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