dce2e4b45d95',
42 'note': 'Test Brightcove downloads and detection in GenericIE',
44 'id': '2371591881001',
46 'title': 'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”',
48 'description': 'md5:a950cc4285c43e44d763d036710cd9cd',
49 'timestamp': 1368213670,
50 'upload_date': '20130510',
51 'uploader_id': '1589608506001',
55 # From http://medianetwork.oracle.com/video/player/1785452137001
56 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1217746023001&flashID=myPlayer&%40videoPlayer=1785452137001',
58 'id': '1785452137001',
60 'title': 'JVMLS 2012: Arrays 2.0 - Opportunities and Challenges',
61 'description': 'John Rose speaks at the JVM Language Summit, August 1, 2012.',
63 'timestamp': 1344975024,
64 'upload_date': '20120814',
65 'uploader_id': '1460825906',
69 # From http://mashable.com/2013/10/26/thermoelectric-bracelet-lets-you-control-your-body-temperature/
70 'url': 'http://c.brightcove.com/services/viewer/federated_f9?&playerID=1265504713001&publisherID=AQ%7E%7E%2CAAABBzUwv1E%7E%2CxP-xFHVUstiMFlNYfvF4G9yFnNaqCw_9&videoID=2750934548001',
72 'id': '2750934548001',
74 'title': 'This Bracelet Acts as a Personal Thermostat',
75 'description': 'md5:547b78c64f4112766ccf4e151c20b6a0',
76 'uploader': 'Mashable',
77 'timestamp': 1382041798,
78 'upload_date': '20131017',
79 'uploader_id': '1130468786001',
83 # test that the default referer works
84 # from http://national.ballet.ca/interact/video/Lost_in_Motion_II/
85 'url': 'http://link.brightcove.com/services/player/bcpid756015033001?bckey=AQ~~,AAAApYJi_Ck~,GxhXCegT1Dp39ilhXuxMJxasUhVNZiil&bctid=2878862109001',
87 'id': '2878862109001',
89 'title': 'Lost in Motion II',
90 'description': 'md5:363109c02998fee92ec02211bd8000df',
91 'uploader': 'National Ballet of Canada',
95 # test flv videos served by akamaihd.net
96 # From http://www.redbull.com/en/bike/stories/1331655643987/replay-uci-dh-world-cup-2014-from-fort-william
97 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?%40videoPlayer=ref%3Aevent-stream-356&linkBaseURL=http%3A%2F%2Fwww.redbull.com%2Fen%2Fbike%2Fvideos%2F1331655630249%2Freplay-uci-fort-william-2014-dh&playerKey=AQ%7E%7E%2CAAAApYJ7UqE%7E%2Cxqr_zXk0I-zzNndy8NlHogrCb5QdyZRf&playerID=1398061561001#__youtubedl_smuggle=%7B%22Referer%22%3A+%22http%3A%2F%2Fwww.redbull.com%2Fen%2Fbike%2Fstories%2F1331655643987%2Freplay-uci-dh-world-cup-2014-from-fort-william%22%7D',
98 # The md5 checksum changes on each download
100 'id': '3750436379001',
102 'title': 'UCI MTB World Cup 2014: Fort William, UK - Downhill Finals',
103 'uploader': 'RBTV Old (do not use)',
104 'description': 'UCI MTB World Cup 2014: Fort William, UK - Downhill Finals',
105 'timestamp': 1409122195,
106 'upload_date': '20140827',
107 'uploader_id': '710858724001',
112 # from http://support.brightcove.com/en/video-cloud/docs/playlist-support-single-video-players
113 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=3550052898001&playerKey=AQ%7E%7E%2CAAABmA9XpXk%7E%2C-Kp7jNgisre1fG5OdqpAFUTcs0lP_ZoL',
116 'id': '3550319591001',
118 'playlist_mincount': 7,
129 def _build_brighcove_url(cls, object_str):
131 Build a Brightcove url from a xml string containing
132 <object class="BrightcoveExperience">{params}</object>
135 # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553
136 object_str = re.sub(r'(<param(?:\s+[a-zA-Z0-9_]+="[^"]*")*)>',
137 lambda m: m.group(1) + '/>', object_str)
138 # Fix up some stupid XML, see https://github.com/rg3/youtube-dl/issues/1608
139 object_str = object_str.replace('<--', '<!--')
140 # remove namespace to simplify extraction
141 object_str = re.sub(r'(<object[^>]*)(xmlns=".*?")', r'\1', object_str)
142 object_str = fix_xml_ampersands(object_str)
145 object_doc = compat_etree_fromstring(object_str.encode('utf-8'))
146 except compat_xml_parse_error:
149 fv_el = find_xpath_attr(object_doc, './param', 'name', 'flashVars')
150 if fv_el is not None:
153 for k, v in compat_parse_qs(fv_el.attrib['value']).items())
157 data_url = object_doc.attrib.get('data', '')
158 data_url_params = compat_parse_qs(compat_urllib_parse_urlparse(data_url).query)
160 def find_param(name):
161 if name in flashvars:
162 return flashvars[name]
163 node = find_xpath_attr(object_doc, './param', 'name', name)
165 return node.attrib['value']
166 return data_url_params.get(name)
170 playerID = find_param('playerID')
172 raise ExtractorError('Cannot find player ID')
173 params['playerID'] = playerID
175 playerKey = find_param('playerKey')
176 # Not all pages define this value
177 if playerKey is not None:
178 params['playerKey'] = playerKey
179 # These fields hold the id of the video
180 videoPlayer = find_param('@videoPlayer') or find_param('videoId') or find_param('videoID') or find_param('@videoList')
181 if videoPlayer is not None:
182 params['@videoPlayer'] = videoPlayer
183 linkBase = find_param('linkBaseURL')
184 if linkBase is not None:
185 params['linkBaseURL'] = linkBase
186 return cls._make_brightcove_url(params)
189 def _build_brighcove_url_from_js(cls, object_js):
190 # The layout of JS is as follows:
191 # customBC.createVideo = function (width, height, playerID, playerKey, videoPlayer, VideoRandomID) {
192 # // build Brightcove <object /> XML
195 r'''(?x)customBC.\createVideo\(
196 .*? # skipping width and height
197 ["\'](?P<playerID>\d+)["\']\s*,\s* # playerID
198 ["\'](?P<playerKey>AQ[^"\']{48})[^"\']*["\']\s*,\s* # playerKey begins with AQ and is 50 characters
199 # in length, however it's appended to itself
200 # in places, so truncate
201 ["\'](?P<videoID>\d+)["\'] # @videoPlayer
204 return cls._make_brightcove_url(m.groupdict())
207 def _make_brightcove_url(cls, params):
208 return update_url_query(cls._FEDERATED_URL, params)
211 def _extract_brightcove_url(cls, webpage):
212 """Try to extract the brightcove url from the webpage, returns None
215 urls = cls._extract_brightcove_urls(webpage)
216 return urls[0] if urls else None
219 def _extract_brightcove_urls(cls, webpage):
220 """Return a list of all Brightcove URLs from the webpage """
223 r'<meta\s+property=[\'"]og:video[\'"]\s+content=[\'"](https?://(?:secure|c)\.brightcove.com/[^\'"]+)[\'"]',
226 url = unescapeHTML(url_m.group(1))
227 # Some sites don't add it, we can't download with this url, for example:
228 # http://www.ktvu.com/videos/news/raw-video-caltrain-releases-video-of-man-almost/vCTZdY/
229 if 'playerKey' in url or 'videoId' in url:
232 matches = re.findall(
235 [^>]+?class=[\'"][^>]*?BrightcoveExperience.*?[\'"] |
236 [^>]*?>\s*<param\s+name="movie"\s+value="https?://[^/]*brightcove\.com/
237 ).+?>\s*</object>''',
240 return list(filter(None, [cls._build_brighcove_url(m) for m in matches]))
242 return list(filter(None, [
243 cls._build_brighcove_url_from_js(custom_bc)
244 for custom_bc in re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)]))
246 def _real_extract(self, url):
247 url, smuggled_data = unsmuggle_url(url, {})
249 # Change the 'videoId' and others field to '@videoPlayer'
250 url = re.sub(r'(?<=[?&])(videoI(d|D)|bctid)', '%40videoPlayer', url)
251 # Change bckey (used by bcove.me urls) to playerKey
252 url = re.sub(r'(?<=[?&])bckey', 'playerKey', url)
253 mobj = re.match(self._VALID_URL, url)
254 query_str = mobj.group('query')
255 query = compat_urlparse.parse_qs(query_str)
257 videoPlayer = query.get('@videoPlayer')
259 # We set the original url as the default 'Referer' header
260 referer = smuggled_data.get('Referer', url)
261 return self._get_video_info(
262 videoPlayer[0], query, referer=referer)
263 elif 'playerKey' in query:
264 player_key = query['playerKey']
265 return self._get_playlist_info(player_key[0])
267 raise ExtractorError(
268 'Cannot find playerKey= variable. Did you forget quotes in a shell invocation?',
271 def _get_video_info(self, video_id, query, referer=None):
273 linkBase = query.get('linkBaseURL')
274 if linkBase is not None:
275 referer = linkBase[0]
276 if referer is not None:
277 headers['Referer'] = referer
278 webpage = self._download_webpage(self._FEDERATED_URL, video_id, headers=headers, query=query)
280 error_msg = self._html_search_regex(
281 r"<h1>We're sorry.</h1>([\s\n]*<p>.*?</p>)+", webpage,
282 'error message', default=None)
283 if error_msg is not None:
284 raise ExtractorError(
285 'brightcove said: %s' % error_msg, expected=True)
287 self.report_extraction(video_id)
288 info = self._search_regex(r'var experienceJSON = ({.*});', webpage, 'json')
289 info = json.loads(info)['data']
290 video_info = info['programmedContent']['videoPlayer']['mediaDTO']
291 video_info['_youtubedl_adServerURL'] = info.get('adServerURL')
293 return self._extract_video_info(video_info)
295 def _get_playlist_info(self, player_key):
296 info_url = 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=%s' % player_key
297 playlist_info = self._download_webpage(
298 info_url, player_key, 'Downloading playlist information')
300 json_data = json.loads(playlist_info)
301 if 'videoList' not in json_data:
302 raise ExtractorError('Empty playlist')
303 playlist_info = json_data['videoList']
304 videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']]
306 return self.playlist_result(videos, playlist_id='%s' % playlist_info['id'],
307 playlist_title=playlist_info['mediaCollectionDTO']['displayName'])
309 def _extract_video_info(self, video_info):
310 publisher_id = video_info.get('publisherId')
312 'id': compat_str(video_info['id']),
313 'title': video_info['displayName'].strip(),
314 'description': video_info.get('shortDescription'),
315 'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'),
316 'uploader': video_info.get('publisherName'),
317 'uploader_id': compat_str(publisher_id) if publisher_id else None,
318 'duration': float_or_none(video_info.get('length'), 1000),
319 'timestamp': int_or_none(video_info.get('creationDate'), 1000),
322 renditions = video_info.get('renditions', []) + video_info.get('IOSRenditions', [])
325 for rend in renditions:
326 url = rend['defaultURL']
331 url_comp = compat_urllib_parse_urlparse(url)
332 if url_comp.path.endswith('.m3u8'):
334 self._extract_m3u8_formats(url, info['id'], 'mp4'))
336 elif 'akamaihd.net' in url_comp.netloc:
337 # This type of renditions are served through
338 # akamaihd.net, but they don't use f4m manifests
339 url = url.replace('control/', '') + '?&v=3.3.0&fp=13&r=FEEFJ&g=RTSJIMBMPFPB'
342 ext = determine_ext(url)
343 tbr = int_or_none(rend.get('encodingRate'), 1000)
345 'format_id': 'http%s' % ('-%s' % tbr if tbr else ''),
348 'filesize': int_or_none(rend.get('size')) or None,
351 if rend.get('audioOnly'):
357 'height': int_or_none(rend.get('frameHeight')),
358 'width': int_or_none(rend.get('frameWidth')),
359 'vcodec': rend.get('videoCodec'),
362 # m3u8 manifests with remote == false are media playlists
363 # Not calling _extract_m3u8_formats here to save network traffic
366 'format_id': 'hls%s' % ('-%s' % tbr if tbr else ''),
371 formats.append(a_format)
372 self._sort_formats(formats)
373 info['formats'] = formats
374 elif video_info.get('FLVFullLengthURL') is not None:
376 'url': video_info['FLVFullLengthURL'],
377 'vcodec': self.FLV_VCODECS.get(video_info.get('FLVFullCodec')),
378 'filesize': int_or_none(video_info.get('FLVFullSize')),
381 if self._downloader.params.get('include_ads', False):
382 adServerURL = video_info.get('_youtubedl_adServerURL')
391 'title': info['title'],
392 'entries': [ad_info, info],
397 if 'url' not in info and not info.get('formats'):
398 raise ExtractorError('Unable to extract video url for %s' % info['id'])
402 class BrightcoveNewIE(InfoExtractor):
403 IE_NAME = 'brightcove:new'
404 _VALID_URL = r'https?://players\.brightcove\.net/(?P<account_id>\d+)/(?P<player_id>[^/]+)_(?P<embed>[^/]+)/index\.html\?.*videoId=(?P<video_id>\d+|ref:[^&]+)'
406 'url': 'http://players.brightcove.net/929656772001/e41d32dc-ec74-459e-a845-6c69f7b724ea_default/index.html?videoId=4463358922001',
407 'md5': 'c8100925723840d4b0d243f7025703be',
409 'id': '4463358922001',
411 'title': 'Meet the man behind Popcorn Time',
412 'description': 'md5:eac376a4fe366edc70279bfb681aea16',
414 'timestamp': 1441391203,
415 'upload_date': '20150904',
416 'uploader_id': '929656772001',
417 'formats': 'mincount:22',
421 'url': 'http://players.brightcove.net/4036320279001/5d112ed9-283f-485f-a7f9-33f42e8bc042_default/index.html?videoId=4279049078001',
423 'id': '4279049078001',
425 'title': 'Titansgrave: Chapter 0',
426 'description': 'Titansgrave: Chapter 0',
427 'duration': 1242.058,
428 'timestamp': 1433556729,
429 'upload_date': '20150606',
430 'uploader_id': '4036320279001',
431 'formats': 'mincount:41',
435 'skip_download': True,
438 # ref: prefixed video id
439 'url': 'http://players.brightcove.net/3910869709001/21519b5c-4b3b-4363-accb-bdc8f358f823_default/index.html?videoId=ref:7069442',
440 'only_matching': True,
442 # non numeric ref: prefixed video id
443 'url': 'http://players.brightcove.net/710858724001/default_default/index.html?videoId=ref:event-stream-356',
444 'only_matching': True,
448 def _extract_url(webpage):
449 urls = BrightcoveNewIE._extract_urls(webpage)
450 return urls[0] if urls else None
453 def _extract_urls(webpage):
455 # 1. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideoiniframe
456 # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideousingjavascript
457 # 3. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/embed-in-page.html
458 # 4. https://support.brightcove.com/en/video-cloud/docs/dynamically-assigning-videos-player
462 # Look for iframe embeds [1]
463 for _, url in re.findall(
464 r'<iframe[^>]+src=(["\'])((?:https?:)?//players\.brightcove\.net/\d+/[^/]+/index\.html.+?)\1', webpage):
465 entries.append(url if url.startswith('http') else 'http:' + url)
467 # Look for embed_in_page embeds [2]
468 for video_id, account_id, player_id, embed in re.findall(
469 # According to examples from [3] it's unclear whether video id
470 # may be optional and what to do when it is
471 # According to [4] data-video-id may be prefixed with ref:
474 data-video-id=["\'](\d+|ref:[^"\']+)["\'][^>]*>.*?
477 src=["\'](?:https?:)?//players\.brightcove\.net/
478 (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
481 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s'
482 % (account_id, player_id, embed, video_id))
486 def _real_extract(self, url):
487 account_id, player_id, embed, video_id = re.match(self._VALID_URL, url).groups()
489 webpage = self._download_webpage(
490 'http://players.brightcove.net/%s/%s_%s/index.min.js'
491 % (account_id, player_id, embed), video_id)
495 catalog = self._search_regex(
496 r'catalog\(({.+?})\);', webpage, 'catalog', default=None)
498 catalog = self._parse_json(
499 js_to_json(catalog), video_id, fatal=False)
501 policy_key = catalog.get('policyKey')
504 policy_key = self._search_regex(
505 r'policyKey\s*:\s*(["\'])(?P<pk>.+?)\1',
506 webpage, 'policy key', group='pk')
508 api_url = 'https://edge.api.brightcove.com/playback/v1/accounts/%s/videos/%s' % (account_id, video_id)
510 json_data = self._download_json(api_url, video_id, headers={
511 'Accept': 'application/json;pk=%s' % policy_key
513 except ExtractorError as e:
514 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
515 json_data = self._parse_json(e.cause.read().decode(), video_id)
516 raise ExtractorError(json_data[0]['message'], expected=True)
519 title = json_data['name'].strip()
522 for source in json_data.get('sources', []):
523 container = source.get('container')
524 source_type = source.get('type')
525 src = source.get('src')
526 if source_type == 'application/x-mpegURL' or container == 'M2TS':
529 formats.extend(self._extract_m3u8_formats(
530 src, video_id, 'mp4', m3u8_id='hls', fatal=False))
531 elif source_type == 'application/dash+xml':
534 formats.extend(self._extract_mpd_formats(src, video_id, 'dash', fatal=False))
536 streaming_src = source.get('streaming_src')
537 stream_name, app_name = source.get('stream_name'), source.get('app_name')
538 if not src and not streaming_src and (not stream_name or not app_name):
540 tbr = float_or_none(source.get('avg_bitrate'), 1000)
541 height = int_or_none(source.get('height'))
542 width = int_or_none(source.get('width'))
545 'filesize': int_or_none(source.get('size')),
546 'container': container,
547 'ext': container.lower(),
549 if width == 0 and height == 0:
557 'vcodec': source.get('codec'),
560 def build_format_id(kind):
563 format_id += '-%dk' % int(tbr)
565 format_id += '-%dp' % height
568 if src or streaming_src:
570 'url': src or streaming_src,
571 'format_id': build_format_id('http' if src else 'http-streaming'),
572 'source_preference': 0 if src else -1,
577 'play_path': stream_name,
578 'format_id': build_format_id('rtmp'),
581 self._sort_formats(formats)
584 for text_track in json_data.get('text_tracks', []):
585 if text_track.get('src'):
586 subtitles.setdefault(text_track.get('srclang'), []).append({
587 'url': text_track['src'],
593 'description': json_data.get('description'),
594 'thumbnail': json_data.get('thumbnail') or json_data.get('poster'),
595 'duration': float_or_none(json_data.get('duration'), 1000),
596 'timestamp': parse_iso8601(json_data.get('published_at')),
597 'uploader_id': account_id,
599 'subtitles': subtitles,
600 'tags': json_data.get('tags', []),