Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/pluralsight.py

   1 from __future__ import unicode_literals
   2
   3 import json
   4 import random
   5 import collections
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_str,
  10     compat_urllib_parse,
  11     compat_urlparse,
  12 )
  13 from ..utils import (
  14     ExtractorError,
  15     int_or_none,
  16     parse_duration,
  17     sanitized_Request,
  18 )
  19
  20
  21 class PluralsightBaseIE(InfoExtractor):
  22     _API_BASE = 'http://app.pluralsight.com'
  23
  24
  25 class PluralsightIE(PluralsightBaseIE):
  26     IE_NAME = 'pluralsight'
  27     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/training/player\?'
  28     _LOGIN_URL = 'https://app.pluralsight.com/id/'
  29
  30     _NETRC_MACHINE = 'pluralsight'
  31
  32     _TESTS = [{
  33         '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',
  34         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
  35         'info_dict': {
  36             'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04',
  37             'ext': 'mp4',
  38             'title': 'Management of SQL Server - Demo Monitoring',
  39             'duration': 338,
  40         },
  41         'skip': 'Requires pluralsight account credentials',
  42     }, {
  43         'url': 'https://app.pluralsight.com/training/player?course=angularjs-get-started&author=scott-allen&name=angularjs-get-started-m1-introduction&clip=0&mode=live',
  44         'only_matching': True,
  45     }, {
  46         # available without pluralsight account
  47         'url': 'http://app.pluralsight.com/training/player?author=scott-allen&name=angularjs-get-started-m1-introduction&mode=live&clip=0&course=angularjs-get-started',
  48         'only_matching': True,
  49     }]
  50
  51     def _real_initialize(self):
  52         self._login()
  53
  54     def _login(self):
  55         (username, password) = self._get_login_info()
  56         if username is None:
  57             return
  58
  59         login_page = self._download_webpage(
  60             self._LOGIN_URL, None, 'Downloading login page')
  61
  62         login_form = self._hidden_inputs(login_page)
  63
  64         login_form.update({
  65             'Username': username.encode('utf-8'),
  66             'Password': password.encode('utf-8'),
  67         })
  68
  69         post_url = self._search_regex(
  70             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
  71             'post url', default=self._LOGIN_URL, group='url')
  72
  73         if not post_url.startswith('http'):
  74             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
  75
  76         request = sanitized_Request(
  77             post_url, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
  78         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  79
  80         response = self._download_webpage(
  81             request, None, 'Logging in as %s' % username)
  82
  83         error = self._search_regex(
  84             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',
  85             response, 'error message', default=None)
  86         if error:
  87             raise ExtractorError('Unable to login: %s' % error, expected=True)
  88
  89         if all(p not in response for p in ('__INITIAL_STATE__', '"currentUser"')):
  90             raise ExtractorError('Unable to log in')
  91
  92     def _real_extract(self, url):
  93         qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
  94
  95         author = qs.get('author', [None])[0]
  96         name = qs.get('name', [None])[0]
  97         clip_id = qs.get('clip', [None])[0]
  98         course = qs.get('course', [None])[0]
  99
 100         if any(not f for f in (author, name, clip_id, course,)):
 101             raise ExtractorError('Invalid URL', expected=True)
 102
 103         display_id = '%s-%s' % (name, clip_id)
 104
 105         webpage = self._download_webpage(url, display_id)
 106
 107         modules = self._search_regex(
 108             r'moduleCollection\s*:\s*new\s+ModuleCollection\((\[.+?\])\s*,\s*\$rootScope\)',
 109             webpage, 'modules', default=None)
 110
 111         if modules:
 112             collection = self._parse_json(modules, display_id)
 113         else:
 114             # Webpage may be served in different layout (see
 115             # https://github.com/rg3/youtube-dl/issues/7607)
 116             collection = self._parse_json(
 117                 self._search_regex(
 118                     r'var\s+initialState\s*=\s*({.+?});\n', webpage, 'initial state'),
 119                 display_id)['course']['modules']
 120
 121         module, clip = None, None
 122
 123         for module_ in collection:
 124             if name in (module_.get('moduleName'), module_.get('name')):
 125                 module = module_
 126                 for clip_ in module_.get('clips', []):
 127                     clip_index = clip_.get('clipIndex')
 128                     if clip_index is None:
 129                         clip_index = clip_.get('index')
 130                     if clip_index is None:
 131                         continue
 132                     if compat_str(clip_index) == clip_id:
 133                         clip = clip_
 134                         break
 135
 136         if not clip:
 137             raise ExtractorError('Unable to resolve clip')
 138
 139         QUALITIES = {
 140             'low': {'width': 640, 'height': 480},
 141             'medium': {'width': 848, 'height': 640},
 142             'high': {'width': 1024, 'height': 768},
 143         }
 144
 145         AllowedQuality = collections.namedtuple('AllowedQuality', ['ext', 'qualities'])
 146
 147         ALLOWED_QUALITIES = (
 148             AllowedQuality('webm', ('high',)),
 149             AllowedQuality('mp4', ('low', 'medium', 'high',)),
 150         )
 151
 152         # In order to minimize the number of calls to ViewClip API and reduce
 153         # the probability of being throttled or banned by Pluralsight we will request
 154         # only single format until formats listing was explicitly requested.
 155         if self._downloader.params.get('listformats', False):
 156             allowed_qualities = ALLOWED_QUALITIES
 157         else:
 158             def guess_allowed_qualities():
 159                 req_format = self._downloader.params.get('format') or 'best'
 160                 req_format_split = req_format.split('-')
 161                 if len(req_format_split) > 1:
 162                     req_ext, req_quality = req_format_split
 163                     for allowed_quality in ALLOWED_QUALITIES:
 164                         if req_ext == allowed_quality.ext and req_quality in allowed_quality.qualities:
 165                             return (AllowedQuality(req_ext, (req_quality, )), )
 166                 req_ext = 'webm' if self._downloader.params.get('prefer_free_formats') else 'mp4'
 167                 return (AllowedQuality(req_ext, ('high', )), )
 168             allowed_qualities = guess_allowed_qualities()
 169
 170         formats = []
 171         for ext, qualities in allowed_qualities:
 172             for quality in qualities:
 173                 f = QUALITIES[quality].copy()
 174                 clip_post = {
 175                     'a': author,
 176                     'cap': 'false',
 177                     'cn': clip_id,
 178                     'course': course,
 179                     'lc': 'en',
 180                     'm': name,
 181                     'mt': ext,
 182                     'q': '%dx%d' % (f['width'], f['height']),
 183                 }
 184                 request = sanitized_Request(
 185                     '%s/training/Player/ViewClip' % self._API_BASE,
 186                     json.dumps(clip_post).encode('utf-8'))
 187                 request.add_header('Content-Type', 'application/json;charset=utf-8')
 188                 format_id = '%s-%s' % (ext, quality)
 189                 clip_url = self._download_webpage(
 190                     request, display_id, 'Downloading %s URL' % format_id, fatal=False)
 191
 192                 # Pluralsight tracks multiple sequential calls to ViewClip API and start
 193                 # to return 429 HTTP errors after some time (see
 194                 # https://github.com/rg3/youtube-dl/pull/6989). Moreover it may even lead
 195                 # to account ban (see https://github.com/rg3/youtube-dl/issues/6842).
 196                 # To somewhat reduce the probability of these consequences
 197                 # we will sleep random amount of time before each call to ViewClip.
 198                 self._sleep(
 199                     random.randint(2, 5), display_id,
 200                     '%(video_id)s: Waiting for %(timeout)s seconds to avoid throttling')
 201
 202                 if not clip_url:
 203                     continue
 204                 f.update({
 205                     'url': clip_url,
 206                     'ext': ext,
 207                     'format_id': format_id,
 208                 })
 209                 formats.append(f)
 210         self._sort_formats(formats)
 211
 212         # TODO: captions
 213         # http://www.pluralsight.com/training/Player/ViewClip + cap = true
 214         # or
 215         # http://www.pluralsight.com/training/Player/Captions
 216         # { a = author, cn = clip_id, lc = end, m = name }
 217
 218         return {
 219             'id': clip['clipName'],
 220             'title': '%s - %s' % (module['title'], clip['title']),
 221             'duration': int_or_none(clip.get('duration')) or parse_duration(clip.get('formattedDuration')),
 222             'creator': author,
 223             'formats': formats
 224         }
 225
 226
 227 class PluralsightCourseIE(PluralsightBaseIE):
 228     IE_NAME = 'pluralsight:course'
 229     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:library/)?courses/(?P<id>[^/]+)'
 230     _TESTS = [{
 231         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
 232         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
 233         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',
 234         'info_dict': {
 235             'id': 'hosting-sql-server-windows-azure-iaas',
 236             'title': 'Hosting SQL Server in Microsoft Azure IaaS Fundamentals',
 237             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
 238         },
 239         'playlist_count': 31,
 240     }, {
 241         # available without pluralsight account
 242         'url': 'https://www.pluralsight.com/courses/angularjs-get-started',
 243         'only_matching': True,
 244     }, {
 245         'url': 'https://app.pluralsight.com/library/courses/understanding-microsoft-azure-amazon-aws/table-of-contents',
 246         'only_matching': True,
 247     }]
 248
 249     def _real_extract(self, url):
 250         course_id = self._match_id(url)
 251
 252         # TODO: PSM cookie
 253
 254         course = self._download_json(
 255             '%s/data/course/%s' % (self._API_BASE, course_id),
 256             course_id, 'Downloading course JSON')
 257
 258         title = course['title']
 259         description = course.get('description') or course.get('shortDescription')
 260
 261         course_data = self._download_json(
 262             '%s/data/course/content/%s' % (self._API_BASE, course_id),
 263             course_id, 'Downloading course data JSON')
 264
 265         entries = []
 266         for module in course_data:
 267             for clip in module.get('clips', []):
 268                 player_parameters = clip.get('playerParameters')
 269                 if not player_parameters:
 270                     continue
 271                 entries.append(self.url_result(
 272                     '%s/training/player?%s' % (self._API_BASE, player_parameters),
 273                     'Pluralsight'))
 274
 275         return self.playlist_result(entries, course_id, title, description)