youtube_dl/extractor/dreisat.py

   1 # coding: utf-8
   2
   3 import re
   4 import xml.etree.ElementTree
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     determine_ext,
   9     unified_strdate,
  10 )
  11
  12
  13 class DreiSatIE(InfoExtractor):
  14     IE_NAME = '3sat'
  15     _VALID_URL = r'(?:http://)?(?:www\.)?3sat.de/mediathek/index.php\?(?:(?:mode|display)=[^&]+&)*obj=(?P<id>[0-9]+)$'
  16     _TEST = {
  17         u"url": u"http://www.3sat.de/mediathek/index.php?obj=36983",
  18         u'file': u'36983.webm',
  19         u'md5': u'57c97d0469d71cf874f6815aa2b7c944',
  20         u'info_dict': {
  21             u"title": u"Kaffeeland Schweiz",
  22             u"description": u"Über 80 Kaffeeröstereien liefern in der Schweiz das Getränk, in das das Land so vernarrt ist: Mehr als 1000 Tassen trinkt ein Schweizer pro Jahr. SCHWEIZWEIT nimmt die Kaffeekultur unter die...",
  23             u"uploader": u"3sat",
  24             u"upload_date": u"20130622"
  25         }
  26     }
  27
  28
  29     def _real_extract(self, url):
  30         mobj = re.match(self._VALID_URL, url)
  31         video_id = mobj.group('id')
  32         details_url = 'http://www.3sat.de/mediathek/xmlservice/web/beitragsDetails?ak=web&id=%s' % video_id
  33         details_xml = self._download_webpage(details_url, video_id, note=u'Downloading video details')
  34         details_doc = xml.etree.ElementTree.fromstring(details_xml.encode('utf-8'))
  35
  36         thumbnail_els = details_doc.findall('.//teaserimage')
  37         thumbnails = [{
  38             'width': te.attrib['key'].partition('x')[0],
  39             'height': te.attrib['key'].partition('x')[2],
  40             'url': te.text,
  41         } for te in thumbnail_els]
  42
  43         information_el = details_doc.find('.//information')
  44         video_title = information_el.find('./title').text
  45         video_description = information_el.find('./detail').text
  46
  47         details_el = details_doc.find('.//details')
  48         video_uploader = details_el.find('./channel').text
  49         upload_date = unified_strdate(details_el.find('./airtime').text)
  50
  51         format_els = details_doc.findall('.//formitaet')
  52         formats = [{
  53             'format_id': fe.attrib['basetype'],
  54             'width': int(fe.find('./width').text),
  55             'height': int(fe.find('./height').text),
  56             'url': fe.find('./url').text,
  57             'ext': determine_ext(fe.find('./url').text),
  58             'filesize': int(fe.find('./filesize').text),
  59             'video_bitrate': int(fe.find('./videoBitrate').text),
  60             '3sat_qualityname': fe.find('./quality').text,
  61         } for fe in format_els
  62             if not fe.find('./url').text.startswith('http://www.metafilegenerator.de/')]
  63
  64         def _sortkey(format):
  65             qidx = ['low', 'med', 'high', 'veryhigh'].index(format['3sat_qualityname'])
  66             prefer_http = 1 if 'rtmp' in format['url'] else 0
  67             return (qidx, prefer_http, format['video_bitrate'])
  68         formats.sort(key=_sortkey)
  69
  70         info = {
  71             '_type': 'video',
  72             'id': video_id,
  73             'title': video_title,
  74             'formats': formats,
  75             'description': video_description,
  76             'thumbnails': thumbnails,
  77             'thumbnail': thumbnails[-1]['url'],
  78             'uploader': video_uploader,
  79             'upload_date': upload_date,
  80         }
  81
  82         # TODO: Remove when #980 has been merged
  83         info.update(formats[-1])
  84
  85         return info