youtube_dl/extractor/lynda.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5
   6 from .subtitles import SubtitlesInfoExtractor
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     compat_urllib_parse,
  10     compat_urllib_request,
  11     ExtractorError,
  12     int_or_none,
  13     compat_str,
  14 )
  15
  16
  17 class LyndaIE(SubtitlesInfoExtractor):
  18     IE_NAME = 'lynda'
  19     IE_DESC = 'lynda.com videos'
  20     _VALID_URL = r'https?://www\.lynda\.com/[^/]+/[^/]+/\d+/(\d+)-\d\.html'
  21     _LOGIN_URL = 'https://www.lynda.com/login/login.aspx'
  22     _NETRC_MACHINE = 'lynda'
  23
  24     _SUCCESSFUL_LOGIN_REGEX = r'isLoggedIn: true'
  25     _TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
  26
  27     ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
  28
  29     _TEST = {
  30         'url': 'http://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
  31         'md5': 'ecfc6862da89489161fb9cd5f5a6fac1',
  32         'info_dict': {
  33             'id': '114408',
  34             'ext': 'mp4',
  35             'title': 'Using the exercise files',
  36             'duration': 68
  37         }
  38     }
  39
  40     def _real_initialize(self):
  41         self._login()
  42
  43     def _real_extract(self, url):
  44         mobj = re.match(self._VALID_URL, url)
  45         video_id = mobj.group(1)
  46
  47         page = self._download_webpage('http://www.lynda.com/ajax/player?videoId=%s&type=video' % video_id, video_id,
  48             'Downloading video JSON')
  49         video_json = json.loads(page)
  50
  51         if 'Status' in video_json:
  52             raise ExtractorError('lynda returned error: %s' % video_json['Message'], expected=True)
  53
  54         if video_json['HasAccess'] is False:
  55             raise ExtractorError(
  56                 'Video %s is only available for members. ' % video_id + self.ACCOUNT_CREDENTIALS_HINT, expected=True)
  57
  58         video_id = compat_str(video_json['ID'])
  59         duration = video_json['DurationInSeconds']
  60         title = video_json['Title']
  61
  62         formats = []
  63
  64         fmts = video_json.get('Formats')
  65         if fmts:
  66             formats.extend([
  67                 {
  68                     'url': fmt['Url'],
  69                     'ext': fmt['Extension'],
  70                     'width': fmt['Width'],
  71                     'height': fmt['Height'],
  72                     'filesize': fmt['FileSize'],
  73                     'format_id': str(fmt['Resolution'])
  74                 } for fmt in fmts])
  75
  76         prioritized_streams = video_json.get('PrioritizedStreams')
  77         if prioritized_streams:
  78             formats.extend([
  79                 {
  80                     'url': video_url,
  81                     'width': int_or_none(format_id),
  82                     'format_id': format_id,
  83                 } for format_id, video_url in prioritized_streams['0'].items()
  84             ])
  85
  86         self._sort_formats(formats)
  87
  88         if self._downloader.params.get('listsubtitles', False):
  89             self._list_available_subtitles(video_id, page)
  90             return
  91
  92         subtitles = self._fix_subtitles(self.extract_subtitles(video_id, page))
  93
  94         return {
  95             'id': video_id,
  96             'title': title,
  97             'duration': duration,
  98             'subtitles': subtitles,
  99             'formats': formats
 100         }
 101
 102     def _login(self):
 103         (username, password) = self._get_login_info()
 104         if username is None:
 105             return
 106
 107         login_form = {
 108             'username': username,
 109             'password': password,
 110             'remember': 'false',
 111             'stayPut': 'false'
 112         }
 113         request = compat_urllib_request.Request(self._LOGIN_URL, compat_urllib_parse.urlencode(login_form))
 114         login_page = self._download_webpage(request, None, 'Logging in as %s' % username)
 115
 116         # Not (yet) logged in
 117         m = re.search(r'loginResultJson = \'(?P<json>[^\']+)\';', login_page)
 118         if m is not None:
 119             response = m.group('json')
 120             response_json = json.loads(response)
 121             state = response_json['state']
 122
 123             if state == 'notlogged':
 124                 raise ExtractorError('Unable to login, incorrect username and/or password', expected=True)
 125
 126             # This is when we get popup:
 127             # > You're already logged in to lynda.com on two devices.
 128             # > If you log in here, we'll log you out of another device.
 129             # So, we need to confirm this.
 130             if state == 'conflicted':
 131                 confirm_form = {
 132                     'username': '',
 133                     'password': '',
 134                     'resolve': 'true',
 135                     'remember': 'false',
 136                     'stayPut': 'false',
 137                 }
 138                 request = compat_urllib_request.Request(self._LOGIN_URL, compat_urllib_parse.urlencode(confirm_form))
 139                 login_page = self._download_webpage(request, None, 'Confirming log in and log out from another device')
 140
 141         if re.search(self._SUCCESSFUL_LOGIN_REGEX, login_page) is None:
 142             raise ExtractorError('Unable to log in')
 143
 144     def _fix_subtitles(self, subtitles):
 145         if subtitles is None:
 146             return subtitles  # subtitles not requested
 147
 148         fixed_subtitles = {}
 149         for k, v in subtitles.items():
 150             subs = json.loads(v)
 151             if len(subs) == 0:
 152                 continue
 153             srt = ''
 154             for pos in range(0, len(subs) - 1):
 155                 seq_current = subs[pos]
 156                 m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
 157                 if m_current is None:
 158                     continue
 159                 seq_next = subs[pos + 1]
 160                 m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
 161                 if m_next is None:
 162                     continue
 163                 appear_time = m_current.group('timecode')
 164                 disappear_time = m_next.group('timecode')
 165                 text = seq_current['Caption']
 166                 srt += '%s\r\n%s --> %s\r\n%s' % (str(pos), appear_time, disappear_time, text)
 167             if srt:
 168                 fixed_subtitles[k] = srt
 169         return fixed_subtitles
 170
 171     def _get_available_subtitles(self, video_id, webpage):
 172         url = 'http://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
 173         sub = self._download_webpage(url, None, False)
 174         sub_json = json.loads(sub)
 175         return {'en': url} if len(sub_json) > 0 else {}
 176
 177
 178 class LyndaCourseIE(InfoExtractor):
 179     IE_NAME = 'lynda:course'
 180     IE_DESC = 'lynda.com online courses'
 181
 182     # Course link equals to welcome/introduction video link of same course
 183     # We will recognize it as course link
 184     _VALID_URL = r'https?://(?:www|m)\.lynda\.com/(?P<coursepath>[^/]+/[^/]+/(?P<courseid>\d+))-\d\.html'
 185
 186     def _real_extract(self, url):
 187         mobj = re.match(self._VALID_URL, url)
 188         course_path = mobj.group('coursepath')
 189         course_id = mobj.group('courseid')
 190
 191         page = self._download_webpage('http://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
 192                                       course_id, 'Downloading course JSON')
 193         course_json = json.loads(page)
 194
 195         if 'Status' in course_json and course_json['Status'] == 'NotFound':
 196             raise ExtractorError('Course %s does not exist' % course_id, expected=True)
 197
 198         unaccessible_videos = 0
 199         videos = []
 200         (username, _) = self._get_login_info()
 201
 202         # Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
 203         # by single video API anymore
 204
 205         for chapter in course_json['Chapters']:
 206             for video in chapter['Videos']:
 207                 if username is None and video['HasAccess'] is False:
 208                     unaccessible_videos += 1
 209                     continue
 210                 videos.append(video['ID'])
 211
 212         if unaccessible_videos > 0:
 213             self._downloader.report_warning('%s videos are only available for members and will not be downloaded. '
 214                                             % unaccessible_videos + LyndaIE.ACCOUNT_CREDENTIALS_HINT)
 215
 216         entries = [
 217             self.url_result('http://www.lynda.com/%s/%s-4.html' %
 218                             (course_path, video_id),
 219                             'Lynda')
 220             for video_id in videos]
 221
 222         course_title = course_json['Title']
 223
 224         return self.playlist_result(entries, course_id, course_title)