Raphaël G. Git Repositories - youtubedl/blob - 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 .generic import GenericIE
   8 from ..utils import (
   9     determine_ext,
  10     ExtractorError,
  11     qualities,
  12     int_or_none,
  13     parse_duration,
  14     unified_strdate,
  15     xpath_text,
  16     parse_xml,
  17 )
  18
  19
  20 class ARDMediathekIE(InfoExtractor):
  21     IE_NAME = 'ARD:mediathek'
  22     _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
  23
  24     _TESTS = [{
  25         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
  26         'file': '22429276.mp4',
  27         'md5': '469751912f1de0816a9fc9df8336476c',
  28         'info_dict': {
  29             'title': 'Vertrauen ist gut, Spionieren ist besser - Geht so deutsch-amerikanische Freundschaft?',
  30             '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',
  31         },
  32         'skip': 'Blocked outside of Germany',
  33     }, {
  34         'url': 'http://www.ardmediathek.de/tv/Tatort/Das-Wunder-von-Wolbeck-Video-tgl-ab-20/Das-Erste/Video?documentId=22490580&bcastId=602916',
  35         'info_dict': {
  36             'id': '22490580',
  37             'ext': 'mp4',
  38             'title': 'Das Wunder von Wolbeck (Video tgl. ab 20 Uhr)',
  39             '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.',
  40         },
  41         'skip': 'Blocked outside of Germany',
  42     }]
  43
  44     def _real_extract(self, url):
  45         # determine video id from url
  46         m = re.match(self._VALID_URL, url)
  47
  48         numid = re.search(r'documentId=([0-9]+)', url)
  49         if numid:
  50             video_id = numid.group(1)
  51         else:
  52             video_id = m.group('video_id')
  53
  54         webpage = self._download_webpage(url, video_id)
  55
  56         if '>Der gewünschte Beitrag ist nicht mehr verfügbar.<' in webpage:
  57             raise ExtractorError('Video %s is no longer available' % video_id, expected=True)
  58
  59         if re.search(r'[\?&]rss($|[=&])', url):
  60             doc = parse_xml(webpage)
  61             if doc.tag == 'rss':
  62                 return GenericIE()._extract_rss(url, video_id, doc)
  63
  64         title = self._html_search_regex(
  65             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
  66              r'<meta name="dcterms.title" content="(.*?)"/>',
  67              r'<h4 class="headline">(.*?)</h4>'],
  68             webpage, 'title')
  69         description = self._html_search_meta(
  70             'dcterms.abstract', webpage, 'description', default=None)
  71         if description is None:
  72             description = self._html_search_meta(
  73                 'description', webpage, 'meta description')
  74
  75         # Thumbnail is sometimes not present.
  76         # It is in the mobile version, but that seems to use a different URL
  77         # structure altogether.
  78         thumbnail = self._og_search_thumbnail(webpage, default=None)
  79
  80         media_streams = re.findall(r'''(?x)
  81             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
  82             "([^"]+)"''', webpage)
  83
  84         if media_streams:
  85             QUALITIES = qualities(['lo', 'hi', 'hq'])
  86             formats = []
  87             for furl in set(media_streams):
  88                 if furl.endswith('.f4m'):
  89                     fid = 'f4m'
  90                 else:
  91                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
  92                     fid = fid_m.group(1) if fid_m else None
  93                 formats.append({
  94                     'quality': QUALITIES(fid),
  95                     'format_id': fid,
  96                     'url': furl,
  97                 })
  98         else:  # request JSON file
  99             media_info = self._download_json(
 100                 'http://www.ardmediathek.de/play/media/%s' % video_id, video_id)
 101             # The second element of the _mediaArray contains the standard http urls
 102             streams = media_info['_mediaArray'][1]['_mediaStreamArray']
 103             if not streams:
 104                 if '"fsk"' in webpage:
 105                     raise ExtractorError('This video is only available after 20:00')
 106
 107             formats = []
 108             for s in streams:
 109                 if type(s['_stream']) == list:
 110                     for index, url in enumerate(s['_stream'][::-1]):
 111                         quality = s['_quality'] + index
 112                         formats.append({
 113                             'quality': quality,
 114                             'url': url,
 115                             'format_id': '%s-%s' % (determine_ext(url), quality)
 116                         })
 117                     continue
 118
 119                 format = {
 120                     'quality': s['_quality'],
 121                     'url': s['_stream'],
 122                 }
 123
 124                 format['format_id'] = '%s-%s' % (
 125                     determine_ext(format['url']), format['quality'])
 126
 127                 formats.append(format)
 128
 129         self._sort_formats(formats)
 130
 131         return {
 132             'id': video_id,
 133             'title': title,
 134             'description': description,
 135             'formats': formats,
 136             'thumbnail': thumbnail,
 137         }
 138
 139
 140 class ARDIE(InfoExtractor):
 141     _VALID_URL = '(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
 142     _TEST = {
 143         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 144         'md5': 'd216c3a86493f9322545e045ddc3eb35',
 145         'info_dict': {
 146             'display_id': 'die-story-im-ersten-mission-unter-falscher-flagge',
 147             'id': '100',
 148             'ext': 'mp4',
 149             'duration': 2600,
 150             'title': 'Die Story im Ersten: Mission unter falscher Flagge',
 151             'upload_date': '20140804',
 152             'thumbnail': 're:^https?://.*\.jpg$',
 153         }
 154     }
 155
 156     def _real_extract(self, url):
 157         mobj = re.match(self._VALID_URL, url)
 158         display_id = mobj.group('display_id')
 159
 160         player_url = mobj.group('mainurl') + '~playerXml.xml'
 161         doc = self._download_xml(player_url, display_id)
 162         video_node = doc.find('./video')
 163         upload_date = unified_strdate(xpath_text(
 164             video_node, './broadcastDate'))
 165         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 166
 167         formats = []
 168         for a in video_node.findall('.//asset'):
 169             f = {
 170                 'format_id': a.attrib['type'],
 171                 'width': int_or_none(a.find('./frameWidth').text),
 172                 'height': int_or_none(a.find('./frameHeight').text),
 173                 'vbr': int_or_none(a.find('./bitrateVideo').text),
 174                 'abr': int_or_none(a.find('./bitrateAudio').text),
 175                 'vcodec': a.find('./codecVideo').text,
 176                 'tbr': int_or_none(a.find('./totalBitrate').text),
 177             }
 178             if a.find('./serverPrefix').text:
 179                 f['url'] = a.find('./serverPrefix').text
 180                 f['playpath'] = a.find('./fileName').text
 181             else:
 182                 f['url'] = a.find('./fileName').text
 183             formats.append(f)
 184         self._sort_formats(formats)
 185
 186         return {
 187             'id': mobj.group('id'),
 188             'formats': formats,
 189             'display_id': display_id,
 190             'title': video_node.find('./title').text,
 191             'duration': parse_duration(video_node.find('./duration').text),
 192             'upload_date': upload_date,
 193             'thumbnail': thumbnail,
 194         }