_ Git - youtube-dl/blob - 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     ExtractorError,
  16     float_or_none,
  17     int_or_none,
  18     parse_duration,
  19     qualities,
  20     sanitized_Request,
  21     srt_subtitles_timecode,
  22     urlencode_postdata,
  23 )
  24
  25
  26 class PluralsightBaseIE(InfoExtractor):
  27     _API_BASE = 'http://app.pluralsight.com'
  28
  29
  30 class PluralsightIE(PluralsightBaseIE):
  31     IE_NAME = 'pluralsight'
  32     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/training/player\?'
  33     _LOGIN_URL = 'https://app.pluralsight.com/id/'
  34
  35     _NETRC_MACHINE = 'pluralsight'
  36
  37     _TESTS = [{
  38         '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',
  39         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
  40         'info_dict': {
  41             'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04',
  42             'ext': 'mp4',
  43             'title': 'Management of SQL Server - Demo Monitoring',
  44             'duration': 338,
  45         },
  46         'skip': 'Requires pluralsight account credentials',
  47     }, {
  48         'url': 'https://app.pluralsight.com/training/player?course=angularjs-get-started&author=scott-allen&name=angularjs-get-started-m1-introduction&clip=0&mode=live',
  49         'only_matching': True,
  50     }, {
  51         # available without pluralsight account
  52         'url': 'http://app.pluralsight.com/training/player?author=scott-allen&name=angularjs-get-started-m1-introduction&mode=live&clip=0&course=angularjs-get-started',
  53         'only_matching': True,
  54     }]
  55
  56     def _real_initialize(self):
  57         self._login()
  58
  59     def _login(self):
  60         (username, password) = self._get_login_info()
  61         if username is None:
  62             return
  63
  64         login_page = self._download_webpage(
  65             self._LOGIN_URL, None, 'Downloading login page')
  66
  67         login_form = self._hidden_inputs(login_page)
  68
  69         login_form.update({
  70             'Username': username,
  71             'Password': password,
  72         })
  73
  74         post_url = self._search_regex(
  75             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
  76             'post url', default=self._LOGIN_URL, group='url')
  77
  78         if not post_url.startswith('http'):
  79             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
  80
  81         request = sanitized_Request(
  82             post_url, urlencode_postdata(login_form))
  83         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  84
  85         response = self._download_webpage(
  86             request, None, 'Logging in as %s' % username)
  87
  88         error = self._search_regex(
  89             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',
  90             response, 'error message', default=None)
  91         if error:
  92             raise ExtractorError('Unable to login: %s' % error, expected=True)
  93
  94         if all(p not in response for p in ('__INITIAL_STATE__', '"currentUser"')):
  95             raise ExtractorError('Unable to log in')
  96
  97     def _get_subtitles(self, author, clip_id, lang, name, duration, video_id):
  98         captions_post = {
  99             'a': author,
 100             'cn': clip_id,
 101             'lc': lang,
 102             'm': name,
 103         }
 104         captions = self._download_json(
 105             '%s/training/Player/Captions' % self._API_BASE, video_id,
 106             'Downloading captions JSON', 'Unable to download captions JSON',
 107             fatal=False, data=json.dumps(captions_post).encode('utf-8'),
 108             headers={'Content-Type': 'application/json;charset=utf-8'})
 109         if captions:
 110             return {
 111                 lang: [{
 112                     'ext': 'json',
 113                     'data': json.dumps(captions),
 114                 }, {
 115                     'ext': 'srt',
 116                     'data': self._convert_subtitles(duration, captions),
 117                 }]
 118             }
 119
 120     @staticmethod
 121     def _convert_subtitles(duration, subs):
 122         srt = ''
 123         for num, current in enumerate(subs):
 124             current = subs[num]
 125             start, text = float_or_none(
 126                 current.get('DisplayTimeOffset')), current.get('Text')
 127             if start is None or text is None:
 128                 continue
 129             end = duration if num == len(subs) - 1 else float_or_none(
 130                 subs[num + 1].get('DisplayTimeOffset'))
 131             srt += os.linesep.join(
 132                 (
 133                     '%d' % num,
 134                     '%s --> %s' % (
 135                         srt_subtitles_timecode(start),
 136                         srt_subtitles_timecode(end)),
 137                     text,
 138                     os.linesep,
 139                 ))
 140         return srt
 141
 142     def _real_extract(self, url):
 143         qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 144
 145         author = qs.get('author', [None])[0]
 146         name = qs.get('name', [None])[0]
 147         clip_id = qs.get('clip', [None])[0]
 148         course = qs.get('course', [None])[0]
 149
 150         if any(not f for f in (author, name, clip_id, course,)):
 151             raise ExtractorError('Invalid URL', expected=True)
 152
 153         display_id = '%s-%s' % (name, clip_id)
 154
 155         webpage = self._download_webpage(url, display_id)
 156
 157         modules = self._search_regex(
 158             r'moduleCollection\s*:\s*new\s+ModuleCollection\((\[.+?\])\s*,\s*\$rootScope\)',
 159             webpage, 'modules', default=None)
 160
 161         if modules:
 162             collection = self._parse_json(modules, display_id)
 163         else:
 164             # Webpage may be served in different layout (see
 165             # https://github.com/rg3/youtube-dl/issues/7607)
 166             collection = self._parse_json(
 167                 self._search_regex(
 168                     r'var\s+initialState\s*=\s*({.+?});\n', webpage, 'initial state'),
 169                 display_id)['course']['modules']
 170
 171         module, clip = None, None
 172
 173         for module_ in collection:
 174             if name in (module_.get('moduleName'), module_.get('name')):
 175                 module = module_
 176                 for clip_ in module_.get('clips', []):
 177                     clip_index = clip_.get('clipIndex')
 178                     if clip_index is None:
 179                         clip_index = clip_.get('index')
 180                     if clip_index is None:
 181                         continue
 182                     if compat_str(clip_index) == clip_id:
 183                         clip = clip_
 184                         break
 185
 186         if not clip:
 187             raise ExtractorError('Unable to resolve clip')
 188
 189         title = '%s - %s' % (module['title'], clip['title'])
 190
 191         QUALITIES = {
 192             'low': {'width': 640, 'height': 480},
 193             'medium': {'width': 848, 'height': 640},
 194             'high': {'width': 1024, 'height': 768},
 195             'high-widescreen': {'width': 1280, 'height': 720},
 196         }
 197
 198         QUALITIES_PREFERENCE = ('low', 'medium', 'high', 'high-widescreen',)
 199         quality_key = qualities(QUALITIES_PREFERENCE)
 200
 201         AllowedQuality = collections.namedtuple('AllowedQuality', ['ext', 'qualities'])
 202
 203         ALLOWED_QUALITIES = (
 204             AllowedQuality('webm', ['high', ]),
 205             AllowedQuality('mp4', ['low', 'medium', 'high', ]),
 206         )
 207
 208         # Some courses also offer widescreen resolution for high quality (see
 209         # https://github.com/rg3/youtube-dl/issues/7766)
 210         widescreen = True if re.search(
 211             r'courseSupportsWidescreenVideoFormats\s*:\s*true', webpage) else False
 212         best_quality = 'high-widescreen' if widescreen else 'high'
 213         if widescreen:
 214             for allowed_quality in ALLOWED_QUALITIES:
 215                 allowed_quality.qualities.append(best_quality)
 216
 217         # In order to minimize the number of calls to ViewClip API and reduce
 218         # the probability of being throttled or banned by Pluralsight we will request
 219         # only single format until formats listing was explicitly requested.
 220         if self._downloader.params.get('listformats', False):
 221             allowed_qualities = ALLOWED_QUALITIES
 222         else:
 223             def guess_allowed_qualities():
 224                 req_format = self._downloader.params.get('format') or 'best'
 225                 req_format_split = req_format.split('-', 1)
 226                 if len(req_format_split) > 1:
 227                     req_ext, req_quality = req_format_split
 228                     for allowed_quality in ALLOWED_QUALITIES:
 229                         if req_ext == allowed_quality.ext and req_quality in allowed_quality.qualities:
 230                             return (AllowedQuality(req_ext, (req_quality, )), )
 231                 req_ext = 'webm' if self._downloader.params.get('prefer_free_formats') else 'mp4'
 232                 return (AllowedQuality(req_ext, (best_quality, )), )
 233             allowed_qualities = guess_allowed_qualities()
 234
 235         formats = []
 236         for ext, qualities_ in allowed_qualities:
 237             for quality in qualities_:
 238                 f = QUALITIES[quality].copy()
 239                 clip_post = {
 240                     'a': author,
 241                     'cap': 'false',
 242                     'cn': clip_id,
 243                     'course': course,
 244                     'lc': 'en',
 245                     'm': name,
 246                     'mt': ext,
 247                     'q': '%dx%d' % (f['width'], f['height']),
 248                 }
 249                 request = sanitized_Request(
 250                     '%s/training/Player/ViewClip' % self._API_BASE,
 251                     json.dumps(clip_post).encode('utf-8'))
 252                 request.add_header('Content-Type', 'application/json;charset=utf-8')
 253                 format_id = '%s-%s' % (ext, quality)
 254                 clip_url = self._download_webpage(
 255                     request, display_id, 'Downloading %s URL' % format_id, fatal=False)
 256
 257                 # Pluralsight tracks multiple sequential calls to ViewClip API and start
 258                 # to return 429 HTTP errors after some time (see
 259                 # https://github.com/rg3/youtube-dl/pull/6989). Moreover it may even lead
 260                 # to account ban (see https://github.com/rg3/youtube-dl/issues/6842).
 261                 # To somewhat reduce the probability of these consequences
 262                 # we will sleep random amount of time before each call to ViewClip.
 263                 self._sleep(
 264                     random.randint(2, 5), display_id,
 265                     '%(video_id)s: Waiting for %(timeout)s seconds to avoid throttling')
 266
 267                 if not clip_url:
 268                     continue
 269                 f.update({
 270                     'url': clip_url,
 271                     'ext': ext,
 272                     'format_id': format_id,
 273                     'quality': quality_key(quality),
 274                 })
 275                 formats.append(f)
 276         self._sort_formats(formats)
 277
 278         duration = int_or_none(
 279             clip.get('duration')) or parse_duration(clip.get('formattedDuration'))
 280
 281         # TODO: other languages?
 282         subtitles = self.extract_subtitles(
 283             author, clip_id, 'en', name, duration, display_id)
 284
 285         return {
 286             'id': clip.get('clipName') or clip['name'],
 287             'title': title,
 288             'duration': duration,
 289             'creator': author,
 290             'formats': formats,
 291             'subtitles': subtitles,
 292         }
 293
 294
 295 class PluralsightCourseIE(PluralsightBaseIE):
 296     IE_NAME = 'pluralsight:course'
 297     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:library/)?courses/(?P<id>[^/]+)'
 298     _TESTS = [{
 299         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
 300         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
 301         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',
 302         'info_dict': {
 303             'id': 'hosting-sql-server-windows-azure-iaas',
 304             'title': 'Hosting SQL Server in Microsoft Azure IaaS Fundamentals',
 305             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
 306         },
 307         'playlist_count': 31,
 308     }, {
 309         # available without pluralsight account
 310         'url': 'https://www.pluralsight.com/courses/angularjs-get-started',
 311         'only_matching': True,
 312     }, {
 313         'url': 'https://app.pluralsight.com/library/courses/understanding-microsoft-azure-amazon-aws/table-of-contents',
 314         'only_matching': True,
 315     }]
 316
 317     def _real_extract(self, url):
 318         course_id = self._match_id(url)
 319
 320         # TODO: PSM cookie
 321
 322         course = self._download_json(
 323             '%s/data/course/%s' % (self._API_BASE, course_id),
 324             course_id, 'Downloading course JSON')
 325
 326         title = course['title']
 327         description = course.get('description') or course.get('shortDescription')
 328
 329         course_data = self._download_json(
 330             '%s/data/course/content/%s' % (self._API_BASE, course_id),
 331             course_id, 'Downloading course data JSON')
 332
 333         entries = []
 334         for num, module in enumerate(course_data, 1):
 335             for clip in module.get('clips', []):
 336                 player_parameters = clip.get('playerParameters')
 337                 if not player_parameters:
 338                     continue
 339                 entries.append({
 340                     '_type': 'url_transparent',
 341                     'url': '%s/training/player?%s' % (self._API_BASE, player_parameters),
 342                     'ie_key': PluralsightIE.ie_key(),
 343                     'chapter': module.get('title'),
 344                     'chapter_number': num,
 345                     'chapter_id': module.get('moduleRef'),
 346                 })
 347
 348         return self.playlist_result(entries, course_id, title, description)