5 import xml.etree.ElementTree
7 from .common import InfoExtractor
17 class BrightcoveIE(InfoExtractor):
18 _VALID_URL = r'https?://.*brightcove\.com/(services|viewer).*\?(?P<query>.*)'
19 _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
20 _PLAYLIST_URL_TEMPLATE = 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=%s'
24 # From http://www.8tv.cat/8aldia/videos/xavier-sala-i-martin-aquesta-tarda-a-8-al-dia/
25 u'url': u'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1654948606001&flashID=myExperience&%40videoPlayer=2371591881001',
26 u'file': u'2371591881001.mp4',
27 u'md5': u'8eccab865181d29ec2958f32a6a754f5',
28 u'note': u'Test Brightcove downloads and detection in GenericIE',
30 u'title': u'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”',
32 u'description': u'md5:a950cc4285c43e44d763d036710cd9cd',
36 # From http://medianetwork.oracle.com/video/player/1785452137001
37 u'url': u'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1217746023001&flashID=myPlayer&%40videoPlayer=1785452137001',
38 u'file': u'1785452137001.flv',
40 u'title': u'JVMLS 2012: Arrays 2.0 - Opportunities and Challenges',
41 u'description': u'John Rose speaks at the JVM Language Summit, August 1, 2012.',
42 u'uploader': u'Oracle',
48 def _build_brighcove_url(cls, object_str):
50 Build a Brightcove url from a xml string containing
51 <object class="BrightcoveExperience">{params}</object>
54 # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553
55 object_str = re.sub(r'(<param name="[^"]+" value="[^"]+")>',
56 lambda m: m.group(1) + '/>', object_str)
57 # Fix up some stupid XML, see https://github.com/rg3/youtube-dl/issues/1608
58 object_str = object_str.replace(u'<--', u'<!--')
60 object_doc = xml.etree.ElementTree.fromstring(object_str)
61 assert u'BrightcoveExperience' in object_doc.attrib['class']
62 params = {'flashID': object_doc.attrib['id'],
63 'playerID': find_xpath_attr(object_doc, './param', 'name', 'playerID').attrib['value'],
65 playerKey = find_xpath_attr(object_doc, './param', 'name', 'playerKey')
66 # Not all pages define this value
67 if playerKey is not None:
68 params['playerKey'] = playerKey.attrib['value']
69 videoPlayer = find_xpath_attr(object_doc, './param', 'name', '@videoPlayer')
70 if videoPlayer is not None:
71 params['@videoPlayer'] = videoPlayer.attrib['value']
72 data = compat_urllib_parse.urlencode(params)
73 return cls._FEDERATED_URL_TEMPLATE % data
76 def _extract_brightcove_url(cls, webpage):
77 """Try to extract the brightcove url from the wepbage, returns None
80 m_brightcove = re.search(
81 r'<object[^>]+?class=([\'"])[^>]*?BrightcoveExperience.*?\1.+?</object>',
83 if m_brightcove is not None:
84 return cls._build_brighcove_url(m_brightcove.group())
88 def _real_extract(self, url):
89 mobj = re.match(self._VALID_URL, url)
90 query_str = mobj.group('query')
91 query = compat_urlparse.parse_qs(query_str)
93 videoPlayer = query.get('@videoPlayer')
95 return self._get_video_info(videoPlayer[0], query_str)
97 player_key = query['playerKey']
98 return self._get_playlist_info(player_key[0])
100 def _get_video_info(self, video_id, query):
101 request_url = self._FEDERATED_URL_TEMPLATE % query
102 webpage = self._download_webpage(request_url, video_id)
104 self.report_extraction(video_id)
105 info = self._search_regex(r'var experienceJSON = ({.*?});', webpage, 'json')
106 info = json.loads(info)['data']
107 video_info = info['programmedContent']['videoPlayer']['mediaDTO']
109 return self._extract_video_info(video_info)
111 def _get_playlist_info(self, player_key):
112 playlist_info = self._download_webpage(self._PLAYLIST_URL_TEMPLATE % player_key,
113 player_key, u'Downloading playlist information')
115 json_data = json.loads(playlist_info)
116 if 'videoList' not in json_data:
117 raise ExtractorError(u'Empty playlist')
118 playlist_info = json_data['videoList']
119 videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']]
121 return self.playlist_result(videos, playlist_id=playlist_info['id'],
122 playlist_title=playlist_info['mediaCollectionDTO']['displayName'])
124 def _extract_video_info(self, video_info):
126 'id': compat_str(video_info['id']),
127 'title': video_info['displayName'],
128 'description': video_info.get('shortDescription'),
129 'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'),
130 'uploader': video_info.get('publisherName'),
133 renditions = video_info.get('renditions')
135 renditions = sorted(renditions, key=lambda r: r['size'])
136 best_format = renditions[-1]
138 'url': best_format['defaultURL'],
140 elif video_info.get('FLVFullLengthURL') is not None:
142 'url': video_info['FLVFullLengthURL'],
145 raise ExtractorError(u'Unable to extract video url for %s' % info['id'])