yt_dlp/extractor/pluralsight.py

   1 import collections
   2 import json
   3 import os
   4 import random
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_str,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     dict_get,
  14     ExtractorError,
  15     float_or_none,
  16     int_or_none,
  17     parse_duration,
  18     parse_qs,
  19     qualities,
  20     srt_subtitles_timecode,
  21     try_get,
  22     update_url_query,
  23     urlencode_postdata,
  24 )
  25
  26
  27 class PluralsightBaseIE(InfoExtractor):
  28     _API_BASE = 'https://app.pluralsight.com'
  29
  30     _GRAPHQL_EP = '%s/player/api/graphql' % _API_BASE
  31     _GRAPHQL_HEADERS = {
  32         'Content-Type': 'application/json;charset=UTF-8',
  33     }
  34     _GRAPHQL_COURSE_TMPL = '''
  35 query BootstrapPlayer {
  36   rpc {
  37     bootstrapPlayer {
  38       profile {
  39         firstName
  40         lastName
  41         email
  42         username
  43         userHandle
  44         authed
  45         isAuthed
  46         plan
  47       }
  48       course(courseId: "%s") {
  49         name
  50         title
  51         courseHasCaptions
  52         translationLanguages {
  53           code
  54           name
  55         }
  56         supportsWideScreenVideoFormats
  57         timestamp
  58         modules {
  59           name
  60           title
  61           duration
  62           formattedDuration
  63           author
  64           authorized
  65           clips {
  66             authorized
  67             clipId
  68             duration
  69             formattedDuration
  70             id
  71             index
  72             moduleIndex
  73             moduleTitle
  74             name
  75             title
  76             watched
  77           }
  78         }
  79       }
  80     }
  81   }
  82 }'''
  83
  84     def _download_course(self, course_id, url, display_id):
  85         try:
  86             return self._download_course_rpc(course_id, url, display_id)
  87         except ExtractorError:
  88             # Old API fallback
  89             return self._download_json(
  90                 'https://app.pluralsight.com/player/user/api/v1/player/payload',
  91                 display_id, data=urlencode_postdata({'courseId': course_id}),
  92                 headers={'Referer': url})
  93
  94     def _download_course_rpc(self, course_id, url, display_id):
  95         response = self._download_json(
  96             self._GRAPHQL_EP, display_id, data=json.dumps({
  97                 'query': self._GRAPHQL_COURSE_TMPL % course_id,
  98                 'variables': {}
  99             }).encode('utf-8'), headers=self._GRAPHQL_HEADERS)
 100
 101         course = try_get(
 102             response, lambda x: x['data']['rpc']['bootstrapPlayer']['course'],
 103             dict)
 104         if course:
 105             return course
 106
 107         raise ExtractorError(
 108             '%s said: %s' % (self.IE_NAME, response['error']['message']),
 109             expected=True)
 110
 111
 112 class PluralsightIE(PluralsightBaseIE):
 113     IE_NAME = 'pluralsight'
 114     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:training/)?player\?'
 115     _LOGIN_URL = 'https://app.pluralsight.com/id/'
 116
 117     _NETRC_MACHINE = 'pluralsight'
 118
 119     _TESTS = [{
 120         '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',
 121         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
 122         'info_dict': {
 123             'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04',
 124             'ext': 'mp4',
 125             'title': 'Demo Monitoring',
 126             'duration': 338,
 127         },
 128         'skip': 'Requires pluralsight account credentials',
 129     }, {
 130         'url': 'https://app.pluralsight.com/training/player?course=angularjs-get-started&author=scott-allen&name=angularjs-get-started-m1-introduction&clip=0&mode=live',
 131         'only_matching': True,
 132     }, {
 133         # available without pluralsight account
 134         'url': 'http://app.pluralsight.com/training/player?author=scott-allen&name=angularjs-get-started-m1-introduction&mode=live&clip=0&course=angularjs-get-started',
 135         'only_matching': True,
 136     }, {
 137         'url': 'https://app.pluralsight.com/player?course=ccna-intro-networking&author=ross-bagurdes&name=ccna-intro-networking-m06&clip=0',
 138         'only_matching': True,
 139     }]
 140
 141     GRAPHQL_VIEWCLIP_TMPL = '''
 142 query viewClip {
 143   viewClip(input: {
 144     author: "%(author)s",
 145     clipIndex: %(clipIndex)d,
 146     courseName: "%(courseName)s",
 147     includeCaptions: %(includeCaptions)s,
 148     locale: "%(locale)s",
 149     mediaType: "%(mediaType)s",
 150     moduleName: "%(moduleName)s",
 151     quality: "%(quality)s"
 152   }) {
 153     urls {
 154       url
 155       cdn
 156       rank
 157       source
 158     },
 159     status
 160   }
 161 }'''
 162
 163     def _perform_login(self, username, password):
 164         login_page = self._download_webpage(
 165             self._LOGIN_URL, None, 'Downloading login page')
 166
 167         login_form = self._hidden_inputs(login_page)
 168
 169         login_form.update({
 170             'Username': username,
 171             'Password': password,
 172         })
 173
 174         post_url = self._search_regex(
 175             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
 176             'post url', default=self._LOGIN_URL, group='url')
 177
 178         if not post_url.startswith('http'):
 179             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
 180
 181         response = self._download_webpage(
 182             post_url, None, 'Logging in',
 183             data=urlencode_postdata(login_form),
 184             headers={'Content-Type': 'application/x-www-form-urlencoded'})
 185
 186         error = self._search_regex(
 187             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',
 188             response, 'error message', default=None)
 189         if error:
 190             raise ExtractorError('Unable to login: %s' % error, expected=True)
 191
 192         if all(not re.search(p, response) for p in (
 193                 r'__INITIAL_STATE__', r'["\']currentUser["\']',
 194                 # new layout?
 195                 r'>\s*Sign out\s*<')):
 196             BLOCKED = 'Your account has been blocked due to suspicious activity'
 197             if BLOCKED in response:
 198                 raise ExtractorError(
 199                     'Unable to login: %s' % BLOCKED, expected=True)
 200             MUST_AGREE = 'To continue using Pluralsight, you must agree to'
 201             if any(p in response for p in (MUST_AGREE, '>Disagree<', '>Agree<')):
 202                 raise ExtractorError(
 203                     'Unable to login: %s some documents. Go to pluralsight.com, '
 204                     'log in and agree with what Pluralsight requires.'
 205                     % MUST_AGREE, expected=True)
 206
 207             raise ExtractorError('Unable to log in')
 208
 209     def _get_subtitles(self, author, clip_idx, clip_id, lang, name, duration, video_id):
 210         captions = None
 211         if clip_id:
 212             captions = self._download_json(
 213                 '%s/transcript/api/v1/caption/json/%s/%s'
 214                 % (self._API_BASE, clip_id, lang), video_id,
 215                 'Downloading captions JSON', 'Unable to download captions JSON',
 216                 fatal=False)
 217         if not captions:
 218             captions_post = {
 219                 'a': author,
 220                 'cn': int(clip_idx),
 221                 'lc': lang,
 222                 'm': name,
 223             }
 224             captions = self._download_json(
 225                 '%s/player/retrieve-captions' % self._API_BASE, video_id,
 226                 'Downloading captions JSON', 'Unable to download captions JSON',
 227                 fatal=False, data=json.dumps(captions_post).encode('utf-8'),
 228                 headers={'Content-Type': 'application/json;charset=utf-8'})
 229         if captions:
 230             return {
 231                 lang: [{
 232                     'ext': 'json',
 233                     'data': json.dumps(captions),
 234                 }, {
 235                     'ext': 'srt',
 236                     'data': self._convert_subtitles(duration, captions),
 237                 }]
 238             }
 239
 240     @staticmethod
 241     def _convert_subtitles(duration, subs):
 242         srt = ''
 243         TIME_OFFSET_KEYS = ('displayTimeOffset', 'DisplayTimeOffset')
 244         TEXT_KEYS = ('text', 'Text')
 245         for num, current in enumerate(subs):
 246             current = subs[num]
 247             start, text = (
 248                 float_or_none(dict_get(current, TIME_OFFSET_KEYS, skip_false_values=False)),
 249                 dict_get(current, TEXT_KEYS))
 250             if start is None or text is None:
 251                 continue
 252             end = duration if num == len(subs) - 1 else float_or_none(
 253                 dict_get(subs[num + 1], TIME_OFFSET_KEYS, skip_false_values=False))
 254             if end is None:
 255                 continue
 256             srt += os.linesep.join(
 257                 (
 258                     '%d' % num,
 259                     '%s --> %s' % (
 260                         srt_subtitles_timecode(start),
 261                         srt_subtitles_timecode(end)),
 262                     text,
 263                     os.linesep,
 264                 ))
 265         return srt
 266
 267     def _real_extract(self, url):
 268         qs = parse_qs(url)
 269
 270         author = qs.get('author', [None])[0]
 271         name = qs.get('name', [None])[0]
 272         clip_idx = qs.get('clip', [None])[0]
 273         course_name = qs.get('course', [None])[0]
 274
 275         if any(not f for f in (author, name, clip_idx, course_name,)):
 276             raise ExtractorError('Invalid URL', expected=True)
 277
 278         display_id = '%s-%s' % (name, clip_idx)
 279
 280         course = self._download_course(course_name, url, display_id)
 281
 282         collection = course['modules']
 283
 284         clip = None
 285
 286         for module_ in collection:
 287             if name in (module_.get('moduleName'), module_.get('name')):
 288                 for clip_ in module_.get('clips', []):
 289                     clip_index = clip_.get('clipIndex')
 290                     if clip_index is None:
 291                         clip_index = clip_.get('index')
 292                     if clip_index is None:
 293                         continue
 294                     if compat_str(clip_index) == clip_idx:
 295                         clip = clip_
 296                         break
 297
 298         if not clip:
 299             raise ExtractorError('Unable to resolve clip')
 300
 301         title = clip['title']
 302         clip_id = clip.get('clipName') or clip.get('name') or clip['clipId']
 303
 304         QUALITIES = {
 305             'low': {'width': 640, 'height': 480},
 306             'medium': {'width': 848, 'height': 640},
 307             'high': {'width': 1024, 'height': 768},
 308             'high-widescreen': {'width': 1280, 'height': 720},
 309         }
 310
 311         QUALITIES_PREFERENCE = ('low', 'medium', 'high', 'high-widescreen',)
 312         quality_key = qualities(QUALITIES_PREFERENCE)
 313
 314         AllowedQuality = collections.namedtuple('AllowedQuality', ['ext', 'qualities'])
 315
 316         ALLOWED_QUALITIES = (
 317             AllowedQuality('webm', ['high', ]),
 318             AllowedQuality('mp4', ['low', 'medium', 'high', ]),
 319         )
 320
 321         # Some courses also offer widescreen resolution for high quality (see
 322         # https://github.com/ytdl-org/youtube-dl/issues/7766)
 323         widescreen = course.get('supportsWideScreenVideoFormats') is True
 324         best_quality = 'high-widescreen' if widescreen else 'high'
 325         if widescreen:
 326             for allowed_quality in ALLOWED_QUALITIES:
 327                 allowed_quality.qualities.append(best_quality)
 328
 329         # In order to minimize the number of calls to ViewClip API and reduce
 330         # the probability of being throttled or banned by Pluralsight we will request
 331         # only single format until formats listing was explicitly requested.
 332         if self.get_param('listformats', False):
 333             allowed_qualities = ALLOWED_QUALITIES
 334         else:
 335             def guess_allowed_qualities():
 336                 req_format = self.get_param('format') or 'best'
 337                 req_format_split = req_format.split('-', 1)
 338                 if len(req_format_split) > 1:
 339                     req_ext, req_quality = req_format_split
 340                     req_quality = '-'.join(req_quality.split('-')[:2])
 341                     for allowed_quality in ALLOWED_QUALITIES:
 342                         if req_ext == allowed_quality.ext and req_quality in allowed_quality.qualities:
 343                             return (AllowedQuality(req_ext, (req_quality, )), )
 344                 req_ext = 'webm' if self.get_param('prefer_free_formats') else 'mp4'
 345                 return (AllowedQuality(req_ext, (best_quality, )), )
 346             allowed_qualities = guess_allowed_qualities()
 347
 348         formats = []
 349         for ext, qualities_ in allowed_qualities:
 350             for quality in qualities_:
 351                 f = QUALITIES[quality].copy()
 352                 clip_post = {
 353                     'author': author,
 354                     'includeCaptions': 'false',
 355                     'clipIndex': int(clip_idx),
 356                     'courseName': course_name,
 357                     'locale': 'en',
 358                     'moduleName': name,
 359                     'mediaType': ext,
 360                     'quality': '%dx%d' % (f['width'], f['height']),
 361                 }
 362                 format_id = '%s-%s' % (ext, quality)
 363
 364                 try:
 365                     viewclip = self._download_json(
 366                         self._GRAPHQL_EP, display_id,
 367                         'Downloading %s viewclip graphql' % format_id,
 368                         data=json.dumps({
 369                             'query': self.GRAPHQL_VIEWCLIP_TMPL % clip_post,
 370                             'variables': {}
 371                         }).encode('utf-8'),
 372                         headers=self._GRAPHQL_HEADERS)['data']['viewClip']
 373                 except ExtractorError:
 374                     # Still works but most likely will go soon
 375                     viewclip = self._download_json(
 376                         '%s/video/clips/viewclip' % self._API_BASE, display_id,
 377                         'Downloading %s viewclip JSON' % format_id, fatal=False,
 378                         data=json.dumps(clip_post).encode('utf-8'),
 379                         headers={'Content-Type': 'application/json;charset=utf-8'})
 380
 381                 # Pluralsight tracks multiple sequential calls to ViewClip API and start
 382                 # to return 429 HTTP errors after some time (see
 383                 # https://github.com/ytdl-org/youtube-dl/pull/6989). Moreover it may even lead
 384                 # to account ban (see https://github.com/ytdl-org/youtube-dl/issues/6842).
 385                 # To somewhat reduce the probability of these consequences
 386                 # we will sleep random amount of time before each call to ViewClip.
 387                 self._sleep(
 388                     random.randint(5, 10), display_id,
 389                     '%(video_id)s: Waiting for %(timeout)s seconds to avoid throttling')
 390
 391                 if not viewclip:
 392                     continue
 393
 394                 clip_urls = viewclip.get('urls')
 395                 if not isinstance(clip_urls, list):
 396                     continue
 397
 398                 for clip_url_data in clip_urls:
 399                     clip_url = clip_url_data.get('url')
 400                     if not clip_url:
 401                         continue
 402                     cdn = clip_url_data.get('cdn')
 403                     clip_f = f.copy()
 404                     clip_f.update({
 405                         'url': clip_url,
 406                         'ext': ext,
 407                         'format_id': '%s-%s' % (format_id, cdn) if cdn else format_id,
 408                         'quality': quality_key(quality),
 409                         'source_preference': int_or_none(clip_url_data.get('rank')),
 410                     })
 411                     formats.append(clip_f)
 412
 413         duration = int_or_none(
 414             clip.get('duration')) or parse_duration(clip.get('formattedDuration'))
 415
 416         # TODO: other languages?
 417         subtitles = self.extract_subtitles(
 418             author, clip_idx, clip.get('clipId'), 'en', name, duration, display_id)
 419
 420         return {
 421             'id': clip_id,
 422             'title': title,
 423             'duration': duration,
 424             'creator': author,
 425             'formats': formats,
 426             'subtitles': subtitles,
 427         }
 428
 429
 430 class PluralsightCourseIE(PluralsightBaseIE):
 431     IE_NAME = 'pluralsight:course'
 432     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:library/)?courses/(?P<id>[^/]+)'
 433     _TESTS = [{
 434         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
 435         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
 436         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',
 437         'info_dict': {
 438             'id': 'hosting-sql-server-windows-azure-iaas',
 439             'title': 'Hosting SQL Server in Microsoft Azure IaaS Fundamentals',
 440             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
 441         },
 442         'playlist_count': 31,
 443     }, {
 444         # available without pluralsight account
 445         'url': 'https://www.pluralsight.com/courses/angularjs-get-started',
 446         'only_matching': True,
 447     }, {
 448         'url': 'https://app.pluralsight.com/library/courses/understanding-microsoft-azure-amazon-aws/table-of-contents',
 449         'only_matching': True,
 450     }]
 451
 452     def _real_extract(self, url):
 453         course_id = self._match_id(url)
 454
 455         # TODO: PSM cookie
 456
 457         course = self._download_course(course_id, url, course_id)
 458
 459         title = course['title']
 460         course_name = course['name']
 461         course_data = course['modules']
 462         description = course.get('description') or course.get('shortDescription')
 463
 464         entries = []
 465         for num, module in enumerate(course_data, 1):
 466             author = module.get('author')
 467             module_name = module.get('name')
 468             if not author or not module_name:
 469                 continue
 470             for clip in module.get('clips', []):
 471                 clip_index = int_or_none(clip.get('index'))
 472                 if clip_index is None:
 473                     continue
 474                 clip_url = update_url_query(
 475                     '%s/player' % self._API_BASE, query={
 476                         'mode': 'live',
 477                         'course': course_name,
 478                         'author': author,
 479                         'name': module_name,
 480                         'clip': clip_index,
 481                     })
 482                 entries.append({
 483                     '_type': 'url_transparent',
 484                     'url': clip_url,
 485                     'ie_key': PluralsightIE.ie_key(),
 486                     'chapter': module.get('title'),
 487                     'chapter_number': num,
 488                     'chapter_id': module.get('moduleRef'),
 489                 })
 490
 491         return self.playlist_result(entries, course_id, title, description)