yt_dlp/extractor/pluralsight.py

   1 from __future__ import unicode_literals
   2
   3 import collections
   4 import json
   5 import os
   6 import random
   7 import re
   8
   9 from .common import InfoExtractor
  10 from ..compat import (
  11     compat_str,
  12     compat_urlparse,
  13 )
  14 from ..utils import (
  15     dict_get,
  16     ExtractorError,
  17     float_or_none,
  18     int_or_none,
  19     parse_duration,
  20     parse_qs,
  21     qualities,
  22     srt_subtitles_timecode,
  23     try_get,
  24     update_url_query,
  25     urlencode_postdata,
  26 )
  27
  28
  29 class PluralsightBaseIE(InfoExtractor):
  30     _API_BASE = 'https://app.pluralsight.com'
  31
  32     _GRAPHQL_EP = '%s/player/api/graphql' % _API_BASE
  33     _GRAPHQL_HEADERS = {
  34         'Content-Type': 'application/json;charset=UTF-8',
  35     }
  36     _GRAPHQL_COURSE_TMPL = '''
  37 query BootstrapPlayer {
  38   rpc {
  39     bootstrapPlayer {
  40       profile {
  41         firstName
  42         lastName
  43         email
  44         username
  45         userHandle
  46         authed
  47         isAuthed
  48         plan
  49       }
  50       course(courseId: "%s") {
  51         name
  52         title
  53         courseHasCaptions
  54         translationLanguages {
  55           code
  56           name
  57         }
  58         supportsWideScreenVideoFormats
  59         timestamp
  60         modules {
  61           name
  62           title
  63           duration
  64           formattedDuration
  65           author
  66           authorized
  67           clips {
  68             authorized
  69             clipId
  70             duration
  71             formattedDuration
  72             id
  73             index
  74             moduleIndex
  75             moduleTitle
  76             name
  77             title
  78             watched
  79           }
  80         }
  81       }
  82     }
  83   }
  84 }'''
  85
  86     def _download_course(self, course_id, url, display_id):
  87         try:
  88             return self._download_course_rpc(course_id, url, display_id)
  89         except ExtractorError:
  90             # Old API fallback
  91             return self._download_json(
  92                 'https://app.pluralsight.com/player/user/api/v1/player/payload',
  93                 display_id, data=urlencode_postdata({'courseId': course_id}),
  94                 headers={'Referer': url})
  95
  96     def _download_course_rpc(self, course_id, url, display_id):
  97         response = self._download_json(
  98             self._GRAPHQL_EP, display_id, data=json.dumps({
  99                 'query': self._GRAPHQL_COURSE_TMPL % course_id,
 100                 'variables': {}
 101             }).encode('utf-8'), headers=self._GRAPHQL_HEADERS)
 102
 103         course = try_get(
 104             response, lambda x: x['data']['rpc']['bootstrapPlayer']['course'],
 105             dict)
 106         if course:
 107             return course
 108
 109         raise ExtractorError(
 110             '%s said: %s' % (self.IE_NAME, response['error']['message']),
 111             expected=True)
 112
 113
 114 class PluralsightIE(PluralsightBaseIE):
 115     IE_NAME = 'pluralsight'
 116     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:training/)?player\?'
 117     _LOGIN_URL = 'https://app.pluralsight.com/id/'
 118
 119     _NETRC_MACHINE = 'pluralsight'
 120
 121     _TESTS = [{
 122         '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',
 123         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
 124         'info_dict': {
 125             'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04',
 126             'ext': 'mp4',
 127             'title': 'Demo Monitoring',
 128             'duration': 338,
 129         },
 130         'skip': 'Requires pluralsight account credentials',
 131     }, {
 132         'url': 'https://app.pluralsight.com/training/player?course=angularjs-get-started&author=scott-allen&name=angularjs-get-started-m1-introduction&clip=0&mode=live',
 133         'only_matching': True,
 134     }, {
 135         # available without pluralsight account
 136         'url': 'http://app.pluralsight.com/training/player?author=scott-allen&name=angularjs-get-started-m1-introduction&mode=live&clip=0&course=angularjs-get-started',
 137         'only_matching': True,
 138     }, {
 139         'url': 'https://app.pluralsight.com/player?course=ccna-intro-networking&author=ross-bagurdes&name=ccna-intro-networking-m06&clip=0',
 140         'only_matching': True,
 141     }]
 142
 143     GRAPHQL_VIEWCLIP_TMPL = '''
 144 query viewClip {
 145   viewClip(input: {
 146     author: "%(author)s",
 147     clipIndex: %(clipIndex)d,
 148     courseName: "%(courseName)s",
 149     includeCaptions: %(includeCaptions)s,
 150     locale: "%(locale)s",
 151     mediaType: "%(mediaType)s",
 152     moduleName: "%(moduleName)s",
 153     quality: "%(quality)s"
 154   }) {
 155     urls {
 156       url
 157       cdn
 158       rank
 159       source
 160     },
 161     status
 162   }
 163 }'''
 164
 165     def _real_initialize(self):
 166         self._login()
 167
 168     def _login(self):
 169         username, password = self._get_login_info()
 170         if username is None:
 171             return
 172
 173         login_page = self._download_webpage(
 174             self._LOGIN_URL, None, 'Downloading login page')
 175
 176         login_form = self._hidden_inputs(login_page)
 177
 178         login_form.update({
 179             'Username': username,
 180             'Password': password,
 181         })
 182
 183         post_url = self._search_regex(
 184             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
 185             'post url', default=self._LOGIN_URL, group='url')
 186
 187         if not post_url.startswith('http'):
 188             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
 189
 190         response = self._download_webpage(
 191             post_url, None, 'Logging in',
 192             data=urlencode_postdata(login_form),
 193             headers={'Content-Type': 'application/x-www-form-urlencoded'})
 194
 195         error = self._search_regex(
 196             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',
 197             response, 'error message', default=None)
 198         if error:
 199             raise ExtractorError('Unable to login: %s' % error, expected=True)
 200
 201         if all(not re.search(p, response) for p in (
 202                 r'__INITIAL_STATE__', r'["\']currentUser["\']',
 203                 # new layout?
 204                 r'>\s*Sign out\s*<')):
 205             BLOCKED = 'Your account has been blocked due to suspicious activity'
 206             if BLOCKED in response:
 207                 raise ExtractorError(
 208                     'Unable to login: %s' % BLOCKED, expected=True)
 209             MUST_AGREE = 'To continue using Pluralsight, you must agree to'
 210             if any(p in response for p in (MUST_AGREE, '>Disagree<', '>Agree<')):
 211                 raise ExtractorError(
 212                     'Unable to login: %s some documents. Go to pluralsight.com, '
 213                     'log in and agree with what Pluralsight requires.'
 214                     % MUST_AGREE, expected=True)
 215
 216             raise ExtractorError('Unable to log in')
 217
 218     def _get_subtitles(self, author, clip_idx, clip_id, lang, name, duration, video_id):
 219         captions = None
 220         if clip_id:
 221             captions = self._download_json(
 222                 '%s/transcript/api/v1/caption/json/%s/%s'
 223                 % (self._API_BASE, clip_id, lang), video_id,
 224                 'Downloading captions JSON', 'Unable to download captions JSON',
 225                 fatal=False)
 226         if not captions:
 227             captions_post = {
 228                 'a': author,
 229                 'cn': int(clip_idx),
 230                 'lc': lang,
 231                 'm': name,
 232             }
 233             captions = self._download_json(
 234                 '%s/player/retrieve-captions' % self._API_BASE, video_id,
 235                 'Downloading captions JSON', 'Unable to download captions JSON',
 236                 fatal=False, data=json.dumps(captions_post).encode('utf-8'),
 237                 headers={'Content-Type': 'application/json;charset=utf-8'})
 238         if captions:
 239             return {
 240                 lang: [{
 241                     'ext': 'json',
 242                     'data': json.dumps(captions),
 243                 }, {
 244                     'ext': 'srt',
 245                     'data': self._convert_subtitles(duration, captions),
 246                 }]
 247             }
 248
 249     @staticmethod
 250     def _convert_subtitles(duration, subs):
 251         srt = ''
 252         TIME_OFFSET_KEYS = ('displayTimeOffset', 'DisplayTimeOffset')
 253         TEXT_KEYS = ('text', 'Text')
 254         for num, current in enumerate(subs):
 255             current = subs[num]
 256             start, text = (
 257                 float_or_none(dict_get(current, TIME_OFFSET_KEYS, skip_false_values=False)),
 258                 dict_get(current, TEXT_KEYS))
 259             if start is None or text is None:
 260                 continue
 261             end = duration if num == len(subs) - 1 else float_or_none(
 262                 dict_get(subs[num + 1], TIME_OFFSET_KEYS, skip_false_values=False))
 263             if end is None:
 264                 continue
 265             srt += os.linesep.join(
 266                 (
 267                     '%d' % num,
 268                     '%s --> %s' % (
 269                         srt_subtitles_timecode(start),
 270                         srt_subtitles_timecode(end)),
 271                     text,
 272                     os.linesep,
 273                 ))
 274         return srt
 275
 276     def _real_extract(self, url):
 277         qs = parse_qs(url)
 278
 279         author = qs.get('author', [None])[0]
 280         name = qs.get('name', [None])[0]
 281         clip_idx = qs.get('clip', [None])[0]
 282         course_name = qs.get('course', [None])[0]
 283
 284         if any(not f for f in (author, name, clip_idx, course_name,)):
 285             raise ExtractorError('Invalid URL', expected=True)
 286
 287         display_id = '%s-%s' % (name, clip_idx)
 288
 289         course = self._download_course(course_name, url, display_id)
 290
 291         collection = course['modules']
 292
 293         clip = None
 294
 295         for module_ in collection:
 296             if name in (module_.get('moduleName'), module_.get('name')):
 297                 for clip_ in module_.get('clips', []):
 298                     clip_index = clip_.get('clipIndex')
 299                     if clip_index is None:
 300                         clip_index = clip_.get('index')
 301                     if clip_index is None:
 302                         continue
 303                     if compat_str(clip_index) == clip_idx:
 304                         clip = clip_
 305                         break
 306
 307         if not clip:
 308             raise ExtractorError('Unable to resolve clip')
 309
 310         title = clip['title']
 311         clip_id = clip.get('clipName') or clip.get('name') or clip['clipId']
 312
 313         QUALITIES = {
 314             'low': {'width': 640, 'height': 480},
 315             'medium': {'width': 848, 'height': 640},
 316             'high': {'width': 1024, 'height': 768},
 317             'high-widescreen': {'width': 1280, 'height': 720},
 318         }
 319
 320         QUALITIES_PREFERENCE = ('low', 'medium', 'high', 'high-widescreen',)
 321         quality_key = qualities(QUALITIES_PREFERENCE)
 322
 323         AllowedQuality = collections.namedtuple('AllowedQuality', ['ext', 'qualities'])
 324
 325         ALLOWED_QUALITIES = (
 326             AllowedQuality('webm', ['high', ]),
 327             AllowedQuality('mp4', ['low', 'medium', 'high', ]),
 328         )
 329
 330         # Some courses also offer widescreen resolution for high quality (see
 331         # https://github.com/ytdl-org/youtube-dl/issues/7766)
 332         widescreen = course.get('supportsWideScreenVideoFormats') is True
 333         best_quality = 'high-widescreen' if widescreen else 'high'
 334         if widescreen:
 335             for allowed_quality in ALLOWED_QUALITIES:
 336                 allowed_quality.qualities.append(best_quality)
 337
 338         # In order to minimize the number of calls to ViewClip API and reduce
 339         # the probability of being throttled or banned by Pluralsight we will request
 340         # only single format until formats listing was explicitly requested.
 341         if self.get_param('listformats', False):
 342             allowed_qualities = ALLOWED_QUALITIES
 343         else:
 344             def guess_allowed_qualities():
 345                 req_format = self.get_param('format') or 'best'
 346                 req_format_split = req_format.split('-', 1)
 347                 if len(req_format_split) > 1:
 348                     req_ext, req_quality = req_format_split
 349                     req_quality = '-'.join(req_quality.split('-')[:2])
 350                     for allowed_quality in ALLOWED_QUALITIES:
 351                         if req_ext == allowed_quality.ext and req_quality in allowed_quality.qualities:
 352                             return (AllowedQuality(req_ext, (req_quality, )), )
 353                 req_ext = 'webm' if self.get_param('prefer_free_formats') else 'mp4'
 354                 return (AllowedQuality(req_ext, (best_quality, )), )
 355             allowed_qualities = guess_allowed_qualities()
 356
 357         formats = []
 358         for ext, qualities_ in allowed_qualities:
 359             for quality in qualities_:
 360                 f = QUALITIES[quality].copy()
 361                 clip_post = {
 362                     'author': author,
 363                     'includeCaptions': 'false',
 364                     'clipIndex': int(clip_idx),
 365                     'courseName': course_name,
 366                     'locale': 'en',
 367                     'moduleName': name,
 368                     'mediaType': ext,
 369                     'quality': '%dx%d' % (f['width'], f['height']),
 370                 }
 371                 format_id = '%s-%s' % (ext, quality)
 372
 373                 try:
 374                     viewclip = self._download_json(
 375                         self._GRAPHQL_EP, display_id,
 376                         'Downloading %s viewclip graphql' % format_id,
 377                         data=json.dumps({
 378                             'query': self.GRAPHQL_VIEWCLIP_TMPL % clip_post,
 379                             'variables': {}
 380                         }).encode('utf-8'),
 381                         headers=self._GRAPHQL_HEADERS)['data']['viewClip']
 382                 except ExtractorError:
 383                     # Still works but most likely will go soon
 384                     viewclip = self._download_json(
 385                         '%s/video/clips/viewclip' % self._API_BASE, display_id,
 386                         'Downloading %s viewclip JSON' % format_id, fatal=False,
 387                         data=json.dumps(clip_post).encode('utf-8'),
 388                         headers={'Content-Type': 'application/json;charset=utf-8'})
 389
 390                 # Pluralsight tracks multiple sequential calls to ViewClip API and start
 391                 # to return 429 HTTP errors after some time (see
 392                 # https://github.com/ytdl-org/youtube-dl/pull/6989). Moreover it may even lead
 393                 # to account ban (see https://github.com/ytdl-org/youtube-dl/issues/6842).
 394                 # To somewhat reduce the probability of these consequences
 395                 # we will sleep random amount of time before each call to ViewClip.
 396                 self._sleep(
 397                     random.randint(5, 10), display_id,
 398                     '%(video_id)s: Waiting for %(timeout)s seconds to avoid throttling')
 399
 400                 if not viewclip:
 401                     continue
 402
 403                 clip_urls = viewclip.get('urls')
 404                 if not isinstance(clip_urls, list):
 405                     continue
 406
 407                 for clip_url_data in clip_urls:
 408                     clip_url = clip_url_data.get('url')
 409                     if not clip_url:
 410                         continue
 411                     cdn = clip_url_data.get('cdn')
 412                     clip_f = f.copy()
 413                     clip_f.update({
 414                         'url': clip_url,
 415                         'ext': ext,
 416                         'format_id': '%s-%s' % (format_id, cdn) if cdn else format_id,
 417                         'quality': quality_key(quality),
 418                         'source_preference': int_or_none(clip_url_data.get('rank')),
 419                     })
 420                     formats.append(clip_f)
 421
 422         self._sort_formats(formats)
 423
 424         duration = int_or_none(
 425             clip.get('duration')) or parse_duration(clip.get('formattedDuration'))
 426
 427         # TODO: other languages?
 428         subtitles = self.extract_subtitles(
 429             author, clip_idx, clip.get('clipId'), 'en', name, duration, display_id)
 430
 431         return {
 432             'id': clip_id,
 433             'title': title,
 434             'duration': duration,
 435             'creator': author,
 436             'formats': formats,
 437             'subtitles': subtitles,
 438         }
 439
 440
 441 class PluralsightCourseIE(PluralsightBaseIE):
 442     IE_NAME = 'pluralsight:course'
 443     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:library/)?courses/(?P<id>[^/]+)'
 444     _TESTS = [{
 445         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
 446         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
 447         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',
 448         'info_dict': {
 449             'id': 'hosting-sql-server-windows-azure-iaas',
 450             'title': 'Hosting SQL Server in Microsoft Azure IaaS Fundamentals',
 451             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
 452         },
 453         'playlist_count': 31,
 454     }, {
 455         # available without pluralsight account
 456         'url': 'https://www.pluralsight.com/courses/angularjs-get-started',
 457         'only_matching': True,
 458     }, {
 459         'url': 'https://app.pluralsight.com/library/courses/understanding-microsoft-azure-amazon-aws/table-of-contents',
 460         'only_matching': True,
 461     }]
 462
 463     def _real_extract(self, url):
 464         course_id = self._match_id(url)
 465
 466         # TODO: PSM cookie
 467
 468         course = self._download_course(course_id, url, course_id)
 469
 470         title = course['title']
 471         course_name = course['name']
 472         course_data = course['modules']
 473         description = course.get('description') or course.get('shortDescription')
 474
 475         entries = []
 476         for num, module in enumerate(course_data, 1):
 477             author = module.get('author')
 478             module_name = module.get('name')
 479             if not author or not module_name:
 480                 continue
 481             for clip in module.get('clips', []):
 482                 clip_index = int_or_none(clip.get('index'))
 483                 if clip_index is None:
 484                     continue
 485                 clip_url = update_url_query(
 486                     '%s/player' % self._API_BASE, query={
 487                         'mode': 'live',
 488                         'course': course_name,
 489                         'author': author,
 490                         'name': module_name,
 491                         'clip': clip_index,
 492                     })
 493                 entries.append({
 494                     '_type': 'url_transparent',
 495                     'url': clip_url,
 496                     'ie_key': PluralsightIE.ie_key(),
 497                     'chapter': module.get('title'),
 498                     'chapter_number': num,
 499                     'chapter_id': module.get('moduleRef'),
 500                 })
 501
 502         return self.playlist_result(entries, course_id, title, description)