yt_dlp/extractor/wistia.py

   1 import re
   2 import urllib.error
   3 import urllib.parse
   4 from base64 import b64decode
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     ExtractorError,
   9     float_or_none,
  10     int_or_none,
  11     parse_qs,
  12     traverse_obj,
  13     try_get,
  14     update_url_query,
  15 )
  16
  17
  18 class WistiaBaseIE(InfoExtractor):
  19     _VALID_ID_REGEX = r'(?P<id>[a-z0-9]{10})'
  20     _VALID_URL_BASE = r'https?://(?:\w+\.)?wistia\.(?:net|com)/(?:embed/)?'
  21     _EMBED_BASE_URL = 'http://fast.wistia.net/embed/'
  22
  23     def _download_embed_config(self, config_type, config_id, referer):
  24         base_url = self._EMBED_BASE_URL + '%s/%s' % (config_type, config_id)
  25         embed_config = self._download_json(
  26             base_url + '.json', config_id, headers={
  27                 'Referer': referer if referer.startswith('http') else base_url,  # Some videos require this.
  28             })
  29
  30         error = traverse_obj(embed_config, 'error')
  31         if error:
  32             raise ExtractorError(
  33                 f'Error while getting the playlist: {error}', expected=True)
  34
  35         return embed_config
  36
  37     def _extract_media(self, embed_config):
  38         data = embed_config['media']
  39         video_id = data['hashedId']
  40         title = data['name']
  41
  42         formats = []
  43         thumbnails = []
  44         for a in data['assets']:
  45             aurl = a.get('url')
  46             if not aurl:
  47                 continue
  48             astatus = a.get('status')
  49             atype = a.get('type')
  50             if (astatus is not None and astatus != 2) or atype in ('preview', 'storyboard'):
  51                 continue
  52             elif atype in ('still', 'still_image'):
  53                 thumbnails.append({
  54                     'url': aurl,
  55                     'width': int_or_none(a.get('width')),
  56                     'height': int_or_none(a.get('height')),
  57                     'filesize': int_or_none(a.get('size')),
  58                 })
  59             else:
  60                 aext = a.get('ext')
  61                 display_name = a.get('display_name')
  62                 format_id = atype
  63                 if atype and atype.endswith('_video') and display_name:
  64                     format_id = '%s-%s' % (atype[:-6], display_name)
  65                 f = {
  66                     'format_id': format_id,
  67                     'url': aurl,
  68                     'tbr': int_or_none(a.get('bitrate')) or None,
  69                     'quality': 1 if atype == 'original' else None,
  70                 }
  71                 if display_name == 'Audio':
  72                     f.update({
  73                         'vcodec': 'none',
  74                     })
  75                 else:
  76                     f.update({
  77                         'width': int_or_none(a.get('width')),
  78                         'height': int_or_none(a.get('height')),
  79                         'vcodec': a.get('codec'),
  80                     })
  81                 if a.get('container') == 'm3u8' or aext == 'm3u8':
  82                     ts_f = f.copy()
  83                     ts_f.update({
  84                         'ext': 'ts',
  85                         'format_id': f['format_id'].replace('hls-', 'ts-'),
  86                         'url': f['url'].replace('.bin', '.ts'),
  87                     })
  88                     formats.append(ts_f)
  89                     f.update({
  90                         'ext': 'mp4',
  91                         'protocol': 'm3u8_native',
  92                     })
  93                 else:
  94                     f.update({
  95                         'container': a.get('container'),
  96                         'ext': aext,
  97                         'filesize': int_or_none(a.get('size')),
  98                     })
  99                 formats.append(f)
 100
 101         self._sort_formats(formats)
 102
 103         subtitles = {}
 104         for caption in data.get('captions', []):
 105             language = caption.get('language')
 106             if not language:
 107                 continue
 108             subtitles[language] = [{
 109                 'url': self._EMBED_BASE_URL + 'captions/' + video_id + '.vtt?language=' + language,
 110             }]
 111
 112         return {
 113             'id': video_id,
 114             'title': title,
 115             'description': data.get('seoDescription'),
 116             'formats': formats,
 117             'thumbnails': thumbnails,
 118             'duration': float_or_none(data.get('duration')),
 119             'timestamp': int_or_none(data.get('createdAt')),
 120             'subtitles': subtitles,
 121         }
 122
 123     @classmethod
 124     def _extract_from_webpage(cls, url, webpage):
 125         from .teachable import TeachableIE
 126
 127         if list(TeachableIE._extract_embed_urls(url, webpage)):
 128             return
 129
 130         yield from super()._extract_from_webpage(url, webpage)
 131
 132     @classmethod
 133     def _extract_wistia_async_embed(cls, webpage):
 134         # https://wistia.com/support/embed-and-share/video-on-your-website
 135         # https://wistia.com/support/embed-and-share/channel-embeds
 136         yield from re.finditer(
 137             r'''(?sx)
 138                 <(?:div|section)[^>]+class=([\"'])(?:(?!\1).)*?(?P<type>wistia[a-z_0-9]+)\s*\bwistia_async_(?P<id>[a-z0-9]{10})\b(?:(?!\1).)*?\1
 139             ''', webpage)
 140
 141     @classmethod
 142     def _extract_url_media_id(cls, url):
 143         mobj = re.search(r'(?:wmediaid|wvideo(?:id)?)]?=(?P<id>[a-z0-9]{10})', urllib.parse.unquote_plus(url))
 144         if mobj:
 145             return mobj.group('id')
 146
 147
 148 class WistiaIE(WistiaBaseIE):
 149     _VALID_URL = r'(?:wistia:|%s(?:iframe|medias)/)%s' % (WistiaBaseIE._VALID_URL_BASE, WistiaBaseIE._VALID_ID_REGEX)
 150     _EMBED_REGEX = [
 151         r'''(?x)
 152             <(?:meta[^>]+?content|(?:iframe|script)[^>]+?src)=["\']
 153             (?P<url>(?:https?:)?//(?:fast\.)?wistia\.(?:net|com)/embed/(?:iframe|medias)/[a-z0-9]{10})
 154             ''']
 155     _TESTS = [{
 156         # with hls video
 157         'url': 'wistia:807fafadvk',
 158         'md5': 'daff0f3687a41d9a71b40e0e8c2610fe',
 159         'info_dict': {
 160             'id': '807fafadvk',
 161             'ext': 'mp4',
 162             'title': 'Drip Brennan Dunn Workshop',
 163             'description': 'a JV Webinars video',
 164             'upload_date': '20160518',
 165             'timestamp': 1463607249,
 166             'duration': 4987.11,
 167         },
 168         'skip': 'video unavailable',
 169     }, {
 170         'url': 'wistia:a6ndpko1wg',
 171         'md5': '10c1ce9c4dde638202513ed17a3767bd',
 172         'info_dict': {
 173             'id': 'a6ndpko1wg',
 174             'ext': 'bin',
 175             'title': 'Episode 2: Boxed Water\'s retention is thirsty',
 176             'upload_date': '20210324',
 177             'description': 'md5:da5994c2c2d254833b412469d9666b7a',
 178             'duration': 966.0,
 179             'timestamp': 1616614369,
 180             'thumbnail': 'https://embed-ssl.wistia.com/deliveries/53dc60239348dc9b9fba3755173ea4c2.bin',
 181         }
 182     }, {
 183         'url': 'wistia:5vd7p4bct5',
 184         'md5': 'b9676d24bf30945d97060638fbfe77f0',
 185         'info_dict': {
 186             'id': '5vd7p4bct5',
 187             'ext': 'bin',
 188             'title': 'md5:eaa9f64c4efd7b5f098b9b6118597679',
 189             'description': 'md5:a9bea0315f0616aa5df2dc413ddcdd0f',
 190             'upload_date': '20220915',
 191             'timestamp': 1663258727,
 192             'duration': 623.019,
 193             'thumbnail': r're:https?://embed(?:-ssl)?.wistia.com/.+\.(?:jpg|bin)$',
 194         },
 195     }, {
 196         'url': 'wistia:sh7fpupwlt',
 197         'only_matching': True,
 198     }, {
 199         'url': 'http://fast.wistia.net/embed/iframe/sh7fpupwlt',
 200         'only_matching': True,
 201     }, {
 202         'url': 'http://fast.wistia.com/embed/iframe/sh7fpupwlt',
 203         'only_matching': True,
 204     }, {
 205         'url': 'http://fast.wistia.net/embed/medias/sh7fpupwlt.json',
 206         'only_matching': True,
 207     }]
 208
 209     _WEBPAGE_TESTS = [{
 210         'url': 'https://www.weidert.com/blog/wistia-channels-video-marketing-tool',
 211         'info_dict': {
 212             'id': 'cqwukac3z1',
 213             'ext': 'bin',
 214             'title': 'How Wistia Channels Can Help Capture Inbound Value From Your Video Content',
 215             'duration': 158.125,
 216             'timestamp': 1618974400,
 217             'description': 'md5:27abc99a758573560be72600ef95cece',
 218             'upload_date': '20210421',
 219             'thumbnail': 'https://embed-ssl.wistia.com/deliveries/6c551820ae950cdee2306d6cbe9ef742.bin',
 220         }
 221     }, {
 222         'url': 'https://study.com/academy/lesson/north-american-exploration-failed-colonies-of-spain-france-england.html#lesson',
 223         'md5': 'b9676d24bf30945d97060638fbfe77f0',
 224         'info_dict': {
 225             'id': '5vd7p4bct5',
 226             'ext': 'bin',
 227             'title': 'paywall_north-american-exploration-failed-colonies-of-spain-france-england',
 228             'upload_date': '20220915',
 229             'timestamp': 1663258727,
 230             'duration': 623.019,
 231             'thumbnail': 'https://embed-ssl.wistia.com/deliveries/83e6ec693e2c05a0ce65809cbaead86a.bin',
 232             'description': 'a Paywall Videos video',
 233         },
 234     }]
 235
 236     def _real_extract(self, url):
 237         video_id = self._match_id(url)
 238         embed_config = self._download_embed_config('medias', video_id, url)
 239         return self._extract_media(embed_config)
 240
 241     @classmethod
 242     def _extract_embed_urls(cls, url, webpage):
 243         urls = list(super()._extract_embed_urls(url, webpage))
 244         for match in cls._extract_wistia_async_embed(webpage):
 245             if match.group('type') != 'wistia_channel':
 246                 urls.append('wistia:%s' % match.group('id'))
 247         for match in re.finditer(r'(?:data-wistia-?id=["\']|Wistia\.embed\(["\']|id=["\']wistia_)(?P<id>[a-z0-9]{10})',
 248                                  webpage):
 249             urls.append('wistia:%s' % match.group('id'))
 250         if not WistiaChannelIE._extract_embed_urls(url, webpage):  # Fallback
 251             media_id = cls._extract_url_media_id(url)
 252             if media_id:
 253                 urls.append('wistia:%s' % match.group('id'))
 254         return urls
 255
 256
 257 class WistiaPlaylistIE(WistiaBaseIE):
 258     _VALID_URL = r'%splaylists/%s' % (WistiaBaseIE._VALID_URL_BASE, WistiaBaseIE._VALID_ID_REGEX)
 259
 260     _TEST = {
 261         'url': 'https://fast.wistia.net/embed/playlists/aodt9etokc',
 262         'info_dict': {
 263             'id': 'aodt9etokc',
 264         },
 265         'playlist_count': 3,
 266     }
 267
 268     def _real_extract(self, url):
 269         playlist_id = self._match_id(url)
 270         playlist = self._download_embed_config('playlists', playlist_id, url)
 271
 272         entries = []
 273         for media in (try_get(playlist, lambda x: x[0]['medias']) or []):
 274             embed_config = media.get('embed_config')
 275             if not embed_config:
 276                 continue
 277             entries.append(self._extract_media(embed_config))
 278
 279         return self.playlist_result(entries, playlist_id)
 280
 281
 282 class WistiaChannelIE(WistiaBaseIE):
 283     _VALID_URL = r'(?:wistiachannel:|%schannel/)%s' % (WistiaBaseIE._VALID_URL_BASE, WistiaBaseIE._VALID_ID_REGEX)
 284
 285     _TESTS = [{
 286         # JSON Embed API returns 403, should fall back to webpage
 287         'url': 'https://fast.wistia.net/embed/channel/yvyvu7wjbg?wchannelid=yvyvu7wjbg',
 288         'info_dict': {
 289             'id': 'yvyvu7wjbg',
 290             'title': 'Copysmith Tutorials and Education!',
 291             'description': 'Learn all things Copysmith via short and informative videos!'
 292         },
 293         'playlist_mincount': 7,
 294         'expected_warnings': ['falling back to webpage'],
 295     }, {
 296         'url': 'https://fast.wistia.net/embed/channel/3802iirk0l',
 297         'info_dict': {
 298             'id': '3802iirk0l',
 299             'title': 'The Roof',
 300         },
 301         'playlist_mincount': 20,
 302     }, {
 303         # link to popup video, follow --no-playlist
 304         'url': 'https://fast.wistia.net/embed/channel/3802iirk0l?wchannelid=3802iirk0l&wmediaid=sp5dqjzw3n',
 305         'info_dict': {
 306             'id': 'sp5dqjzw3n',
 307             'ext': 'bin',
 308             'title': 'The Roof S2: The Modern CRO',
 309             'thumbnail': 'https://embed-ssl.wistia.com/deliveries/dadfa9233eaa505d5e0c85c23ff70741.bin',
 310             'duration': 86.487,
 311             'description': 'A sales leader on The Roof? Man, they really must be letting anyone up here this season.\n',
 312             'timestamp': 1619790290,
 313             'upload_date': '20210430',
 314         },
 315         'params': {'noplaylist': True, 'skip_download': True},
 316     }]
 317     _WEBPAGE_TESTS = [{
 318         'url': 'https://www.profitwell.com/recur/boxed-out',
 319         'info_dict': {
 320             'id': '6jyvmqz6zs',
 321             'title': 'Boxed Out',
 322             'description': 'md5:14a8a93a1dbe236718e6a59f8c8c7bae',
 323         },
 324         'playlist_mincount': 30,
 325     }, {
 326         # section instead of div
 327         'url': 'https://360learning.com/studio/onboarding-joei/',
 328         'info_dict': {
 329             'id': 'z874k93n2o',
 330             'title': 'Onboarding Joei.',
 331             'description': 'Coming to you weekly starting Feb 19th.',
 332         },
 333         'playlist_mincount': 20,
 334     }, {
 335         'url': 'https://amplitude.com/amplify-sessions?amp%5Bwmediaid%5D=pz0m0l0if3&amp%5Bwvideo%5D=pz0m0l0if3&wchannelid=emyjmwjf79&wmediaid=i8um783bdt',
 336         'info_dict': {
 337             'id': 'pz0m0l0if3',
 338             'title': 'A Framework for Improving Product Team Performance',
 339             'ext': 'bin',
 340             'timestamp': 1653935275,
 341             'upload_date': '20220530',
 342             'description': 'Learn how to help your company improve and achieve your product related goals.',
 343             'duration': 1854.39,
 344             'thumbnail': 'https://embed-ssl.wistia.com/deliveries/12fd19e56413d9d6f04e2185c16a6f8854e25226.bin',
 345         },
 346         'params': {'noplaylist': True, 'skip_download': True},
 347     }]
 348
 349     def _real_extract(self, url):
 350         channel_id = self._match_id(url)
 351         media_id = self._extract_url_media_id(url)
 352         if not self._yes_playlist(channel_id, media_id, playlist_label='channel'):
 353             return self.url_result(f'wistia:{media_id}', 'Wistia')
 354
 355         try:
 356             data = self._download_embed_config('channel', channel_id, url)
 357         except (ExtractorError, urllib.error.HTTPError):
 358             # Some channels give a 403 from the JSON API
 359             self.report_warning('Failed to download channel data from API, falling back to webpage.')
 360             webpage = self._download_webpage(f'https://fast.wistia.net/embed/channel/{channel_id}', channel_id)
 361             data = self._parse_json(
 362                 self._search_regex(r'wchanneljsonp-%s\'\]\s*=[^\"]*\"([A-Za-z0-9=/]*)' % channel_id, webpage, 'jsonp', channel_id),
 363                 channel_id, transform_source=lambda x: urllib.parse.unquote_plus(b64decode(x).decode('utf-8')))
 364
 365         # XXX: can there be more than one series?
 366         series = traverse_obj(data, ('series', 0), default={})
 367
 368         entries = [
 369             self.url_result(f'wistia:{video["hashedId"]}', WistiaIE, title=video.get('name'))
 370             for video in traverse_obj(series, ('sections', ..., 'videos', ...)) or []
 371             if video.get('hashedId')
 372         ]
 373
 374         return self.playlist_result(
 375             entries, channel_id, playlist_title=series.get('title'), playlist_description=series.get('description'))
 376
 377     @classmethod
 378     def _extract_embed_urls(cls, url, webpage):
 379         yield from super()._extract_embed_urls(url, webpage)
 380         for match in cls._extract_wistia_async_embed(webpage):
 381             if match.group('type') == 'wistia_channel':
 382                 # original url may contain wmediaid query param
 383                 yield update_url_query(f'wistiachannel:{match.group("id")}', parse_qs(url))