Raphaël G. Git Repositories - youtubedl/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\.(?P<country>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         # geo restricted, bypassable via X-Forwarded-For
  93         'url': 'https://www.dplay.dk/videoer/singleliv/season-5-episode-3',
  94         'only_matching': True,
  95     }]
  96
  97     def _real_extract(self, url):
  98         mobj = re.match(self._VALID_URL, url)
  99         display_id = mobj.group('id')
 100         domain = mobj.group('domain')
 101
 102         self._initialize_geo_bypass([mobj.group('country').upper()])
 103
 104         webpage = self._download_webpage(url, display_id)
 105
 106         video_id = self._search_regex(
 107             r'data-video-id=["\'](\d+)', webpage, 'video id', default=None)
 108
 109         if not video_id:
 110             host = mobj.group('host')
 111             disco_base = 'https://disco-api.%s' % host
 112             self._download_json(
 113                 '%s/token' % disco_base, display_id, 'Downloading token',
 114                 query={
 115                     'realm': host.replace('.', ''),
 116                 })
 117             video = self._download_json(
 118                 '%s/content/videos/%s' % (disco_base, display_id), display_id,
 119                 headers={
 120                     'Referer': url,
 121                     'x-disco-client': 'WEB:UNKNOWN:dplay-client:0.0.1',
 122                 }, query={
 123                     'include': 'show'
 124                 })
 125             video_id = video['data']['id']
 126             info = video['data']['attributes']
 127             title = info['name']
 128             formats = []
 129             for format_id, format_dict in self._download_json(
 130                     '%s/playback/videoPlaybackInfo/%s' % (disco_base, video_id),
 131                     display_id)['data']['attributes']['streaming'].items():
 132                 if not isinstance(format_dict, dict):
 133                     continue
 134                 format_url = format_dict.get('url')
 135                 if not format_url:
 136                     continue
 137                 ext = determine_ext(format_url)
 138                 if format_id == 'dash' or ext == 'mpd':
 139                     formats.extend(self._extract_mpd_formats(
 140                         format_url, display_id, mpd_id='dash', fatal=False))
 141                 elif format_id == 'hls' or ext == 'm3u8':
 142                     formats.extend(self._extract_m3u8_formats(
 143                         format_url, display_id, 'mp4',
 144                         entry_protocol='m3u8_native', m3u8_id='hls',
 145                         fatal=False))
 146                 else:
 147                     formats.append({
 148                         'url': format_url,
 149                         'format_id': format_id,
 150                     })
 151             self._sort_formats(formats)
 152
 153             series = None
 154             try:
 155                 included = video.get('included')
 156                 if isinstance(included, list):
 157                     show = next(e for e in included if e.get('type') == 'show')
 158                     series = try_get(
 159                         show, lambda x: x['attributes']['name'], compat_str)
 160             except StopIteration:
 161                 pass
 162
 163             return {
 164                 'id': video_id,
 165                 'display_id': display_id,
 166                 'title': title,
 167                 'description': info.get('description'),
 168                 'duration': float_or_none(
 169                     info.get('videoDuration'), scale=1000),
 170                 'timestamp': unified_timestamp(info.get('publishStart')),
 171                 'series': series,
 172                 'season_number': int_or_none(info.get('seasonNumber')),
 173                 'episode_number': int_or_none(info.get('episodeNumber')),
 174                 'age_limit': int_or_none(info.get('minimum_age')),
 175                 'formats': formats,
 176             }
 177
 178         info = self._download_json(
 179             'http://%s/api/v2/ajax/videos?video_id=%s' % (domain, video_id),
 180             video_id)['data'][0]
 181
 182         title = info['title']
 183
 184         PROTOCOLS = ('hls', 'hds')
 185         formats = []
 186
 187         def extract_formats(protocol, manifest_url):
 188             if protocol == 'hls':
 189                 m3u8_formats = self._extract_m3u8_formats(
 190                     manifest_url, video_id, ext='mp4',
 191                     entry_protocol='m3u8_native', m3u8_id=protocol, fatal=False)
 192                 # Sometimes final URLs inside m3u8 are unsigned, let's fix this
 193                 # ourselves. Also fragments' URLs are only served signed for
 194                 # Safari user agent.
 195                 query = compat_urlparse.parse_qs(compat_urlparse.urlparse(manifest_url).query)
 196                 for m3u8_format in m3u8_formats:
 197                     m3u8_format.update({
 198                         'url': update_url_query(m3u8_format['url'], query),
 199                         'http_headers': {
 200                             'User-Agent': USER_AGENTS['Safari'],
 201                         },
 202                     })
 203                 formats.extend(m3u8_formats)
 204             elif protocol == 'hds':
 205                 formats.extend(self._extract_f4m_formats(
 206                     manifest_url + '&hdcore=3.8.0&plugin=flowplayer-3.8.0.0',
 207                     video_id, f4m_id=protocol, fatal=False))
 208
 209         domain_tld = domain.split('.')[-1]
 210         if domain_tld in ('se', 'dk', 'no'):
 211             for protocol in PROTOCOLS:
 212                 # Providing dsc-geo allows to bypass geo restriction in some cases
 213                 self._set_cookie(
 214                     'secure.dplay.%s' % domain_tld, 'dsc-geo',
 215                     json.dumps({
 216                         'countryCode': domain_tld.upper(),
 217                         'expiry': (time.time() + 20 * 60) * 1000,
 218                     }))
 219                 stream = self._download_json(
 220                     'https://secure.dplay.%s/secure/api/v2/user/authorization/stream/%s?stream_type=%s'
 221                     % (domain_tld, video_id, protocol), video_id,
 222                     'Downloading %s stream JSON' % protocol, fatal=False)
 223                 if stream and stream.get(protocol):
 224                     extract_formats(protocol, stream[protocol])
 225
 226         # The last resort is to try direct unsigned hls/hds URLs from info dictionary.
 227         # Sometimes this does work even when secure API with dsc-geo has failed (e.g.
 228         # http://www.dplay.no/pga-tour/season-1-hoydepunkter-18-21-februar/).
 229         if not formats:
 230             for protocol in PROTOCOLS:
 231                 if info.get(protocol):
 232                     extract_formats(protocol, info[protocol])
 233
 234         self._sort_formats(formats)
 235
 236         subtitles = {}
 237         for lang in ('se', 'sv', 'da', 'nl', 'no'):
 238             for format_id in ('web_vtt', 'vtt', 'srt'):
 239                 subtitle_url = info.get('subtitles_%s_%s' % (lang, format_id))
 240                 if subtitle_url:
 241                     subtitles.setdefault(lang, []).append({'url': subtitle_url})
 242
 243         return {
 244             'id': video_id,
 245             'display_id': display_id,
 246             'title': title,
 247             'description': info.get('video_metadata_longDescription'),
 248             'duration': int_or_none(info.get('video_metadata_length'), scale=1000),
 249             'timestamp': int_or_none(info.get('video_publish_date')),
 250             'creator': info.get('video_metadata_homeChannel'),
 251             'series': info.get('video_metadata_show'),
 252             'season_number': int_or_none(info.get('season')),
 253             'episode_number': int_or_none(info.get('episode')),
 254             'age_limit': int_or_none(info.get('minimum_age')),
 255             'formats': formats,
 256             'subtitles': subtitles,
 257         }
 258
 259
 260 class DPlayItIE(InfoExtractor):
 261     _VALID_URL = r'https?://it\.dplay\.com/[^/]+/[^/]+/(?P<id>[^/?#]+)'
 262     _GEO_COUNTRIES = ['IT']
 263     _TEST = {
 264         'url': 'http://it.dplay.com/nove/biografie-imbarazzanti/luigi-di-maio-la-psicosi-di-stanislawskij/',
 265         'md5': '2b808ffb00fc47b884a172ca5d13053c',
 266         'info_dict': {
 267             'id': '6918',
 268             'display_id': 'luigi-di-maio-la-psicosi-di-stanislawskij',
 269             'ext': 'mp4',
 270             'title': 'Biografie imbarazzanti: Luigi Di Maio: la psicosi di Stanislawskij',
 271             'description': 'md5:3c7a4303aef85868f867a26f5cc14813',
 272             'thumbnail': r're:^https?://.*\.jpe?g',
 273             'upload_date': '20160524',
 274             'series': 'Biografie imbarazzanti',
 275             'season_number': 1,
 276             'episode': 'Luigi Di Maio: la psicosi di Stanislawskij',
 277             'episode_number': 1,
 278         },
 279     }
 280
 281     def _real_extract(self, url):
 282         display_id = self._match_id(url)
 283
 284         webpage = self._download_webpage(url, display_id)
 285
 286         title = remove_end(self._og_search_title(webpage), ' | Dplay')
 287
 288         video_id = None
 289
 290         info = self._search_regex(
 291             r'playback_json\s*:\s*JSON\.parse\s*\(\s*("(?:\\.|[^"\\])+?")',
 292             webpage, 'playback JSON', default=None)
 293         if info:
 294             for _ in range(2):
 295                 info = self._parse_json(info, display_id, fatal=False)
 296                 if not info:
 297                     break
 298             else:
 299                 video_id = try_get(info, lambda x: x['data']['id'])
 300
 301         if not info:
 302             info_url = self._search_regex(
 303                 r'url\s*[:=]\s*["\']((?:https?:)?//[^/]+/playback/videoPlaybackInfo/\d+)',
 304                 webpage, 'info url')
 305
 306             video_id = info_url.rpartition('/')[-1]
 307
 308             try:
 309                 info = self._download_json(
 310                     info_url, display_id, headers={
 311                         'Authorization': 'Bearer %s' % self._get_cookies(url).get(
 312                             'dplayit_token').value,
 313                         'Referer': url,
 314                     })
 315             except ExtractorError as e:
 316                 if isinstance(e.cause, compat_HTTPError) and e.cause.code in (400, 403):
 317                     info = self._parse_json(e.cause.read().decode('utf-8'), display_id)
 318                     error = info['errors'][0]
 319                     if error.get('code') == 'access.denied.geoblocked':
 320                         self.raise_geo_restricted(
 321                             msg=error.get('detail'), countries=self._GEO_COUNTRIES)
 322                     raise ExtractorError(info['errors'][0]['detail'], expected=True)
 323                 raise
 324
 325         hls_url = info['data']['attributes']['streaming']['hls']['url']
 326
 327         formats = self._extract_m3u8_formats(
 328             hls_url, display_id, ext='mp4', entry_protocol='m3u8_native',
 329             m3u8_id='hls')
 330
 331         series = self._html_search_regex(
 332             r'(?s)<h1[^>]+class=["\'].*?\bshow_title\b.*?["\'][^>]*>(.+?)</h1>',
 333             webpage, 'series', fatal=False)
 334         episode = self._search_regex(
 335             r'<p[^>]+class=["\'].*?\bdesc_ep\b.*?["\'][^>]*>\s*<br/>\s*<b>([^<]+)',
 336             webpage, 'episode', fatal=False)
 337
 338         mobj = re.search(
 339             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})',
 340             webpage)
 341         if mobj:
 342             season_number = int(mobj.group('season_number'))
 343             episode_number = int(mobj.group('episode_number'))
 344             upload_date = unified_strdate(mobj.group('upload_date'))
 345         else:
 346             season_number = episode_number = upload_date = None
 347
 348         return {
 349             'id': compat_str(video_id or display_id),
 350             'display_id': display_id,
 351             'title': title,
 352             'description': self._og_search_description(webpage),
 353             'thumbnail': self._og_search_thumbnail(webpage),
 354             'series': series,
 355             'season_number': season_number,
 356             'episode': episode,
 357             'episode_number': episode_number,
 358             'upload_date': upload_date,
 359             'formats': formats,
 360         }