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 )
  17 from ..compat import compat_etree_fromstring
  18
  19
  20 class ARDMediathekIE(InfoExtractor):
  21     IE_NAME = 'ARD:mediathek'
  22     _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
  23
  24     _TESTS = [{
  25         'url': 'http://www.ardmediathek.de/tv/Dokumentation-und-Reportage/Ich-liebe-das-Leben-trotzdem/rbb-Fernsehen/Video?documentId=29582122&bcastId=3822114',
  26         'info_dict': {
  27             'id': '29582122',
  28             'ext': 'mp4',
  29             'title': 'Ich liebe das Leben trotzdem',
  30             'description': 'md5:45e4c225c72b27993314b31a84a5261c',
  31             'duration': 4557,
  32         },
  33         'params': {
  34             # m3u8 download
  35             'skip_download': True,
  36         },
  37     }, {
  38         'url': 'http://www.ardmediathek.de/tv/Tatort/Tatort-Scheinwelten-H%C3%B6rfassung-Video/Das-Erste/Video?documentId=29522730&bcastId=602916',
  39         'md5': 'f4d98b10759ac06c0072bbcd1f0b9e3e',
  40         'info_dict': {
  41             'id': '29522730',
  42             'ext': 'mp4',
  43             'title': 'Tatort: Scheinwelten - Hörfassung (Video tgl. ab 20 Uhr)',
  44             'description': 'md5:196392e79876d0ac94c94e8cdb2875f1',
  45             'duration': 5252,
  46         },
  47     }, {
  48         # audio
  49         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
  50         'md5': '219d94d8980b4f538c7fcb0865eb7f2c',
  51         'info_dict': {
  52             'id': '28488308',
  53             'ext': 'mp3',
  54             'title': 'Tod eines Fußballers',
  55             'description': 'md5:f6e39f3461f0e1f54bfa48c8875c86ef',
  56             'duration': 3240,
  57         },
  58     }, {
  59         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
  60         'only_matching': True,
  61     }]
  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
  81         subtitles = {}
  82         subtitle_url = media_info.get('_subtitleUrl')
  83         if subtitle_url:
  84             subtitles['de'] = [{
  85                 'ext': 'ttml',
  86                 'url': subtitle_url,
  87             }]
  88
  89         return {
  90             'id': video_id,
  91             'duration': duration,
  92             'thumbnail': thumbnail,
  93             'formats': formats,
  94             'subtitles': subtitles,
  95         }
  96
  97     def _extract_formats(self, media_info, video_id):
  98         type_ = media_info.get('_type')
  99         media_array = media_info.get('_mediaArray', [])
 100         formats = []
 101         for num, media in enumerate(media_array):
 102             for stream in media.get('_mediaStreamArray', []):
 103                 stream_urls = stream.get('_stream')
 104                 if not stream_urls:
 105                     continue
 106                 if not isinstance(stream_urls, list):
 107                     stream_urls = [stream_urls]
 108                 quality = stream.get('_quality')
 109                 server = stream.get('_server')
 110                 for stream_url in stream_urls:
 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                             stream_url + '?hdcore=3.1.1&plugin=aasp-3.1.1.69.124',
 117                             video_id, preference=-1, f4m_id='hds', fatal=False))
 118                     elif ext == 'm3u8':
 119                         formats.extend(self._extract_m3u8_formats(
 120                             stream_url, video_id, 'mp4', preference=1, 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                         elif stream_url.startswith('http'):
 129                             f = {
 130                                 'url': stream_url,
 131                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 132                             }
 133                         else:
 134                             continue
 135                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
 136                         if m:
 137                             f.update({
 138                                 'width': int(m.group('width')),
 139                                 'height': int(m.group('height')),
 140                             })
 141                         if type_ == 'audio':
 142                             f['vcodec'] = 'none'
 143                         formats.append(f)
 144         return formats
 145
 146     def _real_extract(self, url):
 147         # determine video id from url
 148         m = re.match(self._VALID_URL, url)
 149
 150         numid = re.search(r'documentId=([0-9]+)', url)
 151         if numid:
 152             video_id = numid.group(1)
 153         else:
 154             video_id = m.group('video_id')
 155
 156         webpage = self._download_webpage(url, video_id)
 157
 158         if '>Der gewünschte Beitrag ist nicht mehr verfügbar.<' in webpage:
 159             raise ExtractorError('Video %s is no longer available' % video_id, expected=True)
 160
 161         if 'Diese Sendung ist für Jugendliche unter 12 Jahren nicht geeignet. Der Clip ist deshalb nur von 20 bis 6 Uhr verfügbar.' in webpage:
 162             raise ExtractorError('This program is only suitable for those aged 12 and older. Video %s is therefore only available between 20 pm and 6 am.' % video_id, expected=True)
 163
 164         if re.search(r'[\?&]rss($|[=&])', url):
 165             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 166             if doc.tag == 'rss':
 167                 return GenericIE()._extract_rss(url, video_id, doc)
 168
 169         title = self._html_search_regex(
 170             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 171              r'<meta name="dcterms.title" content="(.*?)"/>',
 172              r'<h4 class="headline">(.*?)</h4>'],
 173             webpage, 'title')
 174         description = self._html_search_meta(
 175             'dcterms.abstract', webpage, 'description', default=None)
 176         if description is None:
 177             description = self._html_search_meta(
 178                 'description', webpage, 'meta description')
 179
 180         # Thumbnail is sometimes not present.
 181         # It is in the mobile version, but that seems to use a different URL
 182         # structure altogether.
 183         thumbnail = self._og_search_thumbnail(webpage, default=None)
 184
 185         media_streams = re.findall(r'''(?x)
 186             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 187             "([^"]+)"''', webpage)
 188
 189         if media_streams:
 190             QUALITIES = qualities(['lo', 'hi', 'hq'])
 191             formats = []
 192             for furl in set(media_streams):
 193                 if furl.endswith('.f4m'):
 194                     fid = 'f4m'
 195                 else:
 196                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 197                     fid = fid_m.group(1) if fid_m else None
 198                 formats.append({
 199                     'quality': QUALITIES(fid),
 200                     'format_id': fid,
 201                     'url': furl,
 202                 })
 203             self._sort_formats(formats)
 204             info = {
 205                 'formats': formats,
 206             }
 207         else:  # request JSON file
 208             info = self._extract_media_info(
 209                 'http://www.ardmediathek.de/play/media/%s' % video_id, webpage, video_id)
 210
 211         info.update({
 212             'id': video_id,
 213             'title': title,
 214             'description': description,
 215             'thumbnail': thumbnail,
 216         })
 217
 218         return info
 219
 220
 221 class ARDIE(InfoExtractor):
 222     _VALID_URL = '(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
 223     _TEST = {
 224         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 225         'md5': 'd216c3a86493f9322545e045ddc3eb35',
 226         'info_dict': {
 227             'display_id': 'die-story-im-ersten-mission-unter-falscher-flagge',
 228             'id': '100',
 229             'ext': 'mp4',
 230             'duration': 2600,
 231             'title': 'Die Story im Ersten: Mission unter falscher Flagge',
 232             'upload_date': '20140804',
 233             'thumbnail': 're:^https?://.*\.jpg$',
 234         }
 235     }
 236
 237     def _real_extract(self, url):
 238         mobj = re.match(self._VALID_URL, url)
 239         display_id = mobj.group('display_id')
 240
 241         player_url = mobj.group('mainurl') + '~playerXml.xml'
 242         doc = self._download_xml(player_url, display_id)
 243         video_node = doc.find('./video')
 244         upload_date = unified_strdate(xpath_text(
 245             video_node, './broadcastDate'))
 246         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 247
 248         formats = []
 249         for a in video_node.findall('.//asset'):
 250             f = {
 251                 'format_id': a.attrib['type'],
 252                 'width': int_or_none(a.find('./frameWidth').text),
 253                 'height': int_or_none(a.find('./frameHeight').text),
 254                 'vbr': int_or_none(a.find('./bitrateVideo').text),
 255                 'abr': int_or_none(a.find('./bitrateAudio').text),
 256                 'vcodec': a.find('./codecVideo').text,
 257                 'tbr': int_or_none(a.find('./totalBitrate').text),
 258             }
 259             if a.find('./serverPrefix').text:
 260                 f['url'] = a.find('./serverPrefix').text
 261                 f['playpath'] = a.find('./fileName').text
 262             else:
 263                 f['url'] = a.find('./fileName').text
 264             formats.append(f)
 265         self._sort_formats(formats)
 266
 267         return {
 268             'id': mobj.group('id'),
 269             'formats': formats,
 270             'display_id': display_id,
 271             'title': video_node.find('./title').text,
 272             'duration': parse_duration(video_node.find('./duration').text),
 273             'upload_date': upload_date,
 274             'thumbnail': thumbnail,
 275         }