youtube_dl/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     qualities,
  21     srt_subtitles_timecode,
  22     try_get,
  23     update_url_query,
  24     urlencode_postdata,
  25 )
  26
  27
  28 class PluralsightBaseIE(InfoExtractor):
  29     _API_BASE = 'https://app.pluralsight.com'
  30
  31     _GRAPHQL_EP = '%s/player/api/graphql' % _API_BASE
  32     _GRAPHQL_HEADERS = {
  33         'Content-Type': 'application/json;charset=UTF-8',
  34     }
  35     _GRAPHQL_COURSE_TMPL = '''
  36 query BootstrapPlayer {
  37   rpc {
  38     bootstrapPlayer {
  39       profile {
  40         firstName
  41         lastName
  42         email
  43         username
  44         userHandle
  45         authed
  46         isAuthed
  47         plan
  48       }
  49       course(courseId: "%s") {
  50         name
  51         title
  52         courseHasCaptions
  53         translationLanguages {
  54           code
  55           name
  56         }
  57         supportsWideScreenVideoFormats
  58         timestamp
  59         modules {
  60           name
  61           title
  62           duration
  63           formattedDuration
  64           author
  65           authorized
  66           clips {
  67             authorized
  68             clipId
  69             duration
  70             formattedDuration
  71             id
  72             index
  73             moduleIndex
  74             moduleTitle
  75             name
  76             title
  77             watched
  78           }
  79         }
  80       }
  81     }
  82   }
  83 }'''
  84
  85     def _download_course(self, course_id, url, display_id):
  86         try:
  87             return self._download_course_rpc(course_id, url, display_id)
  88         except ExtractorError:
  89             # Old API fallback
  90             return self._download_json(
  91                 'https://app.pluralsight.com/player/user/api/v1/player/payload',
  92                 display_id, data=urlencode_postdata({'courseId': course_id}),
  93                 headers={'Referer': url})
  94
  95     def _download_course_rpc(self, course_id, url, display_id):
  96         response = self._download_json(
  97             self._GRAPHQL_EP, display_id, data=json.dumps({
  98                 'query': self._GRAPHQL_COURSE_TMPL % course_id,
  99                 'variables': {}
 100             }).encode('utf-8'), headers=self._GRAPHQL_HEADERS)
 101
 102         course = try_get(
 103             response, lambda x: x['data']['rpc']['bootstrapPlayer']['course'],
 104             dict)
 105         if course:
 106             return course
 107
 108         raise ExtractorError(
 109             '%s said: %s' % (self.IE_NAME, response['error']['message']),
 110             expected=True)
 111
 112
 113 class PluralsightIE(PluralsightBaseIE):
 114     IE_NAME = 'pluralsight'
 115     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:training/)?player\?'
 116     _LOGIN_URL = 'https://app.pluralsight.com/id/'
 117
 118     _NETRC_MACHINE = 'pluralsight'
 119
 120     _TESTS = [{
 121         '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',
 122         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
 123         'info_dict': {
 124             'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04',
 125             'ext': 'mp4',
 126             'title': 'Demo Monitoring',
 127             'duration': 338,
 128         },
 129         'skip': 'Requires pluralsight account credentials',
 130     }, {
 131         'url': 'https://app.pluralsight.com/training/player?course=angularjs-get-started&author=scott-allen&name=angularjs-get-started-m1-introduction&clip=0&mode=live',
 132         'only_matching': True,
 133     }, {
 134         # available without pluralsight account
 135         'url': 'http://app.pluralsight.com/training/player?author=scott-allen&name=angularjs-get-started-m1-introduction&mode=live&clip=0&course=angularjs-get-started',
 136         'only_matching': True,
 137     }, {
 138         'url': 'https://app.pluralsight.com/player?course=ccna-intro-networking&author=ross-bagurdes&name=ccna-intro-networking-m06&clip=0',
 139         'only_matching': True,
 140     }]
 141
 142     GRAPHQL_VIEWCLIP_TMPL = '''
 143 query viewClip {
 144   viewClip(input: {
 145     author: "%(author)s",
 146     clipIndex: %(clipIndex)d,
 147     courseName: "%(courseName)s",
 148     includeCaptions: %(includeCaptions)s,
 149     locale: "%(locale)s",
 150     mediaType: "%(mediaType)s",
 151     moduleName: "%(moduleName)s",
 152     quality: "%(quality)s"
 153   }) {
 154     urls {
 155       url
 156       cdn
 157       rank
 158       source
 159     },
 160     status
 161   }
 162 }'''
 163
 164     def _real_initialize(self):
 165         self._login()
 166
 167     def _login(self):
 168         username, password = self._get_login_info()
 169         if username is None:
 170             return
 171
 172         login_page = self._download_webpage(
 173             self._LOGIN_URL, None, 'Downloading login page')
 174
 175         login_form = self._hidden_inputs(login_page)
 176
 177         login_form.update({
 178             'Username': username,
 179             'Password': password,
 180         })
 181
 182         post_url = self._search_regex(
 183             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
 184             'post url', default=self._LOGIN_URL, group='url')
 185
 186         if not post_url.startswith('http'):
 187             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
 188
 189         response = self._download_webpage(
 190             post_url, None, 'Logging in',
 191             data=urlencode_postdata(login_form),
 192             headers={'Content-Type': 'application/x-www-form-urlencoded'})
 193
 194         error = self._search_regex(
 195             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',
 196             response, 'error message', default=None)
 197         if error:
 198             raise ExtractorError('Unable to login: %s' % error, expected=True)
 199
 200         if all(not re.search(p, response) for p in (
 201                 r'__INITIAL_STATE__', r'["\']currentUser["\']',
 202                 # new layout?
 203                 r'>\s*Sign out\s*<')):
 204             BLOCKED = 'Your account has been blocked due to suspicious activity'
 205             if BLOCKED in response:
 206                 raise ExtractorError(
 207                     'Unable to login: %s' % BLOCKED, expected=True)
 208             MUST_AGREE = 'To continue using Pluralsight, you must agree to'
 209             if any(p in response for p in (MUST_AGREE, '>Disagree<', '>Agree<')):
 210                 raise ExtractorError(
 211                     'Unable to login: %s some documents. Go to pluralsight.com, '
 212                     'log in and agree with what Pluralsight requires.'
 213                     % MUST_AGREE, expected=True)
 214
 215             raise ExtractorError('Unable to log in')
 216
 217     def _get_subtitles(self, author, clip_idx, clip_id, lang, name, duration, video_id):
 218         captions = None
 219         if clip_id:
 220             captions = self._download_json(
 221                 '%s/transcript/api/v1/caption/json/%s/%s'
 222                 % (self._API_BASE, clip_id, lang), video_id,
 223                 'Downloading captions JSON', 'Unable to download captions JSON',
 224                 fatal=False)
 225         if not captions:
 226             captions_post = {
 227                 'a': author,
 228                 'cn': int(clip_idx),
 229                 'lc': lang,
 230                 'm': name,
 231             }
 232             captions = self._download_json(
 233                 '%s/player/retrieve-captions' % self._API_BASE, video_id,
 234                 'Downloading captions JSON', 'Unable to download captions JSON',
 235                 fatal=False, data=json.dumps(captions_post).encode('utf-8'),
 236                 headers={'Content-Type': 'application/json;charset=utf-8'})
 237         if captions:
 238             return {
 239                 lang: [{
 240                     'ext': 'json',
 241                     'data': json.dumps(captions),
 242                 }, {
 243                     'ext': 'srt',
 244                     'data': self._convert_subtitles(duration, captions),
 245                 }]
 246             }
 247
 248     @staticmethod
 249     def _convert_subtitles(duration, subs):
 250         srt = ''
 251         TIME_OFFSET_KEYS = ('displayTimeOffset', 'DisplayTimeOffset')
 252         TEXT_KEYS = ('text', 'Text')
 253         for num, current in enumerate(subs):
 254             current = subs[num]
 255             start, text = (
 256                 float_or_none(dict_get(current, TIME_OFFSET_KEYS, skip_false_values=False)),
 257                 dict_get(current, TEXT_KEYS))
 258             if start is None or text is None:
 259                 continue
 260             end = duration if num == len(subs) - 1 else float_or_none(
 261                 dict_get(subs[num + 1], TIME_OFFSET_KEYS, skip_false_values=False))
 262             if end is None:
 263                 continue
 264             srt += os.linesep.join(
 265                 (
 266                     '%d' % num,
 267                     '%s --> %s' % (
 268                         srt_subtitles_timecode(start),
 269                         srt_subtitles_timecode(end)),
 270                     text,
 271                     os.linesep,
 272                 ))
 273         return srt
 274
 275     def _real_extract(self, url):
 276         qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 277
 278         author = qs.get('author', [None])[0]
 279         name = qs.get('name', [None])[0]
 280         clip_idx = qs.get('clip', [None])[0]
 281         course_name = qs.get('course', [None])[0]
 282
 283         if any(not f for f in (author, name, clip_idx, course_name,)):
 284             raise ExtractorError('Invalid URL', expected=True)
 285
 286         display_id = '%s-%s' % (name, clip_idx)
 287
 288         course = self._download_course(course_name, url, display_id)
 289
 290         collection = course['modules']
 291
 292         clip = None
 293
 294         for module_ in collection:
 295             if name in (module_.get('moduleName'), module_.get('name')):
 296                 for clip_ in module_.get('clips', []):
 297                     clip_index = clip_.get('clipIndex')
 298                     if clip_index is None:
 299                         clip_index = clip_.get('index')
 300                     if clip_index is None:
 301                         continue
 302                     if compat_str(clip_index) == clip_idx:
 303                         clip = clip_
 304                         break
 305
 306         if not clip:
 307             raise ExtractorError('Unable to resolve clip')
 308
 309         title = clip['title']
 310         clip_id = clip.get('clipName') or clip.get('name') or clip['clipId']
 311
 312         QUALITIES = {
 313             'low': {'width': 640, 'height': 480},
 314             'medium': {'width': 848, 'height': 640},
 315             'high': {'width': 1024, 'height': 768},
 316             'high-widescreen': {'width': 1280, 'height': 720},
 317         }
 318
 319         QUALITIES_PREFERENCE = ('low', 'medium', 'high', 'high-widescreen',)
 320         quality_key = qualities(QUALITIES_PREFERENCE)
 321
 322         AllowedQuality = collections.namedtuple('AllowedQuality', ['ext', 'qualities'])
 323
 324         ALLOWED_QUALITIES = (
 325             AllowedQuality('webm', ['high', ]),
 326             AllowedQuality('mp4', ['low', 'medium', 'high', ]),
 327         )
 328
 329         # Some courses also offer widescreen resolution for high quality (see
 330         # https://github.com/rg3/youtube-dl/issues/7766)
 331         widescreen = course.get('supportsWideScreenVideoFormats') is True
 332         best_quality = 'high-widescreen' if widescreen else 'high'
 333         if widescreen:
 334             for allowed_quality in ALLOWED_QUALITIES:
 335                 allowed_quality.qualities.append(best_quality)
 336
 337         # In order to minimize the number of calls to ViewClip API and reduce
 338         # the probability of being throttled or banned by Pluralsight we will request
 339         # only single format until formats listing was explicitly requested.
 340         if self._downloader.params.get('listformats', False):
 341             allowed_qualities = ALLOWED_QUALITIES
 342         else:
 343             def guess_allowed_qualities():
 344                 req_format = self._downloader.params.get('format') or 'best'
 345                 req_format_split = req_format.split('-', 1)
 346                 if len(req_format_split) > 1:
 347                     req_ext, req_quality = req_format_split
 348                     req_quality = '-'.join(req_quality.split('-')[:2])
 349                     for allowed_quality in ALLOWED_QUALITIES:
 350                         if req_ext == allowed_quality.ext and req_quality in allowed_quality.qualities:
 351                             return (AllowedQuality(req_ext, (req_quality, )), )
 352                 req_ext = 'webm' if self._downloader.params.get('prefer_free_formats') else 'mp4'
 353                 return (AllowedQuality(req_ext, (best_quality, )), )
 354             allowed_qualities = guess_allowed_qualities()
 355
 356         formats = []
 357         for ext, qualities_ in allowed_qualities:
 358             for quality in qualities_:
 359                 f = QUALITIES[quality].copy()
 360                 clip_post = {
 361                     'author': author,
 362                     'includeCaptions': 'false',
 363                     'clipIndex': int(clip_idx),
 364                     'courseName': course_name,
 365                     'locale': 'en',
 366                     'moduleName': name,
 367                     'mediaType': ext,
 368                     'quality': '%dx%d' % (f['width'], f['height']),
 369                 }
 370                 format_id = '%s-%s' % (ext, quality)
 371
 372                 try:
 373                     viewclip = self._download_json(
 374                         self._GRAPHQL_EP, display_id,
 375                         'Downloading %s viewclip graphql' % format_id,
 376                         data=json.dumps({
 377                             'query': self.GRAPHQL_VIEWCLIP_TMPL % clip_post,
 378                             'variables': {}
 379                         }).encode('utf-8'),
 380                         headers=self._GRAPHQL_HEADERS)['data']['viewClip']
 381                 except ExtractorError:
 382                     # Still works but most likely will go soon
 383                     viewclip = self._download_json(
 384                         '%s/video/clips/viewclip' % self._API_BASE, display_id,
 385                         'Downloading %s viewclip JSON' % format_id, fatal=False,
 386                         data=json.dumps(clip_post).encode('utf-8'),
 387                         headers={'Content-Type': 'application/json;charset=utf-8'})
 388
 389                 # Pluralsight tracks multiple sequential calls to ViewClip API and start
 390                 # to return 429 HTTP errors after some time (see
 391                 # https://github.com/rg3/youtube-dl/pull/6989). Moreover it may even lead
 392                 # to account ban (see https://github.com/rg3/youtube-dl/issues/6842).
 393                 # To somewhat reduce the probability of these consequences
 394                 # we will sleep random amount of time before each call to ViewClip.
 395                 self._sleep(
 396                     random.randint(2, 5), display_id,
 397                     '%(video_id)s: Waiting for %(timeout)s seconds to avoid throttling')
 398
 399                 if not viewclip:
 400                     continue
 401
 402                 clip_urls = viewclip.get('urls')
 403                 if not isinstance(clip_urls, list):
 404                     continue
 405
 406                 for clip_url_data in clip_urls:
 407                     clip_url = clip_url_data.get('url')
 408                     if not clip_url:
 409                         continue
 410                     cdn = clip_url_data.get('cdn')
 411                     clip_f = f.copy()
 412                     clip_f.update({
 413                         'url': clip_url,
 414                         'ext': ext,
 415                         'format_id': '%s-%s' % (format_id, cdn) if cdn else format_id,
 416                         'quality': quality_key(quality),
 417                         'source_preference': int_or_none(clip_url_data.get('rank')),
 418                     })
 419                     formats.append(clip_f)
 420
 421         self._sort_formats(formats)
 422
 423         duration = int_or_none(
 424             clip.get('duration')) or parse_duration(clip.get('formattedDuration'))
 425
 426         # TODO: other languages?
 427         subtitles = self.extract_subtitles(
 428             author, clip_idx, clip.get('clipId'), 'en', name, duration, display_id)
 429
 430         return {
 431             'id': clip_id,
 432             'title': title,
 433             'duration': duration,
 434             'creator': author,
 435             'formats': formats,
 436             'subtitles': subtitles,
 437         }
 438
 439
 440 class PluralsightCourseIE(PluralsightBaseIE):
 441     IE_NAME = 'pluralsight:course'
 442     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:library/)?courses/(?P<id>[^/]+)'
 443     _TESTS = [{
 444         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
 445         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
 446         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',
 447         'info_dict': {
 448             'id': 'hosting-sql-server-windows-azure-iaas',
 449             'title': 'Hosting SQL Server in Microsoft Azure IaaS Fundamentals',
 450             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
 451         },
 452         'playlist_count': 31,
 453     }, {
 454         # available without pluralsight account
 455         'url': 'https://www.pluralsight.com/courses/angularjs-get-started',
 456         'only_matching': True,
 457     }, {
 458         'url': 'https://app.pluralsight.com/library/courses/understanding-microsoft-azure-amazon-aws/table-of-contents',
 459         'only_matching': True,
 460     }]
 461
 462     def _real_extract(self, url):
 463         course_id = self._match_id(url)
 464
 465         # TODO: PSM cookie
 466
 467         course = self._download_course(course_id, url, course_id)
 468
 469         title = course['title']
 470         course_name = course['name']
 471         course_data = course['modules']
 472         description = course.get('description') or course.get('shortDescription')
 473
 474         entries = []
 475         for num, module in enumerate(course_data, 1):
 476             author = module.get('author')
 477             module_name = module.get('name')
 478             if not author or not module_name:
 479                 continue
 480             for clip in module.get('clips', []):
 481                 clip_index = int_or_none(clip.get('index'))
 482                 if clip_index is None:
 483                     continue
 484                 clip_url = update_url_query(
 485                     '%s/player' % self._API_BASE, query={
 486                         'mode': 'live',
 487                         'course': course_name,
 488                         'author': author,
 489                         'name': module_name,
 490                         'clip': clip_index,
 491                     })
 492                 entries.append({
 493                     '_type': 'url_transparent',
 494                     'url': clip_url,
 495                     'ie_key': PluralsightIE.ie_key(),
 496                     'chapter': module.get('title'),
 497                     'chapter_number': num,
 498                     'chapter_id': module.get('moduleRef'),
 499                 })
 500
 501         return self.playlist_result(entries, course_id, title, description)