youtube_dl/extractor/nfb.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     compat_urllib_request,
   8     compat_urllib_parse,
   9 )
  10
  11
  12 class NFBIE(InfoExtractor):
  13     IE_NAME = 'nfb'
  14     IE_DESC = 'National Film Board of Canada'
  15     _VALID_URL = r'https?://(?:www\.)?(nfb|onf)\.ca/film/(?P<id>[\da-z_-]+)'
  16
  17     _TEST = {
  18         'url': 'https://www.nfb.ca/film/qallunaat_why_white_people_are_funny',
  19         'info_dict': {
  20             'id': 'qallunaat_why_white_people_are_funny',
  21             'ext': 'mp4',
  22             'title': 'Qallunaat! Why White People Are Funny ',
  23             'description': 'md5:836d8aff55e087d04d9f6df554d4e038',
  24             'duration': 3128,
  25             'uploader': 'Mark Sandiford',
  26             'uploader_id': 'mark-sandiford',
  27         },
  28         'params': {
  29             # rtmp download
  30             'skip_download': True,
  31         }
  32     }
  33
  34     def _real_extract(self, url):
  35         mobj = re.match(self._VALID_URL, url)
  36         video_id = mobj.group('id')
  37
  38         page = self._download_webpage('https://www.nfb.ca/film/%s' % video_id, video_id, 'Downloading film page')
  39
  40         uploader_id = self._html_search_regex(r'<a class="director-link" href="/explore-all-directors/([^/]+)/"',
  41             page, 'director id', fatal=False)
  42         uploader = self._html_search_regex(r'<em class="director-name" itemprop="name">([^<]+)</em>',
  43             page, 'director name', fatal=False)
  44
  45         request = compat_urllib_request.Request('https://www.nfb.ca/film/%s/player_config' % video_id,
  46             compat_urllib_parse.urlencode({'getConfig': 'true'}).encode('ascii'))
  47         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  48         request.add_header('X-NFB-Referer', 'http://www.nfb.ca/medias/flash/NFBVideoPlayer.swf')
  49
  50         config = self._download_xml(request, video_id, 'Downloading player config XML')
  51
  52         title = None
  53         description = None
  54         thumbnail = None
  55         duration = None
  56         formats = []
  57
  58         def extract_thumbnail(media):
  59             thumbnails = {}
  60             for asset in media.findall('assets/asset'):
  61                 thumbnails[asset.get('quality')] = asset.find('default/url').text
  62             if not thumbnails:
  63                 return None
  64             if 'high' in thumbnails:
  65                 return thumbnails['high']
  66             return list(thumbnails.values())[0]
  67
  68         for media in config.findall('./player/stream/media'):
  69             if media.get('type') == 'posterImage':
  70                 thumbnail = extract_thumbnail(media)
  71             elif media.get('type') == 'video':
  72                 duration = int(media.get('duration'))
  73                 title = media.find('title').text
  74                 description = media.find('description').text
  75                 # It seems assets always go from lower to better quality, so no need to sort
  76                 for asset in media.findall('assets/asset'):
  77                     for x in asset:
  78                         formats.append({
  79                             'url': x.find('streamerURI').text,
  80                             'app': x.find('streamerURI').text.split('/', 3)[3],
  81                             'play_path': x.find('url').text,
  82                             'rtmp_live': False,
  83                             'ext': 'mp4',
  84                             'format_id': '%s-%s' % (x.tag, asset.get('quality')),
  85                         })
  86
  87         return {
  88             'id': video_id,
  89             'title': title,
  90             'description': description,
  91             'thumbnail': thumbnail,
  92             'duration': duration,
  93             'uploader': uploader,
  94             'uploader_id': uploader_id,
  95             'formats': formats,
  96         }