Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/canvas.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5
   6 from .common import InfoExtractor
   7 from .gigya import GigyaBaseIE
   8 from ..compat import compat_HTTPError
   9 from ..utils import (
  10     ExtractorError,
  11     strip_or_none,
  12     float_or_none,
  13     int_or_none,
  14     merge_dicts,
  15     parse_iso8601,
  16     str_or_none,
  17     url_or_none,
  18 )
  19
  20
  21 class CanvasIE(InfoExtractor):
  22     _VALID_URL = r'https?://mediazone\.vrt\.be/api/v1/(?P<site_id>canvas|een|ketnet|vrt(?:video|nieuws)|sporza)/assets/(?P<id>[^/?#&]+)'
  23     _TESTS = [{
  24         'url': 'https://mediazone.vrt.be/api/v1/ketnet/assets/md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
  25         'md5': '68993eda72ef62386a15ea2cf3c93107',
  26         'info_dict': {
  27             'id': 'md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
  28             'display_id': 'md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
  29             'ext': 'mp4',
  30             'title': 'Nachtwacht: De Greystook',
  31             'description': 'Nachtwacht: De Greystook',
  32             'thumbnail': r're:^https?://.*\.jpg$',
  33             'duration': 1468.04,
  34         },
  35         'expected_warnings': ['is not a supported codec', 'Unknown MIME type'],
  36     }, {
  37         'url': 'https://mediazone.vrt.be/api/v1/canvas/assets/mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
  38         'only_matching': True,
  39     }]
  40     _HLS_ENTRY_PROTOCOLS_MAP = {
  41         'HLS': 'm3u8_native',
  42         'HLS_AES': 'm3u8',
  43     }
  44     _REST_API_BASE = 'https://media-services-public.vrt.be/vualto-video-aggregator-web/rest/external/v1'
  45
  46     def _real_extract(self, url):
  47         mobj = re.match(self._VALID_URL, url)
  48         site_id, video_id = mobj.group('site_id'), mobj.group('id')
  49
  50         # Old API endpoint, serves more formats but may fail for some videos
  51         data = self._download_json(
  52             'https://mediazone.vrt.be/api/v1/%s/assets/%s'
  53             % (site_id, video_id), video_id, 'Downloading asset JSON',
  54             'Unable to download asset JSON', fatal=False)
  55
  56         # New API endpoint
  57         if not data:
  58             token = self._download_json(
  59                 '%s/tokens' % self._REST_API_BASE, video_id,
  60                 'Downloading token', data=b'',
  61                 headers={'Content-Type': 'application/json'})['vrtPlayerToken']
  62             data = self._download_json(
  63                 '%s/videos/%s' % (self._REST_API_BASE, video_id),
  64                 video_id, 'Downloading video JSON', fatal=False, query={
  65                     'vrtPlayerToken': token,
  66                     'client': '%s@PROD' % site_id,
  67                 }, expected_status=400)
  68             message = data.get('message')
  69             if message and not data.get('title'):
  70                 if data.get('code') == 'AUTHENTICATION_REQUIRED':
  71                     self.raise_login_required(message)
  72                 raise ExtractorError(message, expected=True)
  73
  74         title = data['title']
  75         description = data.get('description')
  76
  77         formats = []
  78         for target in data['targetUrls']:
  79             format_url, format_type = url_or_none(target.get('url')), str_or_none(target.get('type'))
  80             if not format_url or not format_type:
  81                 continue
  82             format_type = format_type.upper()
  83             if format_type in self._HLS_ENTRY_PROTOCOLS_MAP:
  84                 formats.extend(self._extract_m3u8_formats(
  85                     format_url, video_id, 'mp4', self._HLS_ENTRY_PROTOCOLS_MAP[format_type],
  86                     m3u8_id=format_type, fatal=False))
  87             elif format_type == 'HDS':
  88                 formats.extend(self._extract_f4m_formats(
  89                     format_url, video_id, f4m_id=format_type, fatal=False))
  90             elif format_type == 'MPEG_DASH':
  91                 formats.extend(self._extract_mpd_formats(
  92                     format_url, video_id, mpd_id=format_type, fatal=False))
  93             elif format_type == 'HSS':
  94                 formats.extend(self._extract_ism_formats(
  95                     format_url, video_id, ism_id='mss', fatal=False))
  96             else:
  97                 formats.append({
  98                     'format_id': format_type,
  99                     'url': format_url,
 100                 })
 101         self._sort_formats(formats)
 102
 103         subtitles = {}
 104         subtitle_urls = data.get('subtitleUrls')
 105         if isinstance(subtitle_urls, list):
 106             for subtitle in subtitle_urls:
 107                 subtitle_url = subtitle.get('url')
 108                 if subtitle_url and subtitle.get('type') == 'CLOSED':
 109                     subtitles.setdefault('nl', []).append({'url': subtitle_url})
 110
 111         return {
 112             'id': video_id,
 113             'display_id': video_id,
 114             'title': title,
 115             'description': description,
 116             'formats': formats,
 117             'duration': float_or_none(data.get('duration'), 1000),
 118             'thumbnail': data.get('posterImageUrl'),
 119             'subtitles': subtitles,
 120         }
 121
 122
 123 class CanvasEenIE(InfoExtractor):
 124     IE_DESC = 'canvas.be and een.be'
 125     _VALID_URL = r'https?://(?:www\.)?(?P<site_id>canvas|een)\.be/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 126     _TESTS = [{
 127         'url': 'http://www.canvas.be/video/de-afspraak/najaar-2015/de-afspraak-veilt-voor-de-warmste-week',
 128         'md5': 'ed66976748d12350b118455979cca293',
 129         'info_dict': {
 130             'id': 'mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
 131             'display_id': 'de-afspraak-veilt-voor-de-warmste-week',
 132             'ext': 'flv',
 133             'title': 'De afspraak veilt voor de Warmste Week',
 134             'description': 'md5:24cb860c320dc2be7358e0e5aa317ba6',
 135             'thumbnail': r're:^https?://.*\.jpg$',
 136             'duration': 49.02,
 137         },
 138         'expected_warnings': ['is not a supported codec'],
 139     }, {
 140         # with subtitles
 141         'url': 'http://www.canvas.be/video/panorama/2016/pieter-0167',
 142         'info_dict': {
 143             'id': 'mz-ast-5240ff21-2d30-4101-bba6-92b5ec67c625',
 144             'display_id': 'pieter-0167',
 145             'ext': 'mp4',
 146             'title': 'Pieter 0167',
 147             'description': 'md5:943cd30f48a5d29ba02c3a104dc4ec4e',
 148             'thumbnail': r're:^https?://.*\.jpg$',
 149             'duration': 2553.08,
 150             'subtitles': {
 151                 'nl': [{
 152                     'ext': 'vtt',
 153                 }],
 154             },
 155         },
 156         'params': {
 157             'skip_download': True,
 158         },
 159         'skip': 'Pagina niet gevonden',
 160     }, {
 161         'url': 'https://www.een.be/thuis/emma-pakt-thilly-aan',
 162         'info_dict': {
 163             'id': 'md-ast-3a24ced2-64d7-44fb-b4ed-ed1aafbf90b8',
 164             'display_id': 'emma-pakt-thilly-aan',
 165             'ext': 'mp4',
 166             'title': 'Emma pakt Thilly aan',
 167             'description': 'md5:c5c9b572388a99b2690030afa3f3bad7',
 168             'thumbnail': r're:^https?://.*\.jpg$',
 169             'duration': 118.24,
 170         },
 171         'params': {
 172             'skip_download': True,
 173         },
 174         'expected_warnings': ['is not a supported codec'],
 175     }, {
 176         'url': 'https://www.canvas.be/check-point/najaar-2016/de-politie-uw-vriend',
 177         'only_matching': True,
 178     }]
 179
 180     def _real_extract(self, url):
 181         mobj = re.match(self._VALID_URL, url)
 182         site_id, display_id = mobj.group('site_id'), mobj.group('id')
 183
 184         webpage = self._download_webpage(url, display_id)
 185
 186         title = strip_or_none(self._search_regex(
 187             r'<h1[^>]+class="video__body__header__title"[^>]*>(.+?)</h1>',
 188             webpage, 'title', default=None) or self._og_search_title(
 189             webpage, default=None))
 190
 191         video_id = self._html_search_regex(
 192             r'data-video=(["\'])(?P<id>(?:(?!\1).)+)\1', webpage, 'video id',
 193             group='id')
 194
 195         return {
 196             '_type': 'url_transparent',
 197             'url': 'https://mediazone.vrt.be/api/v1/%s/assets/%s' % (site_id, video_id),
 198             'ie_key': CanvasIE.ie_key(),
 199             'id': video_id,
 200             'display_id': display_id,
 201             'title': title,
 202             'description': self._og_search_description(webpage),
 203         }
 204
 205
 206 class VrtNUIE(GigyaBaseIE):
 207     IE_DESC = 'VrtNU.be'
 208     _VALID_URL = r'https?://(?:www\.)?vrt\.be/(?P<site_id>vrtnu)/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 209     _TESTS = [{
 210         # Available via old API endpoint
 211         'url': 'https://www.vrt.be/vrtnu/a-z/postbus-x/1/postbus-x-s1a1/',
 212         'info_dict': {
 213             'id': 'pbs-pub-2e2d8c27-df26-45c9-9dc6-90c78153044d$vid-90c932b1-e21d-4fb8-99b1-db7b49cf74de',
 214             'ext': 'mp4',
 215             'title': 'De zwarte weduwe',
 216             'description': 'md5:db1227b0f318c849ba5eab1fef895ee4',
 217             'duration': 1457.04,
 218             'thumbnail': r're:^https?://.*\.jpg$',
 219             'season': 'Season 1',
 220             'season_number': 1,
 221             'episode_number': 1,
 222         },
 223         'skip': 'This video is only available for registered users',
 224         'params': {
 225             'username': '<snip>',
 226             'password': '<snip>',
 227         },
 228         'expected_warnings': ['is not a supported codec'],
 229     }, {
 230         # Only available via new API endpoint
 231         'url': 'https://www.vrt.be/vrtnu/a-z/kamp-waes/1/kamp-waes-s1a5/',
 232         'info_dict': {
 233             'id': 'pbs-pub-0763b56c-64fb-4d38-b95b-af60bf433c71$vid-ad36a73c-4735-4f1f-b2c0-a38e6e6aa7e1',
 234             'ext': 'mp4',
 235             'title': 'Aflevering 5',
 236             'description': 'Wie valt door de mand tijdens een missie?',
 237             'duration': 2967.06,
 238             'season': 'Season 1',
 239             'season_number': 1,
 240             'episode_number': 5,
 241         },
 242         'skip': 'This video is only available for registered users',
 243         'params': {
 244             'username': '<snip>',
 245             'password': '<snip>',
 246         },
 247         'expected_warnings': ['Unable to download asset JSON', 'is not a supported codec', 'Unknown MIME type'],
 248     }]
 249     _NETRC_MACHINE = 'vrtnu'
 250     _APIKEY = '3_0Z2HujMtiWq_pkAjgnS2Md2E11a1AwZjYiBETtwNE-EoEHDINgtnvcAOpNgmrVGy'
 251     _CONTEXT_ID = 'R3595707040'
 252
 253     def _real_initialize(self):
 254         self._login()
 255
 256     def _login(self):
 257         username, password = self._get_login_info()
 258         if username is None:
 259             return
 260
 261         auth_data = {
 262             'APIKey': self._APIKEY,
 263             'targetEnv': 'jssdk',
 264             'loginID': username,
 265             'password': password,
 266             'authMode': 'cookie',
 267         }
 268
 269         auth_info = self._gigya_login(auth_data)
 270
 271         # Sometimes authentication fails for no good reason, retry
 272         login_attempt = 1
 273         while login_attempt <= 3:
 274             try:
 275                 # When requesting a token, no actual token is returned, but the
 276                 # necessary cookies are set.
 277                 self._request_webpage(
 278                     'https://token.vrt.be',
 279                     None, note='Requesting a token', errnote='Could not get a token',
 280                     headers={
 281                         'Content-Type': 'application/json',
 282                         'Referer': 'https://www.vrt.be/vrtnu/',
 283                     },
 284                     data=json.dumps({
 285                         'uid': auth_info['UID'],
 286                         'uidsig': auth_info['UIDSignature'],
 287                         'ts': auth_info['signatureTimestamp'],
 288                         'email': auth_info['profile']['email'],
 289                     }).encode('utf-8'))
 290             except ExtractorError as e:
 291                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
 292                     login_attempt += 1
 293                     self.report_warning('Authentication failed')
 294                     self._sleep(1, None, msg_template='Waiting for %(timeout)s seconds before trying again')
 295                 else:
 296                     raise e
 297             else:
 298                 break
 299
 300     def _real_extract(self, url):
 301         display_id = self._match_id(url)
 302
 303         webpage, urlh = self._download_webpage_handle(url, display_id)
 304
 305         info = self._search_json_ld(webpage, display_id, default={})
 306
 307         # title is optional here since it may be extracted by extractor
 308         # that is delegated from here
 309         title = strip_or_none(self._html_search_regex(
 310             r'(?ms)<h1 class="content__heading">(.+?)</h1>',
 311             webpage, 'title', default=None))
 312
 313         description = self._html_search_regex(
 314             r'(?ms)<div class="content__description">(.+?)</div>',
 315             webpage, 'description', default=None)
 316
 317         season = self._html_search_regex(
 318             [r'''(?xms)<div\ class="tabs__tab\ tabs__tab--active">\s*
 319                     <span>seizoen\ (.+?)</span>\s*
 320                 </div>''',
 321              r'<option value="seizoen (\d{1,3})" data-href="[^"]+?" selected>'],
 322             webpage, 'season', default=None)
 323
 324         season_number = int_or_none(season)
 325
 326         episode_number = int_or_none(self._html_search_regex(
 327             r'''(?xms)<div\ class="content__episode">\s*
 328                     <abbr\ title="aflevering">afl</abbr>\s*<span>(\d+)</span>
 329                 </div>''',
 330             webpage, 'episode_number', default=None))
 331
 332         release_date = parse_iso8601(self._html_search_regex(
 333             r'(?ms)<div class="content__broadcastdate">\s*<time\ datetime="(.+?)"',
 334             webpage, 'release_date', default=None))
 335
 336         # If there's a ? or a # in the URL, remove them and everything after
 337         clean_url = urlh.geturl().split('?')[0].split('#')[0].strip('/')
 338         securevideo_url = clean_url + '.mssecurevideo.json'
 339
 340         try:
 341             video = self._download_json(securevideo_url, display_id)
 342         except ExtractorError as e:
 343             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
 344                 self.raise_login_required()
 345             raise
 346
 347         # We are dealing with a '../<show>.relevant' URL
 348         redirect_url = video.get('url')
 349         if redirect_url:
 350             return self.url_result(self._proto_relative_url(redirect_url, 'https:'))
 351
 352         # There is only one entry, but with an unknown key, so just get
 353         # the first one
 354         video_id = list(video.values())[0].get('videoid')
 355
 356         return merge_dicts(info, {
 357             '_type': 'url_transparent',
 358             'url': 'https://mediazone.vrt.be/api/v1/vrtvideo/assets/%s' % video_id,
 359             'ie_key': CanvasIE.ie_key(),
 360             'id': video_id,
 361             'display_id': display_id,
 362             'title': title,
 363             'description': description,
 364             'season': season,
 365             'season_number': season_number,
 366             'episode_number': episode_number,
 367             'release_date': release_date,
 368         })