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