_ Git - youtube-dl/blob - youtube_dl/extractor/dplay.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import re
   6 import time
   7
   8 from .common import InfoExtractor
   9 from ..compat import (
  10     compat_HTTPError,
  11     compat_str,
  12     compat_urlparse,
  13 )
  14 from ..utils import (
  15     determine_ext,
  16     ExtractorError,
  17     float_or_none,
  18     int_or_none,
  19     remove_end,
  20     try_get,
  21     unified_strdate,
  22     unified_timestamp,
  23     update_url_query,
  24     USER_AGENTS,
  25 )
  26
  27
  28 class DPlayIE(InfoExtractor):
  29     _VALID_URL = r'https?://(?P<domain>www\.(?P<host>dplay\.(?:dk|se|no)))/(?:videoer/)?(?P<id>[^/]+/[^/?#]+)'
  30
  31     _TESTS = [{
  32         # non geo restricted, via secure api, unsigned download hls URL
  33         'url': 'http://www.dplay.se/nugammalt-77-handelser-som-format-sverige/season-1-svensken-lar-sig-njuta-av-livet/',
  34         'info_dict': {
  35             'id': '3172',
  36             'display_id': 'nugammalt-77-handelser-som-format-sverige/season-1-svensken-lar-sig-njuta-av-livet',
  37             'ext': 'mp4',
  38             'title': 'Svensken lär sig njuta av livet',
  39             'description': 'md5:d3819c9bccffd0fe458ca42451dd50d8',
  40             'duration': 2650,
  41             'timestamp': 1365454320,
  42             'upload_date': '20130408',
  43             'creator': 'Kanal 5 (Home)',
  44             'series': 'Nugammalt - 77 händelser som format Sverige',
  45             'season_number': 1,
  46             'episode_number': 1,
  47             'age_limit': 0,
  48         },
  49     }, {
  50         # geo restricted, via secure api, unsigned download hls URL
  51         'url': 'http://www.dplay.dk/mig-og-min-mor/season-6-episode-12/',
  52         'info_dict': {
  53             'id': '70816',
  54             'display_id': 'mig-og-min-mor/season-6-episode-12',
  55             'ext': 'mp4',
  56             'title': 'Episode 12',
  57             'description': 'md5:9c86e51a93f8a4401fc9641ef9894c90',
  58             'duration': 2563,
  59             'timestamp': 1429696800,
  60             'upload_date': '20150422',
  61             'creator': 'Kanal 4 (Home)',
  62             'series': 'Mig og min mor',
  63             'season_number': 6,
  64             'episode_number': 12,
  65             'age_limit': 0,
  66         },
  67     }, {
  68         # geo restricted, via direct unsigned hls URL
  69         'url': 'http://www.dplay.no/pga-tour/season-1-hoydepunkter-18-21-februar/',
  70         'only_matching': True,
  71     }, {
  72         # disco-api
  73         'url': 'https://www.dplay.no/videoer/i-kongens-klr/sesong-1-episode-7',
  74         'info_dict': {
  75             'id': '40206',
  76             'display_id': 'i-kongens-klr/sesong-1-episode-7',
  77             'ext': 'mp4',
  78             'title': 'Episode 7',
  79             'description': 'md5:e3e1411b2b9aebeea36a6ec5d50c60cf',
  80             'duration': 2611.16,
  81             'timestamp': 1516726800,
  82             'upload_date': '20180123',
  83             'series': 'I kongens klær',
  84             'season_number': 1,
  85             'episode_number': 7,
  86         },
  87         'params': {
  88             'format': 'bestvideo',
  89             'skip_download': True,
  90         },
  91     }]
  92
  93     def _real_extract(self, url):
  94         mobj = re.match(self._VALID_URL, url)
  95         display_id = mobj.group('id')
  96         domain = mobj.group('domain')
  97
  98         webpage = self._download_webpage(url, display_id)
  99
 100         video_id = self._search_regex(
 101             r'data-video-id=["\'](\d+)', webpage, 'video id', default=None)
 102
 103         if not video_id:
 104             host = mobj.group('host')
 105             disco_base = 'https://disco-api.%s' % host
 106             self._download_json(
 107                 '%s/token' % disco_base, display_id, 'Downloading token',
 108                 query={
 109                     'realm': host.replace('.', ''),
 110                 })
 111             video = self._download_json(
 112                 '%s/content/videos/%s' % (disco_base, display_id), display_id,
 113                 headers={
 114                     'Referer': url,
 115                     'x-disco-client': 'WEB:UNKNOWN:dplay-client:0.0.1',
 116                 }, query={
 117                     'include': 'show'
 118                 })
 119             video_id = video['data']['id']
 120             info = video['data']['attributes']
 121             title = info['name']
 122             formats = []
 123             for format_id, format_dict in self._download_json(
 124                     '%s/playback/videoPlaybackInfo/%s' % (disco_base, video_id),
 125                     display_id)['data']['attributes']['streaming'].items():
 126                 if not isinstance(format_dict, dict):
 127                     continue
 128                 format_url = format_dict.get('url')
 129                 if not format_url:
 130                     continue
 131                 ext = determine_ext(format_url)
 132                 if format_id == 'dash' or ext == 'mpd':
 133                     formats.extend(self._extract_mpd_formats(
 134                         format_url, display_id, mpd_id='dash', fatal=False))
 135                 elif format_id == 'hls' or ext == 'm3u8':
 136                     formats.extend(self._extract_m3u8_formats(
 137                         format_url, display_id, 'mp4',
 138                         entry_protocol='m3u8_native', m3u8_id='hls',
 139                         fatal=False))
 140                 else:
 141                     formats.append({
 142                         'url': format_url,
 143                         'format_id': format_id,
 144                     })
 145             self._sort_formats(formats)
 146
 147             series = None
 148             try:
 149                 included = video.get('included')
 150                 if isinstance(included, list):
 151                     show = next(e for e in included if e.get('type') == 'show')
 152                     series = try_get(
 153                         show, lambda x: x['attributes']['name'], compat_str)
 154             except StopIteration:
 155                 pass
 156
 157             return {
 158                 'id': video_id,
 159                 'display_id': display_id,
 160                 'title': title,
 161                 'description': info.get('description'),
 162                 'duration': float_or_none(
 163                     info.get('videoDuration'), scale=1000),
 164                 'timestamp': unified_timestamp(info.get('publishStart')),
 165                 'series': series,
 166                 'season_number': int_or_none(info.get('seasonNumber')),
 167                 'episode_number': int_or_none(info.get('episodeNumber')),
 168                 'age_limit': int_or_none(info.get('minimum_age')),
 169                 'formats': formats,
 170             }
 171
 172         info = self._download_json(
 173             'http://%s/api/v2/ajax/videos?video_id=%s' % (domain, video_id),
 174             video_id)['data'][0]
 175
 176         title = info['title']
 177
 178         PROTOCOLS = ('hls', 'hds')
 179         formats = []
 180
 181         def extract_formats(protocol, manifest_url):
 182             if protocol == 'hls':
 183                 m3u8_formats = self._extract_m3u8_formats(
 184                     manifest_url, video_id, ext='mp4',
 185                     entry_protocol='m3u8_native', m3u8_id=protocol, fatal=False)
 186                 # Sometimes final URLs inside m3u8 are unsigned, let's fix this
 187                 # ourselves. Also fragments' URLs are only served signed for
 188                 # Safari user agent.
 189                 query = compat_urlparse.parse_qs(compat_urlparse.urlparse(manifest_url).query)
 190                 for m3u8_format in m3u8_formats:
 191                     m3u8_format.update({
 192                         'url': update_url_query(m3u8_format['url'], query),
 193                         'http_headers': {
 194                             'User-Agent': USER_AGENTS['Safari'],
 195                         },
 196                     })
 197                 formats.extend(m3u8_formats)
 198             elif protocol == 'hds':
 199                 formats.extend(self._extract_f4m_formats(
 200                     manifest_url + '&hdcore=3.8.0&plugin=flowplayer-3.8.0.0',
 201                     video_id, f4m_id=protocol, fatal=False))
 202
 203         domain_tld = domain.split('.')[-1]
 204         if domain_tld in ('se', 'dk', 'no'):
 205             for protocol in PROTOCOLS:
 206                 # Providing dsc-geo allows to bypass geo restriction in some cases
 207                 self._set_cookie(
 208                     'secure.dplay.%s' % domain_tld, 'dsc-geo',
 209                     json.dumps({
 210                         'countryCode': domain_tld.upper(),
 211                         'expiry': (time.time() + 20 * 60) * 1000,
 212                     }))
 213                 stream = self._download_json(
 214                     'https://secure.dplay.%s/secure/api/v2/user/authorization/stream/%s?stream_type=%s'
 215                     % (domain_tld, video_id, protocol), video_id,
 216                     'Downloading %s stream JSON' % protocol, fatal=False)
 217                 if stream and stream.get(protocol):
 218                     extract_formats(protocol, stream[protocol])
 219
 220         # The last resort is to try direct unsigned hls/hds URLs from info dictionary.
 221         # Sometimes this does work even when secure API with dsc-geo has failed (e.g.
 222         # http://www.dplay.no/pga-tour/season-1-hoydepunkter-18-21-februar/).
 223         if not formats:
 224             for protocol in PROTOCOLS:
 225                 if info.get(protocol):
 226                     extract_formats(protocol, info[protocol])
 227
 228         self._sort_formats(formats)
 229
 230         subtitles = {}
 231         for lang in ('se', 'sv', 'da', 'nl', 'no'):
 232             for format_id in ('web_vtt', 'vtt', 'srt'):
 233                 subtitle_url = info.get('subtitles_%s_%s' % (lang, format_id))
 234                 if subtitle_url:
 235                     subtitles.setdefault(lang, []).append({'url': subtitle_url})
 236
 237         return {
 238             'id': video_id,
 239             'display_id': display_id,
 240             'title': title,
 241             'description': info.get('video_metadata_longDescription'),
 242             'duration': int_or_none(info.get('video_metadata_length'), scale=1000),
 243             'timestamp': int_or_none(info.get('video_publish_date')),
 244             'creator': info.get('video_metadata_homeChannel'),
 245             'series': info.get('video_metadata_show'),
 246             'season_number': int_or_none(info.get('season')),
 247             'episode_number': int_or_none(info.get('episode')),
 248             'age_limit': int_or_none(info.get('minimum_age')),
 249             'formats': formats,
 250             'subtitles': subtitles,
 251         }
 252
 253
 254 class DPlayItIE(InfoExtractor):
 255     _VALID_URL = r'https?://it\.dplay\.com/[^/]+/[^/]+/(?P<id>[^/?#]+)'
 256     _GEO_COUNTRIES = ['IT']
 257     _TEST = {
 258         'url': 'http://it.dplay.com/nove/biografie-imbarazzanti/luigi-di-maio-la-psicosi-di-stanislawskij/',
 259         'md5': '2b808ffb00fc47b884a172ca5d13053c',
 260         'info_dict': {
 261             'id': '6918',
 262             'display_id': 'luigi-di-maio-la-psicosi-di-stanislawskij',
 263             'ext': 'mp4',
 264             'title': 'Biografie imbarazzanti: Luigi Di Maio: la psicosi di Stanislawskij',
 265             'description': 'md5:3c7a4303aef85868f867a26f5cc14813',
 266             'thumbnail': r're:^https?://.*\.jpe?g',
 267             'upload_date': '20160524',
 268             'series': 'Biografie imbarazzanti',
 269             'season_number': 1,
 270             'episode': 'Luigi Di Maio: la psicosi di Stanislawskij',
 271             'episode_number': 1,
 272         },
 273     }
 274
 275     def _real_extract(self, url):
 276         display_id = self._match_id(url)
 277
 278         webpage = self._download_webpage(url, display_id)
 279
 280         title = remove_end(self._og_search_title(webpage), ' | Dplay')
 281
 282         video_id = None
 283
 284         info = self._search_regex(
 285             r'playback_json\s*:\s*JSON\.parse\s*\(\s*("(?:\\.|[^"\\])+?")',
 286             webpage, 'playback JSON', default=None)
 287         if info:
 288             for _ in range(2):
 289                 info = self._parse_json(info, display_id, fatal=False)
 290                 if not info:
 291                     break
 292             else:
 293                 video_id = try_get(info, lambda x: x['data']['id'])
 294
 295         if not info:
 296             info_url = self._search_regex(
 297                 r'url\s*[:=]\s*["\']((?:https?:)?//[^/]+/playback/videoPlaybackInfo/\d+)',
 298                 webpage, 'info url')
 299
 300             video_id = info_url.rpartition('/')[-1]
 301
 302             try:
 303                 info = self._download_json(
 304                     info_url, display_id, headers={
 305                         'Authorization': 'Bearer %s' % self._get_cookies(url).get(
 306                             'dplayit_token').value,
 307                         'Referer': url,
 308                     })
 309             except ExtractorError as e:
 310                 if isinstance(e.cause, compat_HTTPError) and e.cause.code in (400, 403):
 311                     info = self._parse_json(e.cause.read().decode('utf-8'), display_id)
 312                     error = info['errors'][0]
 313                     if error.get('code') == 'access.denied.geoblocked':
 314                         self.raise_geo_restricted(
 315                             msg=error.get('detail'), countries=self._GEO_COUNTRIES)
 316                     raise ExtractorError(info['errors'][0]['detail'], expected=True)
 317                 raise
 318
 319         hls_url = info['data']['attributes']['streaming']['hls']['url']
 320
 321         formats = self._extract_m3u8_formats(
 322             hls_url, display_id, ext='mp4', entry_protocol='m3u8_native',
 323             m3u8_id='hls')
 324
 325         series = self._html_search_regex(
 326             r'(?s)<h1[^>]+class=["\'].*?\bshow_title\b.*?["\'][^>]*>(.+?)</h1>',
 327             webpage, 'series', fatal=False)
 328         episode = self._search_regex(
 329             r'<p[^>]+class=["\'].*?\bdesc_ep\b.*?["\'][^>]*>\s*<br/>\s*<b>([^<]+)',
 330             webpage, 'episode', fatal=False)
 331
 332         mobj = re.search(
 333             r'(?s)<span[^>]+class=["\']dates["\'][^>]*>.+?\bS\.(?P<season_number>\d+)\s+E\.(?P<episode_number>\d+)\s*-\s*(?P<upload_date>\d{2}/\d{2}/\d{4})',
 334             webpage)
 335         if mobj:
 336             season_number = int(mobj.group('season_number'))
 337             episode_number = int(mobj.group('episode_number'))
 338             upload_date = unified_strdate(mobj.group('upload_date'))
 339         else:
 340             season_number = episode_number = upload_date = None
 341
 342         return {
 343             'id': compat_str(video_id or display_id),
 344             'display_id': display_id,
 345             'title': title,
 346             'description': self._og_search_description(webpage),
 347             'thumbnail': self._og_search_thumbnail(webpage),
 348             'series': series,
 349             'season_number': season_number,
 350             'episode': episode,
 351             'episode_number': episode_number,
 352             'upload_date': upload_date,
 353             'formats': formats,
 354         }