youtube_dl/extractor/ard.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     determine_ext,
   9     ExtractorError,
  10     qualities,
  11     compat_urllib_parse_urlparse,
  12     compat_urllib_parse,
  13 )
  14
  15
  16 class ARDIE(InfoExtractor):
  17     _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
  18
  19     _TESTS = [{
  20         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
  21         'file': '22429276.mp4',
  22         'md5': '469751912f1de0816a9fc9df8336476c',
  23         'info_dict': {
  24             'title': 'Vertrauen ist gut, Spionieren ist besser - Geht so deutsch-amerikanische Freundschaft?',
  25             'description': 'Das Erste Mediathek [ARD]: Vertrauen ist gut, Spionieren ist besser - Geht so deutsch-amerikanische Freundschaft?, Anne Will, Über die Spionage-Affäre diskutieren Clemens Binninger, Katrin Göring-Eckardt, Georg Mascolo, Andrew B. Denison und Constanze Kurz.. Das Video zur Sendung Anne Will am Mittwoch, 16.07.2014',
  26         },
  27         'skip': 'Blocked outside of Germany',
  28     }, {
  29         'url': 'http://www.ardmediathek.de/tv/Tatort/Das-Wunder-von-Wolbeck-Video-tgl-ab-20/Das-Erste/Video?documentId=22490580&bcastId=602916',
  30         'info_dict': {
  31             'id': '22490580',
  32             'ext': 'mp4',
  33             'title': 'Das Wunder von Wolbeck (Video tgl. ab 20 Uhr)',
  34             'description': 'Auf einem restaurierten Hof bei Wolbeck wird der Heilpraktiker Raffael Lembeck eines morgens von seiner Frau Stella tot aufgefunden. Das Opfer war offensichtlich in seiner Praxis zu Fall gekommen und ist dann verblutet, erklärt Prof. Boerne am Tatort.',
  35         },
  36         'skip': 'Blocked outside of Germany',
  37     }]
  38
  39     def _real_extract(self, url):
  40         # determine video id from url
  41         m = re.match(self._VALID_URL, url)
  42
  43         numid = re.search(r'documentId=([0-9]+)', url)
  44         if numid:
  45             video_id = numid.group(1)
  46         else:
  47             video_id = m.group('video_id')
  48
  49         urlp = compat_urllib_parse_urlparse(url)
  50         url = urlp._replace(path=compat_urllib_parse.quote(urlp.path.encode('utf-8'))).geturl()
  51
  52         webpage = self._download_webpage(url, video_id)
  53
  54         if '>Der gewünschte Beitrag ist nicht mehr verfügbar.<' in webpage:
  55             raise ExtractorError('Video %s is no longer available' % video_id, expected=True)
  56
  57         title = self._html_search_regex(
  58             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
  59              r'<meta name="dcterms.title" content="(.*?)"/>',
  60              r'<h4 class="headline">(.*?)</h4>'],
  61             webpage, 'title')
  62         description = self._html_search_meta(
  63             'dcterms.abstract', webpage, 'description', default=None)
  64         if description is None:
  65             description = self._html_search_meta(
  66                 'description', webpage, 'meta description')
  67
  68         # Thumbnail is sometimes not present.
  69         # It is in the mobile version, but that seems to use a different URL
  70         # structure altogether.
  71         thumbnail = self._og_search_thumbnail(webpage, default=None)
  72
  73         media_streams = re.findall(r'''(?x)
  74             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
  75             "([^"]+)"''', webpage)
  76
  77         if media_streams:
  78             QUALITIES = qualities(['lo', 'hi', 'hq'])
  79             formats = []
  80             for furl in set(media_streams):
  81                 if furl.endswith('.f4m'):
  82                     fid = 'f4m'
  83                 else:
  84                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
  85                     fid = fid_m.group(1) if fid_m else None
  86                 formats.append({
  87                     'quality': QUALITIES(fid),
  88                     'format_id': fid,
  89                     'url': furl,
  90                 })
  91         else:  # request JSON file
  92             media_info = self._download_json(
  93                 'http://www.ardmediathek.de/play/media/%s' % video_id, video_id)
  94             # The second element of the _mediaArray contains the standard http urls
  95             streams = media_info['_mediaArray'][1]['_mediaStreamArray']
  96             if not streams:
  97                 if '"fsk"' in webpage:
  98                     raise ExtractorError('This video is only available after 20:00')
  99
 100             formats = []
 101             for s in streams:
 102                 if type(s['_stream']) == list:
 103                     for index, url in enumerate(s['_stream'][::-1]):
 104                         quality = s['_quality'] + index
 105                         formats.append({
 106                             'quality': quality,
 107                             'url': url,
 108                             'format_id': '%s-%s' % (determine_ext(url), quality)
 109                         })
 110                     continue
 111
 112                 format = {
 113                     'quality': s['_quality'],
 114                     'url': s['_stream'],
 115                 }
 116
 117                 format['format_id'] = '%s-%s' % (
 118                     determine_ext(format['url']), format['quality'])
 119
 120                 formats.append(format)
 121
 122         self._sort_formats(formats)
 123
 124         return {
 125             'id': video_id,
 126             'title': title,
 127             'description': description,
 128             'formats': formats,
 129             'thumbnail': thumbnail,
 130         }