Raphaël G. Git Repositories - youtubedl/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 )
  18 from ..compat import compat_etree_fromstring
  19
  20
  21 class ARDMediathekIE(InfoExtractor):
  22     IE_NAME = 'ARD:mediathek'
  23     _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.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         'skip': 'HTTP Error 404: Not Found',
  39     }, {
  40         'url': 'http://www.ardmediathek.de/tv/Tatort/Tatort-Scheinwelten-H%C3%B6rfassung-Video/Das-Erste/Video?documentId=29522730&bcastId=602916',
  41         'md5': 'f4d98b10759ac06c0072bbcd1f0b9e3e',
  42         'info_dict': {
  43             'id': '29522730',
  44             'ext': 'mp4',
  45             'title': 'Tatort: Scheinwelten - Hörfassung (Video tgl. ab 20 Uhr)',
  46             'description': 'md5:196392e79876d0ac94c94e8cdb2875f1',
  47             'duration': 5252,
  48         },
  49         'skip': 'HTTP Error 404: Not Found',
  50     }, {
  51         # audio
  52         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
  53         'md5': '219d94d8980b4f538c7fcb0865eb7f2c',
  54         'info_dict': {
  55             'id': '28488308',
  56             'ext': 'mp3',
  57             'title': 'Tod eines Fußballers',
  58             'description': 'md5:f6e39f3461f0e1f54bfa48c8875c86ef',
  59             'duration': 3240,
  60         },
  61         'skip': 'HTTP Error 404: Not Found',
  62     }, {
  63         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
  64         'only_matching': True,
  65     }, {
  66         # audio
  67         'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
  68         'md5': '4e8f00631aac0395fee17368ac0e9867',
  69         'info_dict': {
  70             'id': '30796318',
  71             'ext': 'mp3',
  72             'title': 'Vor dem Fest',
  73             'description': 'md5:c0c1c8048514deaed2a73b3a60eecacb',
  74             'duration': 3287,
  75         },
  76         'skip': 'Video is no longer available',
  77     }]
  78
  79     def _extract_media_info(self, media_info_url, webpage, video_id):
  80         media_info = self._download_json(
  81             media_info_url, video_id, 'Downloading media JSON')
  82
  83         formats = self._extract_formats(media_info, video_id)
  84
  85         if not formats:
  86             if '"fsk"' in webpage:
  87                 raise ExtractorError(
  88                     'This video is only available after 20:00', expected=True)
  89             elif media_info.get('_geoblocked'):
  90                 raise ExtractorError('This video is not available due to geo restriction', expected=True)
  91
  92         self._sort_formats(formats)
  93
  94         duration = int_or_none(media_info.get('_duration'))
  95         thumbnail = media_info.get('_previewImage')
  96         is_live = media_info.get('_isLive') is True
  97
  98         subtitles = {}
  99         subtitle_url = media_info.get('_subtitleUrl')
 100         if subtitle_url:
 101             subtitles['de'] = [{
 102                 'ext': 'ttml',
 103                 'url': subtitle_url,
 104             }]
 105
 106         return {
 107             'id': video_id,
 108             'duration': duration,
 109             'thumbnail': thumbnail,
 110             'is_live': is_live,
 111             'formats': formats,
 112             'subtitles': subtitles,
 113         }
 114
 115     def _extract_formats(self, media_info, video_id):
 116         type_ = media_info.get('_type')
 117         media_array = media_info.get('_mediaArray', [])
 118         formats = []
 119         for num, media in enumerate(media_array):
 120             for stream in media.get('_mediaStreamArray', []):
 121                 stream_urls = stream.get('_stream')
 122                 if not stream_urls:
 123                     continue
 124                 if not isinstance(stream_urls, list):
 125                     stream_urls = [stream_urls]
 126                 quality = stream.get('_quality')
 127                 server = stream.get('_server')
 128                 for stream_url in stream_urls:
 129                     ext = determine_ext(stream_url)
 130                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
 131                         continue
 132                     if ext == 'f4m':
 133                         formats.extend(self._extract_f4m_formats(
 134                             update_url_query(stream_url, {
 135                                 'hdcore': '3.1.1',
 136                                 'plugin': 'aasp-3.1.1.69.124'
 137                             }),
 138                             video_id, f4m_id='hds', fatal=False))
 139                     elif ext == 'm3u8':
 140                         formats.extend(self._extract_m3u8_formats(
 141                             stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
 142                     else:
 143                         if server and server.startswith('rtmp'):
 144                             f = {
 145                                 'url': server,
 146                                 'play_path': stream_url,
 147                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
 148                             }
 149                         elif stream_url.startswith('http'):
 150                             f = {
 151                                 'url': stream_url,
 152                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 153                             }
 154                         else:
 155                             continue
 156                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
 157                         if m:
 158                             f.update({
 159                                 'width': int(m.group('width')),
 160                                 'height': int(m.group('height')),
 161                             })
 162                         if type_ == 'audio':
 163                             f['vcodec'] = 'none'
 164                         formats.append(f)
 165         return formats
 166
 167     def _real_extract(self, url):
 168         # determine video id from url
 169         m = re.match(self._VALID_URL, url)
 170
 171         document_id = None
 172
 173         numid = re.search(r'documentId=([0-9]+)', url)
 174         if numid:
 175             document_id = video_id = numid.group(1)
 176         else:
 177             video_id = m.group('video_id')
 178
 179         webpage = self._download_webpage(url, video_id)
 180
 181         ERRORS = (
 182             ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
 183             ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
 184              'Video %s is no longer available'),
 185         )
 186
 187         for pattern, message in ERRORS:
 188             if pattern in webpage:
 189                 raise ExtractorError(message % video_id, expected=True)
 190
 191         if re.search(r'[\?&]rss($|[=&])', url):
 192             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 193             if doc.tag == 'rss':
 194                 return GenericIE()._extract_rss(url, video_id, doc)
 195
 196         title = self._html_search_regex(
 197             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 198              r'<meta name="dcterms\.title" content="(.*?)"/>',
 199              r'<h4 class="headline">(.*?)</h4>'],
 200             webpage, 'title')
 201         description = self._html_search_meta(
 202             'dcterms.abstract', webpage, 'description', default=None)
 203         if description is None:
 204             description = self._html_search_meta(
 205                 'description', webpage, 'meta description')
 206
 207         # Thumbnail is sometimes not present.
 208         # It is in the mobile version, but that seems to use a different URL
 209         # structure altogether.
 210         thumbnail = self._og_search_thumbnail(webpage, default=None)
 211
 212         media_streams = re.findall(r'''(?x)
 213             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 214             "([^"]+)"''', webpage)
 215
 216         if media_streams:
 217             QUALITIES = qualities(['lo', 'hi', 'hq'])
 218             formats = []
 219             for furl in set(media_streams):
 220                 if furl.endswith('.f4m'):
 221                     fid = 'f4m'
 222                 else:
 223                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 224                     fid = fid_m.group(1) if fid_m else None
 225                 formats.append({
 226                     'quality': QUALITIES(fid),
 227                     'format_id': fid,
 228                     'url': furl,
 229                 })
 230             self._sort_formats(formats)
 231             info = {
 232                 'formats': formats,
 233             }
 234         else:  # request JSON file
 235             if not document_id:
 236                 video_id = self._search_regex(
 237                     r'/play/(?:config|media)/(\d+)', webpage, 'media id')
 238             info = self._extract_media_info(
 239                 'http://www.ardmediathek.de/play/media/%s' % video_id,
 240                 webpage, video_id)
 241
 242         info.update({
 243             'id': video_id,
 244             'title': self._live_title(title) if info.get('is_live') else title,
 245             'description': description,
 246             'thumbnail': thumbnail,
 247         })
 248
 249         return info
 250
 251
 252 class ARDIE(InfoExtractor):
 253     _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
 254     _TEST = {
 255         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 256         'md5': 'd216c3a86493f9322545e045ddc3eb35',
 257         'info_dict': {
 258             'display_id': 'die-story-im-ersten-mission-unter-falscher-flagge',
 259             'id': '100',
 260             'ext': 'mp4',
 261             'duration': 2600,
 262             'title': 'Die Story im Ersten: Mission unter falscher Flagge',
 263             'upload_date': '20140804',
 264             'thumbnail': r're:^https?://.*\.jpg$',
 265         },
 266         'skip': 'HTTP Error 404: Not Found',
 267     }
 268
 269     def _real_extract(self, url):
 270         mobj = re.match(self._VALID_URL, url)
 271         display_id = mobj.group('display_id')
 272
 273         player_url = mobj.group('mainurl') + '~playerXml.xml'
 274         doc = self._download_xml(player_url, display_id)
 275         video_node = doc.find('./video')
 276         upload_date = unified_strdate(xpath_text(
 277             video_node, './broadcastDate'))
 278         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 279
 280         formats = []
 281         for a in video_node.findall('.//asset'):
 282             f = {
 283                 'format_id': a.attrib['type'],
 284                 'width': int_or_none(a.find('./frameWidth').text),
 285                 'height': int_or_none(a.find('./frameHeight').text),
 286                 'vbr': int_or_none(a.find('./bitrateVideo').text),
 287                 'abr': int_or_none(a.find('./bitrateAudio').text),
 288                 'vcodec': a.find('./codecVideo').text,
 289                 'tbr': int_or_none(a.find('./totalBitrate').text),
 290             }
 291             if a.find('./serverPrefix').text:
 292                 f['url'] = a.find('./serverPrefix').text
 293                 f['playpath'] = a.find('./fileName').text
 294             else:
 295                 f['url'] = a.find('./fileName').text
 296             formats.append(f)
 297         self._sort_formats(formats)
 298
 299         return {
 300             'id': mobj.group('id'),
 301             'formats': formats,
 302             'display_id': display_id,
 303             'title': video_node.find('./title').text,
 304             'duration': parse_duration(video_node.find('./duration').text),
 305             'upload_date': upload_date,
 306             'thumbnail': thumbnail,
 307         }