yt_dlp/extractor/pluralsight.py

   1 import collections
   2 import json
   3 import os
   4 import random
   5 import re
   6 import urllib.parse
   7
   8 from .common import InfoExtractor
   9 from ..utils import (
  10     ExtractorError,
  11     dict_get,
  12     float_or_none,
  13     int_or_none,
  14     parse_duration,
  15     parse_qs,
  16     qualities,
  17     srt_subtitles_timecode,
  18     try_get,
  19     update_url_query,
  20     urlencode_postdata,
  21 )
  22
  23
  24 class PluralsightBaseIE(InfoExtractor):
  25     _API_BASE = 'https://app.pluralsight.com'
  26
  27     _GRAPHQL_EP = f'{_API_BASE}/player/api/graphql'
  28     _GRAPHQL_HEADERS = {
  29         'Content-Type': 'application/json;charset=UTF-8',
  30     }
  31     _GRAPHQL_COURSE_TMPL = '''
  32 query BootstrapPlayer {
  33   rpc {
  34     bootstrapPlayer {
  35       profile {
  36         firstName
  37         lastName
  38         email
  39         username
  40         userHandle
  41         authed
  42         isAuthed
  43         plan
  44       }
  45       course(courseId: "%s") {
  46         name
  47         title
  48         courseHasCaptions
  49         translationLanguages {
  50           code
  51           name
  52         }
  53         supportsWideScreenVideoFormats
  54         timestamp
  55         modules {
  56           name
  57           title
  58           duration
  59           formattedDuration
  60           author
  61           authorized
  62           clips {
  63             authorized
  64             clipId
  65             duration
  66             formattedDuration
  67             id
  68             index
  69             moduleIndex
  70             moduleTitle
  71             name
  72             title
  73             watched
  74           }
  75         }
  76       }
  77     }
  78   }
  79 }'''
  80
  81     def _download_course(self, course_id, url, display_id):
  82         try:
  83             return self._download_course_rpc(course_id, url, display_id)
  84         except ExtractorError:
  85             # Old API fallback
  86             return self._download_json(
  87                 'https://app.pluralsight.com/player/user/api/v1/player/payload',
  88                 display_id, data=urlencode_postdata({'courseId': course_id}),
  89                 headers={'Referer': url})
  90
  91     def _download_course_rpc(self, course_id, url, display_id):
  92         response = self._download_json(
  93             self._GRAPHQL_EP, display_id, data=json.dumps({
  94                 'query': self._GRAPHQL_COURSE_TMPL % course_id,
  95                 'variables': {},
  96             }).encode(), headers=self._GRAPHQL_HEADERS)
  97
  98         course = try_get(
  99             response, lambda x: x['data']['rpc']['bootstrapPlayer']['course'],
 100             dict)
 101         if course:
 102             return course
 103
 104         raise ExtractorError(
 105             '{} said: {}'.format(self.IE_NAME, response['error']['message']),
 106             expected=True)
 107
 108
 109 class PluralsightIE(PluralsightBaseIE):
 110     IE_NAME = 'pluralsight'
 111     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:training/)?player\?'
 112     _LOGIN_URL = 'https://app.pluralsight.com/id/'
 113
 114     _NETRC_MACHINE = 'pluralsight'
 115
 116     _TESTS = [{
 117         'url': 'http://www.pluralsight.com/training/player?author=mike-mckeown&name=hosting-sql-server-windows-azure-iaas-m7-mgmt&mode=live&clip=3&course=hosting-sql-server-windows-azure-iaas',
 118         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
 119         'info_dict': {
 120             'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04',
 121             'ext': 'mp4',
 122             'title': 'Demo Monitoring',
 123             'duration': 338,
 124         },
 125         'skip': 'Requires pluralsight account credentials',
 126     }, {
 127         'url': 'https://app.pluralsight.com/training/player?course=angularjs-get-started&author=scott-allen&name=angularjs-get-started-m1-introduction&clip=0&mode=live',
 128         'only_matching': True,
 129     }, {
 130         # available without pluralsight account
 131         'url': 'http://app.pluralsight.com/training/player?author=scott-allen&name=angularjs-get-started-m1-introduction&mode=live&clip=0&course=angularjs-get-started',
 132         'only_matching': True,
 133     }, {
 134         'url': 'https://app.pluralsight.com/player?course=ccna-intro-networking&author=ross-bagurdes&name=ccna-intro-networking-m06&clip=0',
 135         'only_matching': True,
 136     }]
 137
 138     GRAPHQL_VIEWCLIP_TMPL = '''
 139 query viewClip {
 140   viewClip(input: {
 141     author: "%(author)s",
 142     clipIndex: %(clipIndex)d,
 143     courseName: "%(courseName)s",
 144     includeCaptions: %(includeCaptions)s,
 145     locale: "%(locale)s",
 146     mediaType: "%(mediaType)s",
 147     moduleName: "%(moduleName)s",
 148     quality: "%(quality)s"
 149   }) {
 150     urls {
 151       url
 152       cdn
 153       rank
 154       source
 155     },
 156     status
 157   }
 158 }'''
 159
 160     def _perform_login(self, username, password):
 161         login_page = self._download_webpage(
 162             self._LOGIN_URL, None, 'Downloading login page')
 163
 164         login_form = self._hidden_inputs(login_page)
 165
 166         login_form.update({
 167             'Username': username,
 168             'Password': password,
 169         })
 170
 171         post_url = self._search_regex(
 172             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
 173             'post url', default=self._LOGIN_URL, group='url')
 174
 175         if not post_url.startswith('http'):
 176             post_url = urllib.parse.urljoin(self._LOGIN_URL, post_url)
 177
 178         response = self._download_webpage(
 179             post_url, None, 'Logging in',
 180             data=urlencode_postdata(login_form),
 181             headers={'Content-Type': 'application/x-www-form-urlencoded'})
 182
 183         error = self._search_regex(
 184             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',
 185             response, 'error message', default=None)
 186         if error:
 187             raise ExtractorError(f'Unable to login: {error}', expected=True)
 188
 189         if all(not re.search(p, response) for p in (
 190                 r'__INITIAL_STATE__', r'["\']currentUser["\']',
 191                 # new layout?
 192                 r'>\s*Sign out\s*<')):
 193             BLOCKED = 'Your account has been blocked due to suspicious activity'
 194             if BLOCKED in response:
 195                 raise ExtractorError(
 196                     f'Unable to login: {BLOCKED}', expected=True)
 197             MUST_AGREE = 'To continue using Pluralsight, you must agree to'
 198             if any(p in response for p in (MUST_AGREE, '>Disagree<', '>Agree<')):
 199                 raise ExtractorError(
 200                     f'Unable to login: {MUST_AGREE} some documents. Go to pluralsight.com, '
 201                     'log in and agree with what Pluralsight requires.', expected=True)
 202
 203             raise ExtractorError('Unable to log in')
 204
 205     def _get_subtitles(self, author, clip_idx, clip_id, lang, name, duration, video_id):
 206         captions = None
 207         if clip_id:
 208             captions = self._download_json(
 209                 f'{self._API_BASE}/transcript/api/v1/caption/json/{clip_id}/{lang}', video_id,
 210                 'Downloading captions JSON', 'Unable to download captions JSON',
 211                 fatal=False)
 212         if not captions:
 213             captions_post = {
 214                 'a': author,
 215                 'cn': int(clip_idx),
 216                 'lc': lang,
 217                 'm': name,
 218             }
 219             captions = self._download_json(
 220                 f'{self._API_BASE}/player/retrieve-captions', video_id,
 221                 'Downloading captions JSON', 'Unable to download captions JSON',
 222                 fatal=False, data=json.dumps(captions_post).encode(),
 223                 headers={'Content-Type': 'application/json;charset=utf-8'})
 224         if captions:
 225             return {
 226                 lang: [{
 227                     'ext': 'json',
 228                     'data': json.dumps(captions),
 229                 }, {
 230                     'ext': 'srt',
 231                     'data': self._convert_subtitles(duration, captions),
 232                 }],
 233             }
 234
 235     @staticmethod
 236     def _convert_subtitles(duration, subs):
 237         srt = ''
 238         TIME_OFFSET_KEYS = ('displayTimeOffset', 'DisplayTimeOffset')
 239         TEXT_KEYS = ('text', 'Text')
 240         for num, current in enumerate(subs):
 241             current = subs[num]
 242             start, text = (
 243                 float_or_none(dict_get(current, TIME_OFFSET_KEYS, skip_false_values=False)),
 244                 dict_get(current, TEXT_KEYS))
 245             if start is None or text is None:
 246                 continue
 247             end = duration if num == len(subs) - 1 else float_or_none(
 248                 dict_get(subs[num + 1], TIME_OFFSET_KEYS, skip_false_values=False))
 249             if end is None:
 250                 continue
 251             srt += os.linesep.join(
 252                 (
 253                     f'{num}',
 254                     f'{srt_subtitles_timecode(start)} --> {srt_subtitles_timecode(end)}',
 255                     text,
 256                     os.linesep,
 257                 ))
 258         return srt
 259
 260     def _real_extract(self, url):
 261         qs = parse_qs(url)
 262
 263         author = qs.get('author', [None])[0]
 264         name = qs.get('name', [None])[0]
 265         clip_idx = qs.get('clip', [None])[0]
 266         course_name = qs.get('course', [None])[0]
 267
 268         if any(not f for f in (author, name, clip_idx, course_name)):
 269             raise ExtractorError('Invalid URL', expected=True)
 270
 271         display_id = f'{name}-{clip_idx}'
 272
 273         course = self._download_course(course_name, url, display_id)
 274
 275         collection = course['modules']
 276
 277         clip = None
 278
 279         for module_ in collection:
 280             if name in (module_.get('moduleName'), module_.get('name')):
 281                 for clip_ in module_.get('clips', []):
 282                     clip_index = clip_.get('clipIndex')
 283                     if clip_index is None:
 284                         clip_index = clip_.get('index')
 285                     if clip_index is None:
 286                         continue
 287                     if str(clip_index) == clip_idx:
 288                         clip = clip_
 289                         break
 290
 291         if not clip:
 292             raise ExtractorError('Unable to resolve clip')
 293
 294         title = clip['title']
 295         clip_id = clip.get('clipName') or clip.get('name') or clip['clipId']
 296
 297         QUALITIES = {
 298             'low': {'width': 640, 'height': 480},
 299             'medium': {'width': 848, 'height': 640},
 300             'high': {'width': 1024, 'height': 768},
 301             'high-widescreen': {'width': 1280, 'height': 720},
 302         }
 303
 304         QUALITIES_PREFERENCE = ('low', 'medium', 'high', 'high-widescreen')
 305         quality_key = qualities(QUALITIES_PREFERENCE)
 306
 307         AllowedQuality = collections.namedtuple('AllowedQuality', ['ext', 'qualities'])
 308
 309         ALLOWED_QUALITIES = (
 310             AllowedQuality('webm', ['high']),
 311             AllowedQuality('mp4', ['low', 'medium', 'high']),
 312         )
 313
 314         # Some courses also offer widescreen resolution for high quality (see
 315         # https://github.com/ytdl-org/youtube-dl/issues/7766)
 316         widescreen = course.get('supportsWideScreenVideoFormats') is True
 317         best_quality = 'high-widescreen' if widescreen else 'high'
 318         if widescreen:
 319             for allowed_quality in ALLOWED_QUALITIES:
 320                 allowed_quality.qualities.append(best_quality)
 321
 322         # In order to minimize the number of calls to ViewClip API and reduce
 323         # the probability of being throttled or banned by Pluralsight we will request
 324         # only single format until formats listing was explicitly requested.
 325         if self.get_param('listformats', False):
 326             allowed_qualities = ALLOWED_QUALITIES
 327         else:
 328             def guess_allowed_qualities():
 329                 req_format = self.get_param('format') or 'best'
 330                 req_format_split = req_format.split('-', 1)
 331                 if len(req_format_split) > 1:
 332                     req_ext, req_quality = req_format_split
 333                     req_quality = '-'.join(req_quality.split('-')[:2])
 334                     for allowed_quality in ALLOWED_QUALITIES:
 335                         if req_ext == allowed_quality.ext and req_quality in allowed_quality.qualities:
 336                             return (AllowedQuality(req_ext, (req_quality, )), )
 337                 req_ext = 'webm' if self.get_param('prefer_free_formats') else 'mp4'
 338                 return (AllowedQuality(req_ext, (best_quality, )), )
 339             allowed_qualities = guess_allowed_qualities()
 340
 341         formats = []
 342         for ext, qualities_ in allowed_qualities:
 343             for quality in qualities_:
 344                 f = QUALITIES[quality].copy()
 345                 clip_post = {
 346                     'author': author,
 347                     'includeCaptions': 'false',
 348                     'clipIndex': int(clip_idx),
 349                     'courseName': course_name,
 350                     'locale': 'en',
 351                     'moduleName': name,
 352                     'mediaType': ext,
 353                     'quality': '%dx%d' % (f['width'], f['height']),
 354                 }
 355                 format_id = f'{ext}-{quality}'
 356
 357                 try:
 358                     viewclip = self._download_json(
 359                         self._GRAPHQL_EP, display_id,
 360                         f'Downloading {format_id} viewclip graphql',
 361                         data=json.dumps({
 362                             'query': self.GRAPHQL_VIEWCLIP_TMPL % clip_post,
 363                             'variables': {},
 364                         }).encode(),
 365                         headers=self._GRAPHQL_HEADERS)['data']['viewClip']
 366                 except ExtractorError:
 367                     # Still works but most likely will go soon
 368                     viewclip = self._download_json(
 369                         f'{self._API_BASE}/video/clips/viewclip', display_id,
 370                         f'Downloading {format_id} viewclip JSON', fatal=False,
 371                         data=json.dumps(clip_post).encode(),
 372                         headers={'Content-Type': 'application/json;charset=utf-8'})
 373
 374                 # Pluralsight tracks multiple sequential calls to ViewClip API and start
 375                 # to return 429 HTTP errors after some time (see
 376                 # https://github.com/ytdl-org/youtube-dl/pull/6989). Moreover it may even lead
 377                 # to account ban (see https://github.com/ytdl-org/youtube-dl/issues/6842).
 378                 # To somewhat reduce the probability of these consequences
 379                 # we will sleep random amount of time before each call to ViewClip.
 380                 self._sleep(
 381                     random.randint(5, 10), display_id,
 382                     '%(video_id)s: Waiting for %(timeout)s seconds to avoid throttling')
 383
 384                 if not viewclip:
 385                     continue
 386
 387                 clip_urls = viewclip.get('urls')
 388                 if not isinstance(clip_urls, list):
 389                     continue
 390
 391                 for clip_url_data in clip_urls:
 392                     clip_url = clip_url_data.get('url')
 393                     if not clip_url:
 394                         continue
 395                     cdn = clip_url_data.get('cdn')
 396                     clip_f = f.copy()
 397                     clip_f.update({
 398                         'url': clip_url,
 399                         'ext': ext,
 400                         'format_id': f'{format_id}-{cdn}' if cdn else format_id,
 401                         'quality': quality_key(quality),
 402                         'source_preference': int_or_none(clip_url_data.get('rank')),
 403                     })
 404                     formats.append(clip_f)
 405
 406         duration = int_or_none(
 407             clip.get('duration')) or parse_duration(clip.get('formattedDuration'))
 408
 409         # TODO: other languages?
 410         subtitles = self.extract_subtitles(
 411             author, clip_idx, clip.get('clipId'), 'en', name, duration, display_id)
 412
 413         return {
 414             'id': clip_id,
 415             'title': title,
 416             'duration': duration,
 417             'creator': author,
 418             'formats': formats,
 419             'subtitles': subtitles,
 420         }
 421
 422
 423 class PluralsightCourseIE(PluralsightBaseIE):
 424     IE_NAME = 'pluralsight:course'
 425     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:library/)?courses/(?P<id>[^/]+)'
 426     _TESTS = [{
 427         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
 428         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
 429         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',
 430         'info_dict': {
 431             'id': 'hosting-sql-server-windows-azure-iaas',
 432             'title': 'Hosting SQL Server in Microsoft Azure IaaS Fundamentals',
 433             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
 434         },
 435         'playlist_count': 31,
 436     }, {
 437         # available without pluralsight account
 438         'url': 'https://www.pluralsight.com/courses/angularjs-get-started',
 439         'only_matching': True,
 440     }, {
 441         'url': 'https://app.pluralsight.com/library/courses/understanding-microsoft-azure-amazon-aws/table-of-contents',
 442         'only_matching': True,
 443     }]
 444
 445     def _real_extract(self, url):
 446         course_id = self._match_id(url)
 447
 448         # TODO: PSM cookie
 449
 450         course = self._download_course(course_id, url, course_id)
 451
 452         title = course['title']
 453         course_name = course['name']
 454         course_data = course['modules']
 455         description = course.get('description') or course.get('shortDescription')
 456
 457         entries = []
 458         for num, module in enumerate(course_data, 1):
 459             author = module.get('author')
 460             module_name = module.get('name')
 461             if not author or not module_name:
 462                 continue
 463             for clip in module.get('clips', []):
 464                 clip_index = int_or_none(clip.get('index'))
 465                 if clip_index is None:
 466                     continue
 467                 clip_url = update_url_query(
 468                     f'{self._API_BASE}/player', query={
 469                         'mode': 'live',
 470                         'course': course_name,
 471                         'author': author,
 472                         'name': module_name,
 473                         'clip': clip_index,
 474                     })
 475                 entries.append({
 476                     '_type': 'url_transparent',
 477                     'url': clip_url,
 478                     'ie_key': PluralsightIE.ie_key(),
 479                     'chapter': module.get('title'),
 480                     'chapter_number': num,
 481                     'chapter_id': module.get('moduleRef'),
 482                 })
 483
 484         return self.playlist_result(entries, course_id, title, description)