yt_dlp/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     clean_html,
  12     extract_attributes,
  13     float_or_none,
  14     get_element_by_class,
  15     int_or_none,
  16     merge_dicts,
  17     str_or_none,
  18     strip_or_none,
  19     url_or_none,
  20 )
  21
  22
  23 class CanvasIE(InfoExtractor):
  24     _VALID_URL = r'https?://mediazone\.vrt\.be/api/v1/(?P<site_id>canvas|een|ketnet|vrt(?:video|nieuws)|sporza|dako)/assets/(?P<id>[^/?#&]+)'
  25     _TESTS = [{
  26         'url': 'https://mediazone.vrt.be/api/v1/ketnet/assets/md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
  27         'md5': '68993eda72ef62386a15ea2cf3c93107',
  28         'info_dict': {
  29             'id': 'md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
  30             'display_id': 'md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
  31             'ext': 'mp4',
  32             'title': 'Nachtwacht: De Greystook',
  33             'description': 'Nachtwacht: De Greystook',
  34             'thumbnail': r're:^https?://.*\.jpg$',
  35             'duration': 1468.04,
  36         },
  37         'expected_warnings': ['is not a supported codec', 'Unknown MIME type'],
  38     }, {
  39         'url': 'https://mediazone.vrt.be/api/v1/canvas/assets/mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
  40         'only_matching': True,
  41     }]
  42     _GEO_BYPASS = False
  43     _HLS_ENTRY_PROTOCOLS_MAP = {
  44         'HLS': 'm3u8_native',
  45         'HLS_AES': 'm3u8',
  46     }
  47     _REST_API_BASE = 'https://media-services-public.vrt.be/vualto-video-aggregator-web/rest/external/v1'
  48
  49     def _real_extract(self, url):
  50         mobj = re.match(self._VALID_URL, url)
  51         site_id, video_id = mobj.group('site_id'), mobj.group('id')
  52
  53         data = None
  54         if site_id != 'vrtvideo':
  55             # Old API endpoint, serves more formats but may fail for some videos
  56             data = self._download_json(
  57                 'https://mediazone.vrt.be/api/v1/%s/assets/%s'
  58                 % (site_id, video_id), video_id, 'Downloading asset JSON',
  59                 'Unable to download asset JSON', fatal=False)
  60
  61         # New API endpoint
  62         if not data:
  63             headers = self.geo_verification_headers()
  64             headers.update({'Content-Type': 'application/json'})
  65             token = self._download_json(
  66                 '%s/tokens' % self._REST_API_BASE, video_id,
  67                 'Downloading token', data=b'', headers=headers)['vrtPlayerToken']
  68             data = self._download_json(
  69                 '%s/videos/%s' % (self._REST_API_BASE, video_id),
  70                 video_id, 'Downloading video JSON', query={
  71                     'vrtPlayerToken': token,
  72                     'client': '%s@PROD' % site_id,
  73                 }, expected_status=400)
  74             if not data.get('title'):
  75                 code = data.get('code')
  76                 if code == 'AUTHENTICATION_REQUIRED':
  77                     self.raise_login_required()
  78                 elif code == 'INVALID_LOCATION':
  79                     self.raise_geo_restricted(countries=['BE'])
  80                 raise ExtractorError(data.get('message') or code, expected=True)
  81
  82         title = data['title']
  83         description = data.get('description')
  84
  85         formats = []
  86         for target in data['targetUrls']:
  87             format_url, format_type = url_or_none(target.get('url')), str_or_none(target.get('type'))
  88             if not format_url or not format_type:
  89                 continue
  90             format_type = format_type.upper()
  91             if format_type in self._HLS_ENTRY_PROTOCOLS_MAP:
  92                 formats.extend(self._extract_m3u8_formats(
  93                     format_url, video_id, 'mp4', self._HLS_ENTRY_PROTOCOLS_MAP[format_type],
  94                     m3u8_id=format_type, fatal=False))
  95             elif format_type == 'HDS':
  96                 formats.extend(self._extract_f4m_formats(
  97                     format_url, video_id, f4m_id=format_type, fatal=False))
  98             elif format_type == 'MPEG_DASH':
  99                 formats.extend(self._extract_mpd_formats(
 100                     format_url, video_id, mpd_id=format_type, fatal=False))
 101             elif format_type == 'HSS':
 102                 formats.extend(self._extract_ism_formats(
 103                     format_url, video_id, ism_id='mss', fatal=False))
 104             else:
 105                 formats.append({
 106                     'format_id': format_type,
 107                     'url': format_url,
 108                 })
 109         self._sort_formats(formats)
 110
 111         subtitles = {}
 112         subtitle_urls = data.get('subtitleUrls')
 113         if isinstance(subtitle_urls, list):
 114             for subtitle in subtitle_urls:
 115                 subtitle_url = subtitle.get('url')
 116                 if subtitle_url and subtitle.get('type') == 'CLOSED':
 117                     subtitles.setdefault('nl', []).append({'url': subtitle_url})
 118
 119         return {
 120             'id': video_id,
 121             'display_id': video_id,
 122             'title': title,
 123             'description': description,
 124             'formats': formats,
 125             'duration': float_or_none(data.get('duration'), 1000),
 126             'thumbnail': data.get('posterImageUrl'),
 127             'subtitles': subtitles,
 128         }
 129
 130
 131 class CanvasEenIE(InfoExtractor):
 132     IE_DESC = 'canvas.be and een.be'
 133     _VALID_URL = r'https?://(?:www\.)?(?P<site_id>canvas|een)\.be/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 134     _TESTS = [{
 135         'url': 'http://www.canvas.be/video/de-afspraak/najaar-2015/de-afspraak-veilt-voor-de-warmste-week',
 136         'md5': 'ed66976748d12350b118455979cca293',
 137         'info_dict': {
 138             'id': 'mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
 139             'display_id': 'de-afspraak-veilt-voor-de-warmste-week',
 140             'ext': 'flv',
 141             'title': 'De afspraak veilt voor de Warmste Week',
 142             'description': 'md5:24cb860c320dc2be7358e0e5aa317ba6',
 143             'thumbnail': r're:^https?://.*\.jpg$',
 144             'duration': 49.02,
 145         },
 146         'expected_warnings': ['is not a supported codec'],
 147     }, {
 148         # with subtitles
 149         'url': 'http://www.canvas.be/video/panorama/2016/pieter-0167',
 150         'info_dict': {
 151             'id': 'mz-ast-5240ff21-2d30-4101-bba6-92b5ec67c625',
 152             'display_id': 'pieter-0167',
 153             'ext': 'mp4',
 154             'title': 'Pieter 0167',
 155             'description': 'md5:943cd30f48a5d29ba02c3a104dc4ec4e',
 156             'thumbnail': r're:^https?://.*\.jpg$',
 157             'duration': 2553.08,
 158             'subtitles': {
 159                 'nl': [{
 160                     'ext': 'vtt',
 161                 }],
 162             },
 163         },
 164         'params': {
 165             'skip_download': True,
 166         },
 167         'skip': 'Pagina niet gevonden',
 168     }, {
 169         'url': 'https://www.een.be/thuis/emma-pakt-thilly-aan',
 170         'info_dict': {
 171             'id': 'md-ast-3a24ced2-64d7-44fb-b4ed-ed1aafbf90b8',
 172             'display_id': 'emma-pakt-thilly-aan',
 173             'ext': 'mp4',
 174             'title': 'Emma pakt Thilly aan',
 175             'description': 'md5:c5c9b572388a99b2690030afa3f3bad7',
 176             'thumbnail': r're:^https?://.*\.jpg$',
 177             'duration': 118.24,
 178         },
 179         'params': {
 180             'skip_download': True,
 181         },
 182         'expected_warnings': ['is not a supported codec'],
 183     }, {
 184         'url': 'https://www.canvas.be/check-point/najaar-2016/de-politie-uw-vriend',
 185         'only_matching': True,
 186     }]
 187
 188     def _real_extract(self, url):
 189         mobj = re.match(self._VALID_URL, url)
 190         site_id, display_id = mobj.group('site_id'), mobj.group('id')
 191
 192         webpage = self._download_webpage(url, display_id)
 193
 194         title = strip_or_none(self._search_regex(
 195             r'<h1[^>]+class="video__body__header__title"[^>]*>(.+?)</h1>',
 196             webpage, 'title', default=None) or self._og_search_title(
 197             webpage, default=None))
 198
 199         video_id = self._html_search_regex(
 200             r'data-video=(["\'])(?P<id>(?:(?!\1).)+)\1', webpage, 'video id',
 201             group='id')
 202
 203         return {
 204             '_type': 'url_transparent',
 205             'url': 'https://mediazone.vrt.be/api/v1/%s/assets/%s' % (site_id, video_id),
 206             'ie_key': CanvasIE.ie_key(),
 207             'id': video_id,
 208             'display_id': display_id,
 209             'title': title,
 210             'description': self._og_search_description(webpage),
 211         }
 212
 213
 214 class VrtNUIE(GigyaBaseIE):
 215     IE_DESC = 'VrtNU.be'
 216     _VALID_URL = r'https?://(?:www\.)?vrt\.be/vrtnu/a-z/(?:[^/]+/){2}(?P<id>[^/?#&]+)'
 217     _TESTS = [{
 218         # Available via old API endpoint
 219         'url': 'https://www.vrt.be/vrtnu/a-z/postbus-x/1989/postbus-x-s1989a1/',
 220         'info_dict': {
 221             'id': 'pbs-pub-e8713dac-899e-41de-9313-81269f4c04ac$vid-90c932b1-e21d-4fb8-99b1-db7b49cf74de',
 222             'ext': 'mp4',
 223             'title': 'Postbus X - Aflevering 1 (Seizoen 1989)',
 224             'description': 'md5:b704f669eb9262da4c55b33d7c6ed4b7',
 225             'duration': 1457.04,
 226             'thumbnail': r're:^https?://.*\.jpg$',
 227             'series': 'Postbus X',
 228             'season': 'Seizoen 1989',
 229             'season_number': 1989,
 230             'episode': 'De zwarte weduwe',
 231             'episode_number': 1,
 232             'timestamp': 1595822400,
 233             'upload_date': '20200727',
 234         },
 235         'skip': 'This video is only available for registered users',
 236         'params': {
 237             'username': '<snip>',
 238             'password': '<snip>',
 239         },
 240         'expected_warnings': ['is not a supported codec'],
 241     }, {
 242         # Only available via new API endpoint
 243         'url': 'https://www.vrt.be/vrtnu/a-z/kamp-waes/1/kamp-waes-s1a5/',
 244         'info_dict': {
 245             'id': 'pbs-pub-0763b56c-64fb-4d38-b95b-af60bf433c71$vid-ad36a73c-4735-4f1f-b2c0-a38e6e6aa7e1',
 246             'ext': 'mp4',
 247             'title': 'Aflevering 5',
 248             'description': 'Wie valt door de mand tijdens een missie?',
 249             'duration': 2967.06,
 250             'season': 'Season 1',
 251             'season_number': 1,
 252             'episode_number': 5,
 253         },
 254         'skip': 'This video is only available for registered users',
 255         'params': {
 256             'username': '<snip>',
 257             'password': '<snip>',
 258         },
 259         'expected_warnings': ['Unable to download asset JSON', 'is not a supported codec', 'Unknown MIME type'],
 260     }]
 261     _NETRC_MACHINE = 'vrtnu'
 262     _APIKEY = '3_0Z2HujMtiWq_pkAjgnS2Md2E11a1AwZjYiBETtwNE-EoEHDINgtnvcAOpNgmrVGy'
 263     _CONTEXT_ID = 'R3595707040'
 264
 265     def _real_initialize(self):
 266         self._login()
 267
 268     def _login(self):
 269         username, password = self._get_login_info()
 270         if username is None:
 271             return
 272
 273         auth_data = {
 274             'APIKey': self._APIKEY,
 275             'targetEnv': 'jssdk',
 276             'loginID': username,
 277             'password': password,
 278             'authMode': 'cookie',
 279         }
 280
 281         auth_info = self._gigya_login(auth_data)
 282
 283         # Sometimes authentication fails for no good reason, retry
 284         login_attempt = 1
 285         while login_attempt <= 3:
 286             try:
 287                 # When requesting a token, no actual token is returned, but the
 288                 # necessary cookies are set.
 289                 self._request_webpage(
 290                     'https://token.vrt.be',
 291                     None, note='Requesting a token', errnote='Could not get a token',
 292                     headers={
 293                         'Content-Type': 'application/json',
 294                         'Referer': 'https://www.vrt.be/vrtnu/',
 295                     },
 296                     data=json.dumps({
 297                         'uid': auth_info['UID'],
 298                         'uidsig': auth_info['UIDSignature'],
 299                         'ts': auth_info['signatureTimestamp'],
 300                         'email': auth_info['profile']['email'],
 301                     }).encode('utf-8'))
 302             except ExtractorError as e:
 303                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
 304                     login_attempt += 1
 305                     self.report_warning('Authentication failed')
 306                     self._sleep(1, None, msg_template='Waiting for %(timeout)s seconds before trying again')
 307                 else:
 308                     raise e
 309             else:
 310                 break
 311
 312     def _real_extract(self, url):
 313         display_id = self._match_id(url)
 314
 315         webpage = self._download_webpage(url, display_id)
 316
 317         attrs = extract_attributes(self._search_regex(
 318             r'(<nui-media[^>]+>)', webpage, 'media element'))
 319         video_id = attrs['videoid']
 320         publication_id = attrs.get('publicationid')
 321         if publication_id:
 322             video_id = publication_id + '$' + video_id
 323
 324         page = (self._parse_json(self._search_regex(
 325             r'digitalData\s*=\s*({.+?});', webpage, 'digial data',
 326             default='{}'), video_id, fatal=False) or {}).get('page') or {}
 327
 328         info = self._search_json_ld(webpage, display_id, default={})
 329         return merge_dicts(info, {
 330             '_type': 'url_transparent',
 331             'url': 'https://mediazone.vrt.be/api/v1/vrtvideo/assets/%s' % video_id,
 332             'ie_key': CanvasIE.ie_key(),
 333             'id': video_id,
 334             'display_id': display_id,
 335             'season_number': int_or_none(page.get('episode_season')),
 336         })
 337
 338
 339 class DagelijkseKostIE(InfoExtractor):
 340     IE_DESC = 'dagelijksekost.een.be'
 341     _VALID_URL = r'https?://dagelijksekost\.een\.be/gerechten/(?P<id>[^/?#&]+)'
 342     _TEST = {
 343         'url': 'https://dagelijksekost.een.be/gerechten/hachis-parmentier-met-witloof',
 344         'md5': '30bfffc323009a3e5f689bef6efa2365',
 345         'info_dict': {
 346             'id': 'md-ast-27a4d1ff-7d7b-425e-b84f-a4d227f592fa',
 347             'display_id': 'hachis-parmentier-met-witloof',
 348             'ext': 'mp4',
 349             'title': 'Hachis parmentier met witloof',
 350             'description': 'md5:9960478392d87f63567b5b117688cdc5',
 351             'thumbnail': r're:^https?://.*\.jpg$',
 352             'duration': 283.02,
 353         },
 354         'expected_warnings': ['is not a supported codec'],
 355     }
 356
 357     def _real_extract(self, url):
 358         display_id = self._match_id(url)
 359         webpage = self._download_webpage(url, display_id)
 360
 361         title = strip_or_none(get_element_by_class(
 362             'dish-metadata__title', webpage
 363         ) or self._html_search_meta(
 364             'twitter:title', webpage))
 365
 366         description = clean_html(get_element_by_class(
 367             'dish-description', webpage)
 368         ) or self._html_search_meta(
 369             ('description', 'twitter:description', 'og:description'),
 370             webpage)
 371
 372         video_id = self._html_search_regex(
 373             r'data-url=(["\'])(?P<id>(?:(?!\1).)+)\1', webpage, 'video id',
 374             group='id')
 375
 376         return {
 377             '_type': 'url_transparent',
 378             'url': 'https://mediazone.vrt.be/api/v1/dako/assets/%s' % video_id,
 379             'ie_key': CanvasIE.ie_key(),
 380             'id': video_id,
 381             'display_id': display_id,
 382             'title': title,
 383             'description': description,
 384         }