_ 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|classic)\.)?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         'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
  56         'only_matching': True,
  57     }]
  58
  59     @classmethod
  60     def suitable(cls, url):
  61         return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)
  62
  63     def _extract_media_info(self, media_info_url, webpage, video_id):
  64         media_info = self._download_json(
  65             media_info_url, video_id, 'Downloading media JSON')
  66
  67         formats = self._extract_formats(media_info, video_id)
  68
  69         if not formats:
  70             if '"fsk"' in webpage:
  71                 raise ExtractorError(
  72                     'This video is only available after 20:00', expected=True)
  73             elif media_info.get('_geoblocked'):
  74                 raise ExtractorError('This video is not available due to geo restriction', expected=True)
  75
  76         self._sort_formats(formats)
  77
  78         duration = int_or_none(media_info.get('_duration'))
  79         thumbnail = media_info.get('_previewImage')
  80         is_live = media_info.get('_isLive') is True
  81
  82         subtitles = {}
  83         subtitle_url = media_info.get('_subtitleUrl')
  84         if subtitle_url:
  85             subtitles['de'] = [{
  86                 'ext': 'ttml',
  87                 'url': subtitle_url,
  88             }]
  89
  90         return {
  91             'id': video_id,
  92             'duration': duration,
  93             'thumbnail': thumbnail,
  94             'is_live': is_live,
  95             'formats': formats,
  96             'subtitles': subtitles,
  97         }
  98
  99     def _extract_formats(self, media_info, video_id):
 100         type_ = media_info.get('_type')
 101         media_array = media_info.get('_mediaArray', [])
 102         formats = []
 103         for num, media in enumerate(media_array):
 104             for stream in media.get('_mediaStreamArray', []):
 105                 stream_urls = stream.get('_stream')
 106                 if not stream_urls:
 107                     continue
 108                 if not isinstance(stream_urls, list):
 109                     stream_urls = [stream_urls]
 110                 quality = stream.get('_quality')
 111                 server = stream.get('_server')
 112                 for stream_url in stream_urls:
 113                     if not url_or_none(stream_url):
 114                         continue
 115                     ext = determine_ext(stream_url)
 116                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
 117                         continue
 118                     if ext == 'f4m':
 119                         formats.extend(self._extract_f4m_formats(
 120                             update_url_query(stream_url, {
 121                                 'hdcore': '3.1.1',
 122                                 'plugin': 'aasp-3.1.1.69.124'
 123                             }),
 124                             video_id, f4m_id='hds', fatal=False))
 125                     elif ext == 'm3u8':
 126                         formats.extend(self._extract_m3u8_formats(
 127                             stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
 128                     else:
 129                         if server and server.startswith('rtmp'):
 130                             f = {
 131                                 'url': server,
 132                                 'play_path': stream_url,
 133                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
 134                             }
 135                         else:
 136                             f = {
 137                                 'url': stream_url,
 138                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 139                             }
 140                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
 141                         if m:
 142                             f.update({
 143                                 'width': int(m.group('width')),
 144                                 'height': int(m.group('height')),
 145                             })
 146                         if type_ == 'audio':
 147                             f['vcodec'] = 'none'
 148                         formats.append(f)
 149         return formats
 150
 151     def _real_extract(self, url):
 152         # determine video id from url
 153         m = re.match(self._VALID_URL, url)
 154
 155         document_id = None
 156
 157         numid = re.search(r'documentId=([0-9]+)', url)
 158         if numid:
 159             document_id = video_id = numid.group(1)
 160         else:
 161             video_id = m.group('video_id')
 162
 163         webpage = self._download_webpage(url, video_id)
 164
 165         ERRORS = (
 166             ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
 167             ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
 168              'Video %s is no longer available'),
 169         )
 170
 171         for pattern, message in ERRORS:
 172             if pattern in webpage:
 173                 raise ExtractorError(message % video_id, expected=True)
 174
 175         if re.search(r'[\?&]rss($|[=&])', url):
 176             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 177             if doc.tag == 'rss':
 178                 return GenericIE()._extract_rss(url, video_id, doc)
 179
 180         title = self._html_search_regex(
 181             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 182              r'<meta name="dcterms\.title" content="(.*?)"/>',
 183              r'<h4 class="headline">(.*?)</h4>',
 184              r'<title[^>]*>(.*?)</title>'],
 185             webpage, 'title')
 186         description = self._html_search_meta(
 187             'dcterms.abstract', webpage, 'description', default=None)
 188         if description is None:
 189             description = self._html_search_meta(
 190                 'description', webpage, 'meta description', default=None)
 191         if description is None:
 192             description = self._html_search_regex(
 193                 r'<p\s+class="teasertext">(.+?)</p>',
 194                 webpage, 'teaser text', default=None)
 195
 196         # Thumbnail is sometimes not present.
 197         # It is in the mobile version, but that seems to use a different URL
 198         # structure altogether.
 199         thumbnail = self._og_search_thumbnail(webpage, default=None)
 200
 201         media_streams = re.findall(r'''(?x)
 202             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 203             "([^"]+)"''', webpage)
 204
 205         if media_streams:
 206             QUALITIES = qualities(['lo', 'hi', 'hq'])
 207             formats = []
 208             for furl in set(media_streams):
 209                 if furl.endswith('.f4m'):
 210                     fid = 'f4m'
 211                 else:
 212                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 213                     fid = fid_m.group(1) if fid_m else None
 214                 formats.append({
 215                     'quality': QUALITIES(fid),
 216                     'format_id': fid,
 217                     'url': furl,
 218                 })
 219             self._sort_formats(formats)
 220             info = {
 221                 'formats': formats,
 222             }
 223         else:  # request JSON file
 224             if not document_id:
 225                 video_id = self._search_regex(
 226                     r'/play/(?:config|media)/(\d+)', webpage, 'media id')
 227             info = self._extract_media_info(
 228                 'http://www.ardmediathek.de/play/media/%s' % video_id,
 229                 webpage, video_id)
 230
 231         info.update({
 232             'id': video_id,
 233             'title': self._live_title(title) if info.get('is_live') else title,
 234             'description': description,
 235             'thumbnail': thumbnail,
 236         })
 237
 238         return info
 239
 240
 241 class ARDIE(InfoExtractor):
 242     _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
 243     _TESTS = [{
 244         # available till 14.02.2019
 245         'url': 'http://www.daserste.de/information/talk/maischberger/videos/das-groko-drama-zerlegen-sich-die-volksparteien-video-102.html',
 246         'md5': '8e4ec85f31be7c7fc08a26cdbc5a1f49',
 247         'info_dict': {
 248             'display_id': 'das-groko-drama-zerlegen-sich-die-volksparteien-video',
 249             'id': '102',
 250             'ext': 'mp4',
 251             'duration': 4435.0,
 252             'title': 'Das GroKo-Drama: Zerlegen sich die Volksparteien?',
 253             'upload_date': '20180214',
 254             'thumbnail': r're:^https?://.*\.jpg$',
 255         },
 256     }, {
 257         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 258         'only_matching': True,
 259     }]
 260
 261     def _real_extract(self, url):
 262         mobj = re.match(self._VALID_URL, url)
 263         display_id = mobj.group('display_id')
 264
 265         player_url = mobj.group('mainurl') + '~playerXml.xml'
 266         doc = self._download_xml(player_url, display_id)
 267         video_node = doc.find('./video')
 268         upload_date = unified_strdate(xpath_text(
 269             video_node, './broadcastDate'))
 270         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 271
 272         formats = []
 273         for a in video_node.findall('.//asset'):
 274             f = {
 275                 'format_id': a.attrib['type'],
 276                 'width': int_or_none(a.find('./frameWidth').text),
 277                 'height': int_or_none(a.find('./frameHeight').text),
 278                 'vbr': int_or_none(a.find('./bitrateVideo').text),
 279                 'abr': int_or_none(a.find('./bitrateAudio').text),
 280                 'vcodec': a.find('./codecVideo').text,
 281                 'tbr': int_or_none(a.find('./totalBitrate').text),
 282             }
 283             if a.find('./serverPrefix').text:
 284                 f['url'] = a.find('./serverPrefix').text
 285                 f['playpath'] = a.find('./fileName').text
 286             else:
 287                 f['url'] = a.find('./fileName').text
 288             formats.append(f)
 289         self._sort_formats(formats)
 290
 291         return {
 292             'id': mobj.group('id'),
 293             'formats': formats,
 294             'display_id': display_id,
 295             'title': video_node.find('./title').text,
 296             'duration': parse_duration(video_node.find('./duration').text),
 297             'upload_date': upload_date,
 298             'thumbnail': thumbnail,
 299         }
 300
 301
 302 class ARDBetaMediathekIE(InfoExtractor):
 303     _VALID_URL = r'https://(?:beta|www)\.ardmediathek\.de/[^/]+/(?:player|live)/(?P<video_id>[a-zA-Z0-9]+)(?:/(?P<display_id>[^/?#]+))?'
 304     _TESTS = [{
 305         'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
 306         'md5': '2d02d996156ea3c397cfc5036b5d7f8f',
 307         'info_dict': {
 308             'display_id': 'die-robuste-roswita',
 309             'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
 310             'title': 'Tatort: Die robuste Roswita',
 311             'description': r're:^Der Mord.*trüber ist als die Ilm.',
 312             'duration': 5316,
 313             'thumbnail': 'https://img.ardmediathek.de/standard/00/55/43/59/34/-1774185891/16x9/960?mandant=ard',
 314             'upload_date': '20180826',
 315             'ext': 'mp4',
 316         },
 317     }, {
 318         'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3N3ci5kZS9hZXgvbzEwNzE5MTU/',
 319         'only_matching': True,
 320     }, {
 321         'url': 'https://www.ardmediathek.de/swr/live/Y3JpZDovL3N3ci5kZS8xMzQ4MTA0Mg',
 322         'only_matching': True,
 323     }]
 324
 325     def _real_extract(self, url):
 326         mobj = re.match(self._VALID_URL, url)
 327         video_id = mobj.group('video_id')
 328         display_id = mobj.group('display_id') or video_id
 329
 330         webpage = self._download_webpage(url, display_id)
 331         data_json = self._search_regex(r'window\.__APOLLO_STATE__\s*=\s*(\{.*);\n', webpage, 'json')
 332         data = self._parse_json(data_json, display_id)
 333
 334         res = {
 335             'id': video_id,
 336             'display_id': display_id,
 337         }
 338         formats = []
 339         for widget in data.values():
 340             if widget.get('_geoblocked'):
 341                 raise ExtractorError('This video is not available due to geoblocking', expected=True)
 342
 343             if '_duration' in widget:
 344                 res['duration'] = widget['_duration']
 345             if 'clipTitle' in widget:
 346                 res['title'] = widget['clipTitle']
 347             if '_previewImage' in widget:
 348                 res['thumbnail'] = widget['_previewImage']
 349             if 'broadcastedOn' in widget:
 350                 res['upload_date'] = unified_strdate(widget['broadcastedOn'])
 351             if 'synopsis' in widget:
 352                 res['description'] = widget['synopsis']
 353             if '_subtitleUrl' in widget:
 354                 res['subtitles'] = {'de': [{
 355                     'ext': 'ttml',
 356                     'url': widget['_subtitleUrl'],
 357                 }]}
 358             if '_quality' in widget:
 359                 format_url = widget['_stream']['json'][0]
 360
 361                 if format_url.endswith('.f4m'):
 362                     formats.extend(self._extract_f4m_formats(
 363                         format_url + '?hdcore=3.11.0',
 364                         video_id, f4m_id='hds', fatal=False))
 365                 elif format_url.endswith('m3u8'):
 366                     formats.extend(self._extract_m3u8_formats(
 367                         format_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
 368                 else:
 369                     formats.append({
 370                         'format_id': 'http-' + widget['_quality'],
 371                         'url': format_url,
 372                         'preference': 10,  # Plain HTTP, that's nice
 373                     })
 374
 375         self._sort_formats(formats)
 376         res['formats'] = formats
 377
 378         return res