2 from __future__ import unicode_literals
6 import xml.etree.ElementTree
8 from .common import InfoExtractor
13 compat_urllib_parse_urlparse,
14 compat_urllib_request,
27 class BrightcoveIE(InfoExtractor):
28 _VALID_URL = r'(?:https?://.*brightcove\.com/(services|viewer).*?\?|brightcove:)(?P<query>.*)'
29 _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
33 # From http://www.8tv.cat/8aldia/videos/xavier-sala-i-martin-aquesta-tarda-a-8-al-dia/
34 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1654948606001&flashID=myExperience&%40videoPlayer=2371591881001',
35 'md5': '5423e113865d26e40624dce2e4b45d95',
36 'note': 'Test Brightcove downloads and detection in GenericIE',
38 'id': '2371591881001',
40 'title': 'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”',
42 'description': 'md5:a950cc4285c43e44d763d036710cd9cd',
46 # From http://medianetwork.oracle.com/video/player/1785452137001
47 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1217746023001&flashID=myPlayer&%40videoPlayer=1785452137001',
49 'id': '1785452137001',
51 'title': 'JVMLS 2012: Arrays 2.0 - Opportunities and Challenges',
52 'description': 'John Rose speaks at the JVM Language Summit, August 1, 2012.',
57 # From http://mashable.com/2013/10/26/thermoelectric-bracelet-lets-you-control-your-body-temperature/
58 'url': 'http://c.brightcove.com/services/viewer/federated_f9?&playerID=1265504713001&publisherID=AQ%7E%7E%2CAAABBzUwv1E%7E%2CxP-xFHVUstiMFlNYfvF4G9yFnNaqCw_9&videoID=2750934548001',
60 'id': '2750934548001',
62 'title': 'This Bracelet Acts as a Personal Thermostat',
63 'description': 'md5:547b78c64f4112766ccf4e151c20b6a0',
64 'uploader': 'Mashable',
68 # test that the default referer works
69 # from http://national.ballet.ca/interact/video/Lost_in_Motion_II/
70 'url': 'http://link.brightcove.com/services/player/bcpid756015033001?bckey=AQ~~,AAAApYJi_Ck~,GxhXCegT1Dp39ilhXuxMJxasUhVNZiil&bctid=2878862109001',
72 'id': '2878862109001',
74 'title': 'Lost in Motion II',
75 'description': 'md5:363109c02998fee92ec02211bd8000df',
76 'uploader': 'National Ballet of Canada',
80 # test flv videos served by akamaihd.net
81 # From http://www.redbull.com/en/bike/stories/1331655643987/replay-uci-dh-world-cup-2014-from-fort-william
82 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?%40videoPlayer=ref%3ABC2996102916001&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',
83 # The md5 checksum changes on each download
85 'id': '2996102916001',
87 'title': 'UCI MTB World Cup 2014: Fort William, UK - Downhill Finals',
88 'uploader': 'Red Bull TV',
89 'description': 'UCI MTB World Cup 2014: Fort William, UK - Downhill Finals',
94 # from http://support.brightcove.com/en/video-cloud/docs/playlist-support-single-video-players
95 'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=3550052898001&playerKey=AQ%7E%7E%2CAAABmA9XpXk%7E%2C-Kp7jNgisre1fG5OdqpAFUTcs0lP_ZoL',
98 'id': '3550319591001',
100 'playlist_mincount': 7,
105 def _build_brighcove_url(cls, object_str):
107 Build a Brightcove url from a xml string containing
108 <object class="BrightcoveExperience">{params}</object>
111 # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553
112 object_str = re.sub(r'(<param(?:\s+[a-zA-Z0-9_]+="[^"]*")*)>',
113 lambda m: m.group(1) + '/>', object_str)
114 # Fix up some stupid XML, see https://github.com/rg3/youtube-dl/issues/1608
115 object_str = object_str.replace('<--', '<!--')
116 # remove namespace to simplify extraction
117 object_str = re.sub(r'(<object[^>]*)(xmlns=".*?")', r'\1', object_str)
118 object_str = fix_xml_ampersands(object_str)
121 object_doc = xml.etree.ElementTree.fromstring(object_str.encode('utf-8'))
122 except xml.etree.ElementTree.ParseError:
125 fv_el = find_xpath_attr(object_doc, './param', 'name', 'flashVars')
126 if fv_el is not None:
129 for k, v in compat_parse_qs(fv_el.attrib['value']).items())
133 def find_param(name):
134 if name in flashvars:
135 return flashvars[name]
136 node = find_xpath_attr(object_doc, './param', 'name', name)
138 return node.attrib['value']
143 playerID = find_param('playerID')
145 raise ExtractorError('Cannot find player ID')
146 params['playerID'] = playerID
148 playerKey = find_param('playerKey')
149 # Not all pages define this value
150 if playerKey is not None:
151 params['playerKey'] = playerKey
152 # The three fields hold the id of the video
153 videoPlayer = find_param('@videoPlayer') or find_param('videoId') or find_param('videoID')
154 if videoPlayer is not None:
155 params['@videoPlayer'] = videoPlayer
156 linkBase = find_param('linkBaseURL')
157 if linkBase is not None:
158 params['linkBaseURL'] = linkBase
159 data = compat_urllib_parse.urlencode(params)
160 return cls._FEDERATED_URL_TEMPLATE % data
163 def _extract_brightcove_url(cls, webpage):
164 """Try to extract the brightcove url from the webpage, returns None
167 urls = cls._extract_brightcove_urls(webpage)
168 return urls[0] if urls else None
171 def _extract_brightcove_urls(cls, webpage):
172 """Return a list of all Brightcove URLs from the webpage """
175 r'<meta\s+property=[\'"]og:video[\'"]\s+content=[\'"](https?://(?:secure|c)\.brightcove.com/[^\'"]+)[\'"]',
178 url = unescapeHTML(url_m.group(1))
179 # Some sites don't add it, we can't download with this url, for example:
180 # http://www.ktvu.com/videos/news/raw-video-caltrain-releases-video-of-man-almost/vCTZdY/
181 if 'playerKey' in url or 'videoId' in url:
184 matches = re.findall(
187 [^>]+?class=[\'"][^>]*?BrightcoveExperience.*?[\'"] |
188 [^>]*?>\s*<param\s+name="movie"\s+value="https?://[^/]*brightcove\.com/
189 ).+?>\s*</object>''',
192 return list(filter(None, [cls._build_brighcove_url(m) for m in matches]))
194 custombcs = re.findall(r'customBC.\createVideo\((.+?)\);',webpage)
197 for match in custombcs:
198 # brightcove playerkey begins with AQ and is 50 characters in length,
199 # however it's appended to itself in places, so truncate.
200 f = re.search(r'["\'](AQ[^"\']{48}).*?["\'](\d+)["\']', match)
202 urls.append('brightcove:playerKey='+f.group(1)+'&%40videoPlayer='+f.group(2))
205 def _real_extract(self, url):
206 url, smuggled_data = unsmuggle_url(url, {})
208 # Change the 'videoId' and others field to '@videoPlayer'
209 url = re.sub(r'(?<=[?&])(videoI(d|D)|bctid)', '%40videoPlayer', url)
210 # Change bckey (used by bcove.me urls) to playerKey
211 url = re.sub(r'(?<=[?&])bckey', 'playerKey', url)
212 mobj = re.match(self._VALID_URL, url)
213 query_str = mobj.group('query')
214 query = compat_urlparse.parse_qs(query_str)
216 videoPlayer = query.get('@videoPlayer')
218 # We set the original url as the default 'Referer' header
219 referer = smuggled_data.get('Referer', url)
220 return self._get_video_info(
221 videoPlayer[0], query_str, query, referer=referer)
222 elif 'playerKey' in query:
223 player_key = query['playerKey']
224 return self._get_playlist_info(player_key[0])
226 raise ExtractorError(
227 'Cannot find playerKey= variable. Did you forget quotes in a shell invocation?',
230 def _get_video_info(self, video_id, query_str, query, referer=None):
231 request_url = self._FEDERATED_URL_TEMPLATE % query_str
232 req = compat_urllib_request.Request(request_url)
233 linkBase = query.get('linkBaseURL')
234 if linkBase is not None:
235 referer = linkBase[0]
236 if referer is not None:
237 req.add_header('Referer', referer)
238 webpage = self._download_webpage(req, video_id)
240 error_msg = self._html_search_regex(
241 r"<h1>We're sorry.</h1>([\s\n]*<p>.*?</p>)+", webpage,
242 'error message', default=None)
243 if error_msg is not None:
244 raise ExtractorError(
245 'brightcove said: %s' % error_msg, expected=True)
247 self.report_extraction(video_id)
248 info = self._search_regex(r'var experienceJSON = ({.*});', webpage, 'json')
249 info = json.loads(info)['data']
250 video_info = info['programmedContent']['videoPlayer']['mediaDTO']
251 video_info['_youtubedl_adServerURL'] = info.get('adServerURL')
253 return self._extract_video_info(video_info)
255 def _get_playlist_info(self, player_key):
256 info_url = 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=%s' % player_key
257 playlist_info = self._download_webpage(
258 info_url, player_key, 'Downloading playlist information')
260 json_data = json.loads(playlist_info)
261 if 'videoList' not in json_data:
262 raise ExtractorError('Empty playlist')
263 playlist_info = json_data['videoList']
264 videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']]
266 return self.playlist_result(videos, playlist_id='%s' % playlist_info['id'],
267 playlist_title=playlist_info['mediaCollectionDTO']['displayName'])
269 def _extract_video_info(self, video_info):
271 'id': compat_str(video_info['id']),
272 'title': video_info['displayName'].strip(),
273 'description': video_info.get('shortDescription'),
274 'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'),
275 'uploader': video_info.get('publisherName'),
278 renditions = video_info.get('renditions')
281 for rend in renditions:
282 url = rend['defaultURL']
287 url_comp = compat_urllib_parse_urlparse(url)
288 if url_comp.path.endswith('.m3u8'):
290 self._extract_m3u8_formats(url, info['id'], 'mp4'))
292 elif 'akamaihd.net' in url_comp.netloc:
293 # This type of renditions are served through
294 # akamaihd.net, but they don't use f4m manifests
295 url = url.replace('control/', '') + '?&v=3.3.0&fp=13&r=FEEFJ&g=RTSJIMBMPFPB'
298 ext = determine_ext(url)
299 size = rend.get('size')
303 'height': rend.get('frameHeight'),
304 'width': rend.get('frameWidth'),
305 'filesize': size if size != 0 else None,
307 self._sort_formats(formats)
308 info['formats'] = formats
309 elif video_info.get('FLVFullLengthURL') is not None:
311 'url': video_info['FLVFullLengthURL'],
314 if self._downloader.params.get('include_ads', False):
315 adServerURL = video_info.get('_youtubedl_adServerURL')
324 'title': info['title'],
325 'entries': [ad_info, info],
330 if 'url' not in info and not info.get('formats'):
331 raise ExtractorError('Unable to extract video url for %s' % info['id'])