_ 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     qualities,
  12     int_or_none,
  13     parse_duration,
  14     unified_strdate,
  15     xpath_text,
  16     update_url_query,
  17     url_or_none,
  18 )
  19 from ..compat import compat_etree_fromstring
  20
  21
  22 class ARDMediathekIE(InfoExtractor):
  23     IE_NAME = 'ARD:mediathek'
  24     _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
  25
  26     _TESTS = [{
  27         # available till 26.07.2022
  28         'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',
  29         'info_dict': {
  30             'id': '44726822',
  31             'ext': 'mp4',
  32             'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
  33             'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
  34             'duration': 1740,
  35         },
  36         'params': {
  37             # m3u8 download
  38             'skip_download': True,
  39         }
  40     }, {
  41         'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
  42         'only_matching': True,
  43     }, {
  44         # audio
  45         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
  46         'only_matching': True,
  47     }, {
  48         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
  49         'only_matching': True,
  50     }, {
  51         # audio
  52         'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
  53         'only_matching': True,
  54     }]
  55
  56     def _extract_media_info(self, media_info_url, webpage, video_id):
  57         media_info = self._download_json(
  58             media_info_url, video_id, 'Downloading media JSON')
  59
  60         formats = self._extract_formats(media_info, video_id)
  61
  62         if not formats:
  63             if '"fsk"' in webpage:
  64                 raise ExtractorError(
  65                     'This video is only available after 20:00', expected=True)
  66             elif media_info.get('_geoblocked'):
  67                 raise ExtractorError('This video is not available due to geo restriction', expected=True)
  68
  69         self._sort_formats(formats)
  70
  71         duration = int_or_none(media_info.get('_duration'))
  72         thumbnail = media_info.get('_previewImage')
  73         is_live = media_info.get('_isLive') is True
  74
  75         subtitles = {}
  76         subtitle_url = media_info.get('_subtitleUrl')
  77         if subtitle_url:
  78             subtitles['de'] = [{
  79                 'ext': 'ttml',
  80                 'url': subtitle_url,
  81             }]
  82
  83         return {
  84             'id': video_id,
  85             'duration': duration,
  86             'thumbnail': thumbnail,
  87             'is_live': is_live,
  88             'formats': formats,
  89             'subtitles': subtitles,
  90         }
  91
  92     def _extract_formats(self, media_info, video_id):
  93         type_ = media_info.get('_type')
  94         media_array = media_info.get('_mediaArray', [])
  95         formats = []
  96         for num, media in enumerate(media_array):
  97             for stream in media.get('_mediaStreamArray', []):
  98                 stream_urls = stream.get('_stream')
  99                 if not stream_urls:
 100                     continue
 101                 if not isinstance(stream_urls, list):
 102                     stream_urls = [stream_urls]
 103                 quality = stream.get('_quality')
 104                 server = stream.get('_server')
 105                 for stream_url in stream_urls:
 106                     if not url_or_none(stream_url):
 107                         continue
 108                     ext = determine_ext(stream_url)
 109                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
 110                         continue
 111                     if ext == 'f4m':
 112                         formats.extend(self._extract_f4m_formats(
 113                             update_url_query(stream_url, {
 114                                 'hdcore': '3.1.1',
 115                                 'plugin': 'aasp-3.1.1.69.124'
 116                             }),
 117                             video_id, f4m_id='hds', fatal=False))
 118                     elif ext == 'm3u8':
 119                         formats.extend(self._extract_m3u8_formats(
 120                             stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
 121                     else:
 122                         if server and server.startswith('rtmp'):
 123                             f = {
 124                                 'url': server,
 125                                 'play_path': stream_url,
 126                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
 127                             }
 128                         else:
 129                             f = {
 130                                 'url': stream_url,
 131                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 132                             }
 133                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
 134                         if m:
 135                             f.update({
 136                                 'width': int(m.group('width')),
 137                                 'height': int(m.group('height')),
 138                             })
 139                         if type_ == 'audio':
 140                             f['vcodec'] = 'none'
 141                         formats.append(f)
 142         return formats
 143
 144     def _real_extract(self, url):
 145         # determine video id from url
 146         m = re.match(self._VALID_URL, url)
 147
 148         document_id = None
 149
 150         numid = re.search(r'documentId=([0-9]+)', url)
 151         if numid:
 152             document_id = video_id = numid.group(1)
 153         else:
 154             video_id = m.group('video_id')
 155
 156         webpage = self._download_webpage(url, video_id)
 157
 158         ERRORS = (
 159             ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
 160             ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
 161              'Video %s is no longer available'),
 162         )
 163
 164         for pattern, message in ERRORS:
 165             if pattern in webpage:
 166                 raise ExtractorError(message % video_id, expected=True)
 167
 168         if re.search(r'[\?&]rss($|[=&])', url):
 169             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 170             if doc.tag == 'rss':
 171                 return GenericIE()._extract_rss(url, video_id, doc)
 172
 173         title = self._html_search_regex(
 174             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 175              r'<meta name="dcterms\.title" content="(.*?)"/>',
 176              r'<h4 class="headline">(.*?)</h4>',
 177              r'<title[^>]*>(.*?)</title>'],
 178             webpage, 'title')
 179         description = self._html_search_meta(
 180             'dcterms.abstract', webpage, 'description', default=None)
 181         if description is None:
 182             description = self._html_search_meta(
 183                 'description', webpage, 'meta description', default=None)
 184         if description is None:
 185             description = self._html_search_regex(
 186                 r'<p\s+class="teasertext">(.+?)</p>',
 187                 webpage, 'teaser text', default=None)
 188
 189         # Thumbnail is sometimes not present.
 190         # It is in the mobile version, but that seems to use a different URL
 191         # structure altogether.
 192         thumbnail = self._og_search_thumbnail(webpage, default=None)
 193
 194         media_streams = re.findall(r'''(?x)
 195             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 196             "([^"]+)"''', webpage)
 197
 198         if media_streams:
 199             QUALITIES = qualities(['lo', 'hi', 'hq'])
 200             formats = []
 201             for furl in set(media_streams):
 202                 if furl.endswith('.f4m'):
 203                     fid = 'f4m'
 204                 else:
 205                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 206                     fid = fid_m.group(1) if fid_m else None
 207                 formats.append({
 208                     'quality': QUALITIES(fid),
 209                     'format_id': fid,
 210                     'url': furl,
 211                 })
 212             self._sort_formats(formats)
 213             info = {
 214                 'formats': formats,
 215             }
 216         else:  # request JSON file
 217             if not document_id:
 218                 video_id = self._search_regex(
 219                     r'/play/(?:config|media)/(\d+)', webpage, 'media id')
 220             info = self._extract_media_info(
 221                 'http://www.ardmediathek.de/play/media/%s' % video_id,
 222                 webpage, video_id)
 223
 224         info.update({
 225             'id': video_id,
 226             'title': self._live_title(title) if info.get('is_live') else title,
 227             'description': description,
 228             'thumbnail': thumbnail,
 229         })
 230
 231         return info
 232
 233
 234 class ARDIE(InfoExtractor):
 235     _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
 236     _TESTS = [{
 237         # available till 14.02.2019
 238         'url': 'http://www.daserste.de/information/talk/maischberger/videos/das-groko-drama-zerlegen-sich-die-volksparteien-video-102.html',
 239         'md5': '8e4ec85f31be7c7fc08a26cdbc5a1f49',
 240         'info_dict': {
 241             'display_id': 'das-groko-drama-zerlegen-sich-die-volksparteien-video',
 242             'id': '102',
 243             'ext': 'mp4',
 244             'duration': 4435.0,
 245             'title': 'Das GroKo-Drama: Zerlegen sich die Volksparteien?',
 246             'upload_date': '20180214',
 247             'thumbnail': r're:^https?://.*\.jpg$',
 248         },
 249     }, {
 250         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 251         'only_matching': True,
 252     }]
 253
 254     def _real_extract(self, url):
 255         mobj = re.match(self._VALID_URL, url)
 256         display_id = mobj.group('display_id')
 257
 258         player_url = mobj.group('mainurl') + '~playerXml.xml'
 259         doc = self._download_xml(player_url, display_id)
 260         video_node = doc.find('./video')
 261         upload_date = unified_strdate(xpath_text(
 262             video_node, './broadcastDate'))
 263         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 264
 265         formats = []
 266         for a in video_node.findall('.//asset'):
 267             f = {
 268                 'format_id': a.attrib['type'],
 269                 'width': int_or_none(a.find('./frameWidth').text),
 270                 'height': int_or_none(a.find('./frameHeight').text),
 271                 'vbr': int_or_none(a.find('./bitrateVideo').text),
 272                 'abr': int_or_none(a.find('./bitrateAudio').text),
 273                 'vcodec': a.find('./codecVideo').text,
 274                 'tbr': int_or_none(a.find('./totalBitrate').text),
 275             }
 276             if a.find('./serverPrefix').text:
 277                 f['url'] = a.find('./serverPrefix').text
 278                 f['playpath'] = a.find('./fileName').text
 279             else:
 280                 f['url'] = a.find('./fileName').text
 281             formats.append(f)
 282         self._sort_formats(formats)
 283
 284         return {
 285             'id': mobj.group('id'),
 286             'formats': formats,
 287             'display_id': display_id,
 288             'title': video_node.find('./title').text,
 289             'duration': parse_duration(video_node.find('./duration').text),
 290             'upload_date': upload_date,
 291             'thumbnail': thumbnail,
 292         }
 293
 294
 295 class ARDBetaMediathekIE(InfoExtractor):
 296     _VALID_URL = r'https://beta\.ardmediathek\.de/[a-z]+/player/(?P<video_id>[a-zA-Z0-9]+)/(?P<display_id>[^/?#]+)'
 297     _TESTS = [{
 298         'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
 299         'md5': '2d02d996156ea3c397cfc5036b5d7f8f',
 300         'info_dict': {
 301             'display_id': 'die-robuste-roswita',
 302             'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
 303             'title': 'Tatort: Die robuste Roswita',
 304             'description': r're:^Der Mord.*trüber ist als die Ilm.',
 305             'duration': 5316,
 306             'thumbnail': 'https://img.ardmediathek.de/standard/00/55/43/59/34/-1774185891/16x9/960?mandant=ard',
 307             'upload_date': '20180826',
 308             'ext': 'mp4',
 309         },
 310     }]
 311
 312     def _real_extract(self, url):
 313         mobj = re.match(self._VALID_URL, url)
 314         video_id = mobj.group('video_id')
 315         display_id = mobj.group('display_id')
 316
 317         webpage = self._download_webpage(url, display_id)
 318         data_json = self._search_regex(r'window\.__APOLLO_STATE__\s*=\s*(\{.*);\n', webpage, 'json')
 319         data = self._parse_json(data_json, display_id)
 320
 321         res = {
 322             'id': video_id,
 323             'display_id': display_id,
 324         }
 325         formats = []
 326         for widget in data.values():
 327             if widget.get('_geoblocked'):
 328                 raise ExtractorError('This video is not available due to geoblocking', expected=True)
 329
 330             if '_duration' in widget:
 331                 res['duration'] = widget['_duration']
 332             if 'clipTitle' in widget:
 333                 res['title'] = widget['clipTitle']
 334             if '_previewImage' in widget:
 335                 res['thumbnail'] = widget['_previewImage']
 336             if 'broadcastedOn' in widget:
 337                 res['upload_date'] = unified_strdate(widget['broadcastedOn'])
 338             if 'synopsis' in widget:
 339                 res['description'] = widget['synopsis']
 340             if '_subtitleUrl' in widget:
 341                 res['subtitles'] = {'de': [{
 342                     'ext': 'ttml',
 343                     'url': widget['_subtitleUrl'],
 344                 }]}
 345             if '_quality' in widget:
 346                 format_url = widget['_stream']['json'][0]
 347
 348                 if format_url.endswith('.f4m'):
 349                     formats.extend(self._extract_f4m_formats(
 350                         format_url + '?hdcore=3.11.0',
 351                         video_id, f4m_id='hds', fatal=False))
 352                 elif format_url.endswith('m3u8'):
 353                     formats.extend(self._extract_m3u8_formats(
 354                         format_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
 355                 else:
 356                     formats.append({
 357                         'format_id': 'http-' + widget['_quality'],
 358                         'url': format_url,
 359                         'preference': 10,  # Plain HTTP, that's nice
 360                     })
 361
 362         self._sort_formats(formats)
 363         res['formats'] = formats
 364
 365         return res