git.bitcoin.ninja 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)/(?:.*/)(?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         # audio
  42         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
  43         'only_matching': True,
  44     }, {
  45         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
  46         'only_matching': True,
  47     }, {
  48         # audio
  49         'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
  50         'only_matching': True,
  51     }]
  52
  53     def _extract_media_info(self, media_info_url, webpage, video_id):
  54         media_info = self._download_json(
  55             media_info_url, video_id, 'Downloading media JSON')
  56
  57         formats = self._extract_formats(media_info, video_id)
  58
  59         if not formats:
  60             if '"fsk"' in webpage:
  61                 raise ExtractorError(
  62                     'This video is only available after 20:00', expected=True)
  63             elif media_info.get('_geoblocked'):
  64                 raise ExtractorError('This video is not available due to geo restriction', expected=True)
  65
  66         self._sort_formats(formats)
  67
  68         duration = int_or_none(media_info.get('_duration'))
  69         thumbnail = media_info.get('_previewImage')
  70         is_live = media_info.get('_isLive') is True
  71
  72         subtitles = {}
  73         subtitle_url = media_info.get('_subtitleUrl')
  74         if subtitle_url:
  75             subtitles['de'] = [{
  76                 'ext': 'ttml',
  77                 'url': subtitle_url,
  78             }]
  79
  80         return {
  81             'id': video_id,
  82             'duration': duration,
  83             'thumbnail': thumbnail,
  84             'is_live': is_live,
  85             'formats': formats,
  86             'subtitles': subtitles,
  87         }
  88
  89     def _extract_formats(self, media_info, video_id):
  90         type_ = media_info.get('_type')
  91         media_array = media_info.get('_mediaArray', [])
  92         formats = []
  93         for num, media in enumerate(media_array):
  94             for stream in media.get('_mediaStreamArray', []):
  95                 stream_urls = stream.get('_stream')
  96                 if not stream_urls:
  97                     continue
  98                 if not isinstance(stream_urls, list):
  99                     stream_urls = [stream_urls]
 100                 quality = stream.get('_quality')
 101                 server = stream.get('_server')
 102                 for stream_url in stream_urls:
 103                     if not url_or_none(stream_url):
 104                         continue
 105                     ext = determine_ext(stream_url)
 106                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
 107                         continue
 108                     if ext == 'f4m':
 109                         formats.extend(self._extract_f4m_formats(
 110                             update_url_query(stream_url, {
 111                                 'hdcore': '3.1.1',
 112                                 'plugin': 'aasp-3.1.1.69.124'
 113                             }),
 114                             video_id, f4m_id='hds', fatal=False))
 115                     elif ext == 'm3u8':
 116                         formats.extend(self._extract_m3u8_formats(
 117                             stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
 118                     else:
 119                         if server and server.startswith('rtmp'):
 120                             f = {
 121                                 'url': server,
 122                                 'play_path': stream_url,
 123                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
 124                             }
 125                         else:
 126                             f = {
 127                                 'url': stream_url,
 128                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 129                             }
 130                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
 131                         if m:
 132                             f.update({
 133                                 'width': int(m.group('width')),
 134                                 'height': int(m.group('height')),
 135                             })
 136                         if type_ == 'audio':
 137                             f['vcodec'] = 'none'
 138                         formats.append(f)
 139         return formats
 140
 141     def _real_extract(self, url):
 142         # determine video id from url
 143         m = re.match(self._VALID_URL, url)
 144
 145         document_id = None
 146
 147         numid = re.search(r'documentId=([0-9]+)', url)
 148         if numid:
 149             document_id = video_id = numid.group(1)
 150         else:
 151             video_id = m.group('video_id')
 152
 153         webpage = self._download_webpage(url, video_id)
 154
 155         ERRORS = (
 156             ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
 157             ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
 158              'Video %s is no longer available'),
 159         )
 160
 161         for pattern, message in ERRORS:
 162             if pattern in webpage:
 163                 raise ExtractorError(message % video_id, expected=True)
 164
 165         if re.search(r'[\?&]rss($|[=&])', url):
 166             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 167             if doc.tag == 'rss':
 168                 return GenericIE()._extract_rss(url, video_id, doc)
 169
 170         title = self._html_search_regex(
 171             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 172              r'<meta name="dcterms\.title" content="(.*?)"/>',
 173              r'<h4 class="headline">(.*?)</h4>'],
 174             webpage, 'title')
 175         description = self._html_search_meta(
 176             'dcterms.abstract', webpage, 'description', default=None)
 177         if description is None:
 178             description = self._html_search_meta(
 179                 'description', webpage, 'meta description')
 180
 181         # Thumbnail is sometimes not present.
 182         # It is in the mobile version, but that seems to use a different URL
 183         # structure altogether.
 184         thumbnail = self._og_search_thumbnail(webpage, default=None)
 185
 186         media_streams = re.findall(r'''(?x)
 187             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 188             "([^"]+)"''', webpage)
 189
 190         if media_streams:
 191             QUALITIES = qualities(['lo', 'hi', 'hq'])
 192             formats = []
 193             for furl in set(media_streams):
 194                 if furl.endswith('.f4m'):
 195                     fid = 'f4m'
 196                 else:
 197                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 198                     fid = fid_m.group(1) if fid_m else None
 199                 formats.append({
 200                     'quality': QUALITIES(fid),
 201                     'format_id': fid,
 202                     'url': furl,
 203                 })
 204             self._sort_formats(formats)
 205             info = {
 206                 'formats': formats,
 207             }
 208         else:  # request JSON file
 209             if not document_id:
 210                 video_id = self._search_regex(
 211                     r'/play/(?:config|media)/(\d+)', webpage, 'media id')
 212             info = self._extract_media_info(
 213                 'http://www.ardmediathek.de/play/media/%s' % video_id,
 214                 webpage, video_id)
 215
 216         info.update({
 217             'id': video_id,
 218             'title': self._live_title(title) if info.get('is_live') else title,
 219             'description': description,
 220             'thumbnail': thumbnail,
 221         })
 222
 223         return info
 224
 225
 226 class ARDIE(InfoExtractor):
 227     _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
 228     _TESTS = [{
 229         # available till 14.02.2019
 230         'url': 'http://www.daserste.de/information/talk/maischberger/videos/das-groko-drama-zerlegen-sich-die-volksparteien-video-102.html',
 231         'md5': '8e4ec85f31be7c7fc08a26cdbc5a1f49',
 232         'info_dict': {
 233             'display_id': 'das-groko-drama-zerlegen-sich-die-volksparteien-video',
 234             'id': '102',
 235             'ext': 'mp4',
 236             'duration': 4435.0,
 237             'title': 'Das GroKo-Drama: Zerlegen sich die Volksparteien?',
 238             'upload_date': '20180214',
 239             'thumbnail': r're:^https?://.*\.jpg$',
 240         },
 241     }, {
 242         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 243         'only_matching': True,
 244     }]
 245
 246     def _real_extract(self, url):
 247         mobj = re.match(self._VALID_URL, url)
 248         display_id = mobj.group('display_id')
 249
 250         player_url = mobj.group('mainurl') + '~playerXml.xml'
 251         doc = self._download_xml(player_url, display_id)
 252         video_node = doc.find('./video')
 253         upload_date = unified_strdate(xpath_text(
 254             video_node, './broadcastDate'))
 255         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 256
 257         formats = []
 258         for a in video_node.findall('.//asset'):
 259             f = {
 260                 'format_id': a.attrib['type'],
 261                 'width': int_or_none(a.find('./frameWidth').text),
 262                 'height': int_or_none(a.find('./frameHeight').text),
 263                 'vbr': int_or_none(a.find('./bitrateVideo').text),
 264                 'abr': int_or_none(a.find('./bitrateAudio').text),
 265                 'vcodec': a.find('./codecVideo').text,
 266                 'tbr': int_or_none(a.find('./totalBitrate').text),
 267             }
 268             if a.find('./serverPrefix').text:
 269                 f['url'] = a.find('./serverPrefix').text
 270                 f['playpath'] = a.find('./fileName').text
 271             else:
 272                 f['url'] = a.find('./fileName').text
 273             formats.append(f)
 274         self._sort_formats(formats)
 275
 276         return {
 277             'id': mobj.group('id'),
 278             'formats': formats,
 279             'display_id': display_id,
 280             'title': video_node.find('./title').text,
 281             'duration': parse_duration(video_node.find('./duration').text),
 282             'upload_date': upload_date,
 283             'thumbnail': thumbnail,
 284         }
 285
 286
 287 class ARDBetaMediathekIE(InfoExtractor):
 288     _VALID_URL = r'https://beta\.ardmediathek\.de/[a-z]+/player/(?P<video_id>[a-zA-Z0-9]+)/(?P<display_id>[^/?#]+)'
 289     _TESTS = [{
 290         'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
 291         'md5': '2d02d996156ea3c397cfc5036b5d7f8f',
 292         'info_dict': {
 293             'display_id': 'die-robuste-roswita',
 294             'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
 295             'title': 'Tatort: Die robuste Roswita',
 296             'description': r're:^Der Mord.*trüber ist als die Ilm.',
 297             'duration': 5316,
 298             'thumbnail': 'https://img.ardmediathek.de/standard/00/55/43/59/34/-1774185891/16x9/960?mandant=ard',
 299             'upload_date': '20180826',
 300             'ext': 'mp4',
 301         },
 302     }]
 303
 304     def _real_extract(self, url):
 305         mobj = re.match(self._VALID_URL, url)
 306         video_id = mobj.group('video_id')
 307         display_id = mobj.group('display_id')
 308
 309         webpage = self._download_webpage(url, display_id)
 310         data_json = self._search_regex(r'window\.__APOLLO_STATE__\s*=\s*(\{.*);\n', webpage, 'json')
 311         data = self._parse_json(data_json, display_id)
 312
 313         res = {
 314             'id': video_id,
 315             'display_id': display_id,
 316         }
 317         formats = []
 318         for widget in data.values():
 319             if widget.get('_geoblocked'):
 320                 raise ExtractorError('This video is not available due to geoblocking', expected=True)
 321
 322             if '_duration' in widget:
 323                 res['duration'] = widget['_duration']
 324             if 'clipTitle' in widget:
 325                 res['title'] = widget['clipTitle']
 326             if '_previewImage' in widget:
 327                 res['thumbnail'] = widget['_previewImage']
 328             if 'broadcastedOn' in widget:
 329                 res['upload_date'] = unified_strdate(widget['broadcastedOn'])
 330             if 'synopsis' in widget:
 331                 res['description'] = widget['synopsis']
 332             if '_subtitleUrl' in widget:
 333                 res['subtitles'] = {'de': [{
 334                     'ext': 'ttml',
 335                     'url': widget['_subtitleUrl'],
 336                 }]}
 337             if '_quality' in widget:
 338                 format_url = widget['_stream']['json'][0]
 339
 340                 if format_url.endswith('.f4m'):
 341                     # Skip f4m - these URLs just return a 403
 342                     formats.append({
 343                         'format_id': 'f4m-' + widget['_quality'],
 344                         'url': format_url,
 345                         'preference': -1001,
 346                     })
 347                 elif format_url.endswith('m3u8'):
 348                     formats.extend(self._extract_m3u8_formats(
 349                         format_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
 350                 else:
 351                     formats.append({
 352                         'format_id': 'http-' + widget['_quality'],
 353                         'url': format_url,
 354                         'preference': 10,  # Plain HTTP, that's nice
 355                     })
 356
 357         self._sort_formats(formats)
 358         res['formats'] = formats
 359
 360         return res