]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/nfb.py
5bd15f7a72f5aeb49d91391e11ddffaa1a52f44f
[youtubedl] / youtube_dl / extractor / nfb.py
1 from __future__ import unicode_literals
2
3 from .common import InfoExtractor
4 from ..compat import compat_urllib_parse
5 from ..utils import sanitized_Request
6
7
8 class NFBIE(InfoExtractor):
9 IE_NAME = 'nfb'
10 IE_DESC = 'National Film Board of Canada'
11 _VALID_URL = r'https?://(?:www\.)?(?:nfb|onf)\.ca/film/(?P<id>[\da-z_-]+)'
12
13 _TEST = {
14 'url': 'https://www.nfb.ca/film/qallunaat_why_white_people_are_funny',
15 'info_dict': {
16 'id': 'qallunaat_why_white_people_are_funny',
17 'ext': 'mp4',
18 'title': 'Qallunaat! Why White People Are Funny ',
19 'description': 'md5:836d8aff55e087d04d9f6df554d4e038',
20 'duration': 3128,
21 'uploader': 'Mark Sandiford',
22 'uploader_id': 'mark-sandiford',
23 },
24 'params': {
25 # rtmp download
26 'skip_download': True,
27 }
28 }
29
30 def _real_extract(self, url):
31 video_id = self._match_id(url)
32 page = self._download_webpage(
33 'https://www.nfb.ca/film/%s' % video_id, video_id,
34 'Downloading film page')
35
36 uploader_id = self._html_search_regex(r'<a class="director-link" href="/explore-all-directors/([^/]+)/"',
37 page, 'director id', fatal=False)
38 uploader = self._html_search_regex(r'<em class="director-name" itemprop="name">([^<]+)</em>',
39 page, 'director name', fatal=False)
40
41 request = sanitized_Request(
42 'https://www.nfb.ca/film/%s/player_config' % video_id,
43 compat_urllib_parse.urlencode({'getConfig': 'true'}).encode('ascii'))
44 request.add_header('Content-Type', 'application/x-www-form-urlencoded')
45 request.add_header('X-NFB-Referer', 'http://www.nfb.ca/medias/flash/NFBVideoPlayer.swf')
46
47 config = self._download_xml(request, video_id, 'Downloading player config XML')
48
49 title = None
50 description = None
51 thumbnail = None
52 duration = None
53 formats = []
54
55 def extract_thumbnail(media):
56 thumbnails = {}
57 for asset in media.findall('assets/asset'):
58 thumbnails[asset.get('quality')] = asset.find('default/url').text
59 if not thumbnails:
60 return None
61 if 'high' in thumbnails:
62 return thumbnails['high']
63 return list(thumbnails.values())[0]
64
65 for media in config.findall('./player/stream/media'):
66 if media.get('type') == 'posterImage':
67 thumbnail = extract_thumbnail(media)
68 elif media.get('type') == 'video':
69 duration = int(media.get('duration'))
70 title = media.find('title').text
71 description = media.find('description').text
72 # It seems assets always go from lower to better quality, so no need to sort
73 for asset in media.findall('assets/asset'):
74 for x in asset:
75 formats.append({
76 'url': x.find('streamerURI').text,
77 'app': x.find('streamerURI').text.split('/', 3)[3],
78 'play_path': x.find('url').text,
79 'rtmp_live': False,
80 'ext': 'mp4',
81 'format_id': '%s-%s' % (x.tag, asset.get('quality')),
82 })
83
84 return {
85 'id': video_id,
86 'title': title,
87 'description': description,
88 'thumbnail': thumbnail,
89 'duration': duration,
90 'uploader': uploader,
91 'uploader_id': uploader_id,
92 'formats': formats,
93 }