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     get_element_by_attribute,
  12     qualities,
  13     int_or_none,
  14     parse_duration,
  15     unified_strdate,
  16     xpath_text,
  17 )
  18 from ..compat import compat_etree_fromstring
  19
  20
  21 class ARDMediathekIE(InfoExtractor):
  22     IE_NAME = 'ARD:mediathek'
  23     _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
  24
  25     _TESTS = [{
  26         'url': 'http://www.ardmediathek.de/tv/Dokumentation-und-Reportage/Ich-liebe-das-Leben-trotzdem/rbb-Fernsehen/Video?documentId=29582122&bcastId=3822114',
  27         'info_dict': {
  28             'id': '29582122',
  29             'ext': 'mp4',
  30             'title': 'Ich liebe das Leben trotzdem',
  31             'description': 'md5:45e4c225c72b27993314b31a84a5261c',
  32             'duration': 4557,
  33         },
  34         'params': {
  35             # m3u8 download
  36             'skip_download': True,
  37         },
  38     }, {
  39         'url': 'http://www.ardmediathek.de/tv/Tatort/Tatort-Scheinwelten-H%C3%B6rfassung-Video/Das-Erste/Video?documentId=29522730&bcastId=602916',
  40         'md5': 'f4d98b10759ac06c0072bbcd1f0b9e3e',
  41         'info_dict': {
  42             'id': '29522730',
  43             'ext': 'mp4',
  44             'title': 'Tatort: Scheinwelten - Hörfassung (Video tgl. ab 20 Uhr)',
  45             'description': 'md5:196392e79876d0ac94c94e8cdb2875f1',
  46             'duration': 5252,
  47         },
  48     }, {
  49         # audio
  50         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
  51         'md5': '219d94d8980b4f538c7fcb0865eb7f2c',
  52         'info_dict': {
  53             'id': '28488308',
  54             'ext': 'mp3',
  55             'title': 'Tod eines Fußballers',
  56             'description': 'md5:f6e39f3461f0e1f54bfa48c8875c86ef',
  57             'duration': 3240,
  58         },
  59     }, {
  60         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
  61         'only_matching': True,
  62     }]
  63
  64     def _extract_media_info(self, media_info_url, webpage, video_id):
  65         media_info = self._download_json(
  66             media_info_url, video_id, 'Downloading media JSON')
  67
  68         formats = self._extract_formats(media_info, video_id)
  69
  70         if not formats:
  71             if '"fsk"' in webpage:
  72                 raise ExtractorError(
  73                     'This video is only available after 20:00', expected=True)
  74             elif media_info.get('_geoblocked'):
  75                 raise ExtractorError('This video is not available due to geo restriction', expected=True)
  76
  77         self._sort_formats(formats)
  78
  79         duration = int_or_none(media_info.get('_duration'))
  80         thumbnail = media_info.get('_previewImage')
  81
  82         subtitles = {}
  83         subtitle_url = media_info.get('_subtitleUrl')
  84         if subtitle_url:
  85             subtitles['de'] = [{
  86                 'ext': 'srt',
  87                 'url': subtitle_url,
  88             }]
  89
  90         return {
  91             'id': video_id,
  92             'duration': duration,
  93             'thumbnail': thumbnail,
  94             'formats': formats,
  95             'subtitles': subtitles,
  96         }
  97
  98     def _extract_formats(self, media_info, video_id):
  99         type_ = media_info.get('_type')
 100         media_array = media_info.get('_mediaArray', [])
 101         formats = []
 102         for num, media in enumerate(media_array):
 103             for stream in media.get('_mediaStreamArray', []):
 104                 stream_urls = stream.get('_stream')
 105                 if not stream_urls:
 106                     continue
 107                 if not isinstance(stream_urls, list):
 108                     stream_urls = [stream_urls]
 109                 quality = stream.get('_quality')
 110                 server = stream.get('_server')
 111                 for stream_url in stream_urls:
 112                     ext = determine_ext(stream_url)
 113                     if ext == 'f4m':
 114                         formats.extend(self._extract_f4m_formats(
 115                             stream_url + '?hdcore=3.1.1&plugin=aasp-3.1.1.69.124',
 116                             video_id, preference=-1, f4m_id='hds'))
 117                     elif ext == 'm3u8':
 118                         formats.extend(self._extract_m3u8_formats(
 119                             stream_url, video_id, 'mp4', preference=1, m3u8_id='hls'))
 120                     else:
 121                         if server and server.startswith('rtmp'):
 122                             f = {
 123                                 'url': server,
 124                                 'play_path': stream_url,
 125                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
 126                             }
 127                         elif stream_url.startswith('http'):
 128                             f = {
 129                                 'url': stream_url,
 130                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 131                             }
 132                         else:
 133                             continue
 134                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
 135                         if m:
 136                             f.update({
 137                                 'width': int(m.group('width')),
 138                                 'height': int(m.group('height')),
 139                             })
 140                         if type_ == 'audio':
 141                             f['vcodec'] = 'none'
 142                         formats.append(f)
 143         return formats
 144
 145     def _real_extract(self, url):
 146         # determine video id from url
 147         m = re.match(self._VALID_URL, url)
 148
 149         numid = re.search(r'documentId=([0-9]+)', url)
 150         if numid:
 151             video_id = numid.group(1)
 152         else:
 153             video_id = m.group('video_id')
 154
 155         webpage = self._download_webpage(url, video_id)
 156
 157         if '>Der gewünschte Beitrag ist nicht mehr verfügbar.<' in webpage:
 158             raise ExtractorError('Video %s is no longer available' % video_id, expected=True)
 159
 160         if 'Diese Sendung ist für Jugendliche unter 12 Jahren nicht geeignet. Der Clip ist deshalb nur von 20 bis 6 Uhr verfügbar.' in webpage:
 161             raise ExtractorError('This program is only suitable for those aged 12 and older. Video %s is therefore only available between 20 pm and 6 am.' % video_id, expected=True)
 162
 163         if re.search(r'[\?&]rss($|[=&])', url):
 164             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 165             if doc.tag == 'rss':
 166                 return GenericIE()._extract_rss(url, video_id, doc)
 167
 168         title = self._html_search_regex(
 169             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 170              r'<meta name="dcterms.title" content="(.*?)"/>',
 171              r'<h4 class="headline">(.*?)</h4>'],
 172             webpage, 'title')
 173         description = self._html_search_meta(
 174             'dcterms.abstract', webpage, 'description', default=None)
 175         if description is None:
 176             description = self._html_search_meta(
 177                 'description', webpage, 'meta description')
 178
 179         # Thumbnail is sometimes not present.
 180         # It is in the mobile version, but that seems to use a different URL
 181         # structure altogether.
 182         thumbnail = self._og_search_thumbnail(webpage, default=None)
 183
 184         media_streams = re.findall(r'''(?x)
 185             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 186             "([^"]+)"''', webpage)
 187
 188         if media_streams:
 189             QUALITIES = qualities(['lo', 'hi', 'hq'])
 190             formats = []
 191             for furl in set(media_streams):
 192                 if furl.endswith('.f4m'):
 193                     fid = 'f4m'
 194                 else:
 195                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 196                     fid = fid_m.group(1) if fid_m else None
 197                 formats.append({
 198                     'quality': QUALITIES(fid),
 199                     'format_id': fid,
 200                     'url': furl,
 201                 })
 202             self._sort_formats(formats)
 203             info = {
 204                 'formats': formats,
 205             }
 206         else:  # request JSON file
 207             info = self._extract_media_info(
 208                 'http://www.ardmediathek.de/play/media/%s' % video_id, webpage, video_id)
 209
 210         info.update({
 211             'id': video_id,
 212             'title': title,
 213             'description': description,
 214             'thumbnail': thumbnail,
 215         })
 216
 217         return info
 218
 219
 220 class ARDIE(InfoExtractor):
 221     _VALID_URL = '(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
 222     _TEST = {
 223         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 224         'md5': 'd216c3a86493f9322545e045ddc3eb35',
 225         'info_dict': {
 226             'display_id': 'die-story-im-ersten-mission-unter-falscher-flagge',
 227             'id': '100',
 228             'ext': 'mp4',
 229             'duration': 2600,
 230             'title': 'Die Story im Ersten: Mission unter falscher Flagge',
 231             'upload_date': '20140804',
 232             'thumbnail': 're:^https?://.*\.jpg$',
 233         }
 234     }
 235
 236     def _real_extract(self, url):
 237         mobj = re.match(self._VALID_URL, url)
 238         display_id = mobj.group('display_id')
 239
 240         player_url = mobj.group('mainurl') + '~playerXml.xml'
 241         doc = self._download_xml(player_url, display_id)
 242         video_node = doc.find('./video')
 243         upload_date = unified_strdate(xpath_text(
 244             video_node, './broadcastDate'))
 245         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 246
 247         formats = []
 248         for a in video_node.findall('.//asset'):
 249             f = {
 250                 'format_id': a.attrib['type'],
 251                 'width': int_or_none(a.find('./frameWidth').text),
 252                 'height': int_or_none(a.find('./frameHeight').text),
 253                 'vbr': int_or_none(a.find('./bitrateVideo').text),
 254                 'abr': int_or_none(a.find('./bitrateAudio').text),
 255                 'vcodec': a.find('./codecVideo').text,
 256                 'tbr': int_or_none(a.find('./totalBitrate').text),
 257             }
 258             if a.find('./serverPrefix').text:
 259                 f['url'] = a.find('./serverPrefix').text
 260                 f['playpath'] = a.find('./fileName').text
 261             else:
 262                 f['url'] = a.find('./fileName').text
 263             formats.append(f)
 264         self._sort_formats(formats)
 265
 266         return {
 267             'id': mobj.group('id'),
 268             'formats': formats,
 269             'display_id': display_id,
 270             'title': video_node.find('./title').text,
 271             'duration': parse_duration(video_node.find('./duration').text),
 272             'upload_date': upload_date,
 273             'thumbnail': thumbnail,
 274         }
 275
 276
 277 class SportschauIE(ARDMediathekIE):
 278     IE_NAME = 'Sportschau'
 279     _VALID_URL = r'(?P<baseurl>https?://(?:www\.)?sportschau\.de/(?:[^/]+/)+video(?P<id>[^/#?]+))\.html'
 280     _TESTS = [{
 281         'url': 'http://www.sportschau.de/tourdefrance/videoseppeltkokainhatnichtsmitklassischemdopingzutun100.html',
 282         'info_dict': {
 283             'id': 'seppeltkokainhatnichtsmitklassischemdopingzutun100',
 284             'ext': 'mp4',
 285             'title': 'Seppelt: "Kokain hat nichts mit klassischem Doping zu tun"',
 286             'thumbnail': 're:^https?://.*\.jpg$',
 287             'description': 'Der ARD-Doping Experte Hajo Seppelt gibt seine Einschätzung zum ersten Dopingfall der diesjährigen Tour de France um den Italiener Luca Paolini ab.',
 288         },
 289         'params': {
 290             # m3u8 download
 291             'skip_download': True,
 292         },
 293     }]
 294
 295     def _real_extract(self, url):
 296         mobj = re.match(self._VALID_URL, url)
 297         video_id = mobj.group('id')
 298         base_url = mobj.group('baseurl')
 299
 300         webpage = self._download_webpage(url, video_id)
 301         title = get_element_by_attribute('class', 'headline', webpage)
 302         description = self._html_search_meta('description', webpage, 'description')
 303
 304         info = self._extract_media_info(
 305             base_url + '-mc_defaultQuality-h.json', webpage, video_id)
 306
 307         info.update({
 308             'title': title,
 309             'description': description,
 310         })
 311
 312         return info