_ Git - youtube-dl/blob - 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     ExtractorError,
  10     unified_strdate,
  11 )
  12
  13
  14 class DreiSatIE(InfoExtractor):
  15     IE_NAME = '3sat'
  16     _VALID_URL = r'(?:http://)?(?:www\.)?3sat.de/mediathek/index.php\?(?:(?:mode|display)=[^&]+&)*obj=(?P<id>[0-9]+)$'
  17     _TEST = {
  18         u"url": u"http://www.3sat.de/mediathek/index.php?obj=36983",
  19         u'file': u'36983.webm',
  20         u'md5': u'57c97d0469d71cf874f6815aa2b7c944',
  21         u'info_dict': {
  22             u"title": u"Kaffeeland Schweiz",
  23             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...",
  24             u"uploader": u"3sat",
  25             u"upload_date": u"20130622"
  26         }
  27     }
  28
  29
  30     def _real_extract(self, url):
  31         mobj = re.match(self._VALID_URL, url)
  32         video_id = mobj.group('id')
  33         details_url = 'http://www.3sat.de/mediathek/xmlservice/web/beitragsDetails?ak=web&id=%s' % video_id
  34         details_xml = self._download_webpage(details_url, video_id, note=u'Downloading video details')
  35         details_doc = xml.etree.ElementTree.fromstring(details_xml.encode('utf-8'))
  36
  37         thumbnail_els = details_doc.findall('.//teaserimage')
  38         thumbnails = [{
  39             'width': te.attrib['key'].partition('x')[0],
  40             'height': te.attrib['key'].partition('x')[2],
  41             'url': te.text,
  42         } for te in thumbnail_els]
  43
  44         information_el = details_doc.find('.//information')
  45         video_title = information_el.find('./title').text
  46         video_description = information_el.find('./detail').text
  47
  48         details_el = details_doc.find('.//details')
  49         video_uploader = details_el.find('./channel').text
  50         upload_date = unified_strdate(details_el.find('./airtime').text)
  51
  52         format_els = details_doc.findall('.//formitaet')
  53         formats = [{
  54             'format_id': fe.attrib['basetype'],
  55             'width': int(fe.find('./width').text),
  56             'height': int(fe.find('./height').text),
  57             'url': 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             'id': video_id,
  72             'title': video_title,
  73             'formats': formats,
  74             'description': video_description,
  75             'thumbnails': thumbnails,
  76             'thumbnail': thumbnails[-1]['url'],
  77             'uploader': video_uploader,
  78             'upload_date': upload_date,
  79         }
  80
  81         # TODO: Remove when #980 has been merged
  82         info['url'] = formats[-1]['url']
  83         info['ext'] = determine_ext(formats[-1]['url'])
  84
  85         return self.video_result(info)