class NFBIE(InfoExtractor):
IE_NAME = 'nfb'
IE_DESC = 'National Film Board of Canada'
- _VALID_URL = r'https?://(?:www\.)?nfb\.ca/film/(?P<id>[\da-z_-]+)'
+ _VALID_URL = r'https?://(?:www\.)?(nfb|onf)\.ca/film/(?P<id>[\da-z_-]+)'
_TEST = {
'url': 'https://www.nfb.ca/film/qallunaat_why_white_people_are_funny',
mobj = re.match(self._VALID_URL, url)
video_id = mobj.group('id')
- page = self._download_webpage(url, video_id, 'Downloading film page')
+ page = self._download_webpage('https://www.nfb.ca/film/%s' % video_id, video_id, 'Downloading film page')
uploader_id = self._html_search_regex(r'<a class="director-link" href="/explore-all-directors/([^/]+)/"',
page, 'director id', fatal=False)
page, 'director name', fatal=False)
request = compat_urllib_request.Request('https://www.nfb.ca/film/%s/player_config' % video_id,
- compat_urllib_parse.urlencode({'getConfig': 'true'}))
+ compat_urllib_parse.urlencode({'getConfig': 'true'}).encode('ascii'))
request.add_header('Content-Type', 'application/x-www-form-urlencoded')
request.add_header('X-NFB-Referer', 'http://www.nfb.ca/medias/flash/NFBVideoPlayer.swf')
config = self._download_xml(request, video_id, 'Downloading player config XML')
- thumbnail = config.find("./player/stream/media[@type='posterImage']/assets/asset[@quality='high']/default/url").text
- video = config.find("./player/stream/media[@type='video']")
- duration = int(video.get('duration'))
- title = video.find('title').text
- description = video.find('description').text
+ title = None
+ description = None
+ thumbnail = None
+ duration = None
+ formats = []
- # It seems assets always go from lower to better quality, so no need to sort
- formats = [{
- 'url': x.find('default/streamerURI').text + '/',
- 'play_path': x.find('default/url').text,
- 'rtmp_live': False,
- 'ext': 'mp4',
- 'format_id': x.get('quality'),
- } for x in video.findall('assets/asset')]
+ def extract_thumbnail(media):
+ thumbnails = {}
+ for asset in media.findall('assets/asset'):
+ thumbnails[asset.get('quality')] = asset.find('default/url').text
+ if not thumbnails:
+ return None
+ if 'high' in thumbnails:
+ return thumbnails['high']
+ return list(thumbnails.values())[0]
+
+ for media in config.findall('./player/stream/media'):
+ if media.get('type') == 'posterImage':
+ thumbnail = extract_thumbnail(media)
+ elif media.get('type') == 'video':
+ duration = int(media.get('duration'))
+ title = media.find('title').text
+ description = media.find('description').text
+ # It seems assets always go from lower to better quality, so no need to sort
+ formats = [{
+ 'url': x.find('default/streamerURI').text,
+ 'app': x.find('default/streamerURI').text.split('/', 3)[3],
+ 'play_path': x.find('default/url').text,
+ 'rtmp_live': False,
+ 'ext': 'mp4',
+ 'format_id': x.get('quality'),
+ } for x in media.findall('assets/asset')]
return {
'id': video_id,