_ Git - youtube-dl/blob - youtube_dl/extractor/lynda.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5
   6 from .common import InfoExtractor
   7 from ..compat import compat_str
   8 from ..utils import (
   9     ExtractorError,
  10     clean_html,
  11     int_or_none,
  12     sanitized_Request,
  13     urlencode_postdata,
  14 )
  15
  16
  17 class LyndaBaseIE(InfoExtractor):
  18     _LOGIN_URL = 'https://www.lynda.com/login/login.aspx'
  19     _ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
  20     _NETRC_MACHINE = 'lynda'
  21
  22     def _real_initialize(self):
  23         self._login()
  24
  25     def _login(self):
  26         username, password = self._get_login_info()
  27         if username is None:
  28             return
  29
  30         login_form = {
  31             'username': username.encode('utf-8'),
  32             'password': password.encode('utf-8'),
  33             'remember': 'false',
  34             'stayPut': 'false'
  35         }
  36         request = sanitized_Request(
  37             self._LOGIN_URL, urlencode_postdata(login_form))
  38         login_page = self._download_webpage(
  39             request, None, 'Logging in as %s' % username)
  40
  41         # Not (yet) logged in
  42         m = re.search(r'loginResultJson\s*=\s*\'(?P<json>[^\']+)\';', login_page)
  43         if m is not None:
  44             response = m.group('json')
  45             response_json = json.loads(response)
  46             state = response_json['state']
  47
  48             if state == 'notlogged':
  49                 raise ExtractorError(
  50                     'Unable to login, incorrect username and/or password',
  51                     expected=True)
  52
  53             # This is when we get popup:
  54             # > You're already logged in to lynda.com on two devices.
  55             # > If you log in here, we'll log you out of another device.
  56             # So, we need to confirm this.
  57             if state == 'conflicted':
  58                 confirm_form = {
  59                     'username': '',
  60                     'password': '',
  61                     'resolve': 'true',
  62                     'remember': 'false',
  63                     'stayPut': 'false',
  64                 }
  65                 request = sanitized_Request(
  66                     self._LOGIN_URL, urlencode_postdata(confirm_form))
  67                 login_page = self._download_webpage(
  68                     request, None,
  69                     'Confirming log in and log out from another device')
  70
  71         if all(not re.search(p, login_page) for p in ('isLoggedIn\s*:\s*true', r'logout\.aspx', r'>Log out<')):
  72             if 'login error' in login_page:
  73                 mobj = re.search(
  74                     r'(?s)<h1[^>]+class="topmost">(?P<title>[^<]+)</h1>\s*<div>(?P<description>.+?)</div>',
  75                     login_page)
  76                 if mobj:
  77                     raise ExtractorError(
  78                         'lynda returned error: %s - %s'
  79                         % (mobj.group('title'), clean_html(mobj.group('description'))),
  80                         expected=True)
  81             raise ExtractorError('Unable to log in')
  82
  83     def _logout(self):
  84         username, _ = self._get_login_info()
  85         if username is None:
  86             return
  87
  88         self._download_webpage(
  89             'http://www.lynda.com/ajax/logout.aspx', None,
  90             'Logging out', 'Unable to log out', fatal=False)
  91
  92
  93 class LyndaIE(LyndaBaseIE):
  94     IE_NAME = 'lynda'
  95     IE_DESC = 'lynda.com videos'
  96     _VALID_URL = r'https?://www\.lynda\.com/(?:[^/]+/[^/]+/\d+|player/embed)/(?P<id>\d+)'
  97     _NETRC_MACHINE = 'lynda'
  98
  99     _TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
 100
 101     _TESTS = [{
 102         'url': 'http://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
 103         'md5': 'ecfc6862da89489161fb9cd5f5a6fac1',
 104         'info_dict': {
 105             'id': '114408',
 106             'ext': 'mp4',
 107             'title': 'Using the exercise files',
 108             'duration': 68
 109         }
 110     }, {
 111         'url': 'https://www.lynda.com/player/embed/133770?tr=foo=1;bar=g;fizz=rt&fs=0',
 112         'only_matching': True,
 113     }]
 114
 115     def _real_extract(self, url):
 116         video_id = self._match_id(url)
 117
 118         video = self._download_json(
 119             'http://www.lynda.com/ajax/player?videoId=%s&type=video' % video_id,
 120             video_id, 'Downloading video JSON')
 121
 122         if 'Status' in video:
 123             raise ExtractorError(
 124                 'lynda returned error: %s' % video['Message'], expected=True)
 125
 126         if video.get('HasAccess') is False:
 127             self.raise_login_required('Video %s is only available for members' % video_id)
 128
 129         video_id = compat_str(video.get('ID') or video_id)
 130         duration = int_or_none(video.get('DurationInSeconds'))
 131         title = video['Title']
 132
 133         formats = []
 134
 135         fmts = video.get('Formats')
 136         if fmts:
 137             formats.extend([{
 138                 'url': f['Url'],
 139                 'ext': f.get('Extension'),
 140                 'width': int_or_none(f.get('Width')),
 141                 'height': int_or_none(f.get('Height')),
 142                 'filesize': int_or_none(f.get('FileSize')),
 143                 'format_id': compat_str(f.get('Resolution')) if f.get('Resolution') else None,
 144             } for f in fmts if f.get('Url')])
 145
 146         prioritized_streams = video.get('PrioritizedStreams')
 147         if prioritized_streams:
 148             for prioritized_stream_id, prioritized_stream in prioritized_streams.items():
 149                 formats.extend([{
 150                     'url': video_url,
 151                     'width': int_or_none(format_id),
 152                     'format_id': '%s-%s' % (prioritized_stream_id, format_id),
 153                 } for format_id, video_url in prioritized_stream.items()])
 154
 155         self._check_formats(formats, video_id)
 156         self._sort_formats(formats)
 157
 158         subtitles = self.extract_subtitles(video_id)
 159
 160         return {
 161             'id': video_id,
 162             'title': title,
 163             'duration': duration,
 164             'subtitles': subtitles,
 165             'formats': formats
 166         }
 167
 168     def _fix_subtitles(self, subs):
 169         srt = ''
 170         seq_counter = 0
 171         for pos in range(0, len(subs) - 1):
 172             seq_current = subs[pos]
 173             m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
 174             if m_current is None:
 175                 continue
 176             seq_next = subs[pos + 1]
 177             m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
 178             if m_next is None:
 179                 continue
 180             appear_time = m_current.group('timecode')
 181             disappear_time = m_next.group('timecode')
 182             text = seq_current['Caption'].strip()
 183             if text:
 184                 seq_counter += 1
 185                 srt += '%s\r\n%s --> %s\r\n%s\r\n\r\n' % (seq_counter, appear_time, disappear_time, text)
 186         if srt:
 187             return srt
 188
 189     def _get_subtitles(self, video_id):
 190         url = 'http://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
 191         subs = self._download_json(url, None, False)
 192         if subs:
 193             return {'en': [{'ext': 'srt', 'data': self._fix_subtitles(subs)}]}
 194         else:
 195             return {}
 196
 197
 198 class LyndaCourseIE(LyndaBaseIE):
 199     IE_NAME = 'lynda:course'
 200     IE_DESC = 'lynda.com online courses'
 201
 202     # Course link equals to welcome/introduction video link of same course
 203     # We will recognize it as course link
 204     _VALID_URL = r'https?://(?:www|m)\.lynda\.com/(?P<coursepath>[^/]+/[^/]+/(?P<courseid>\d+))-\d\.html'
 205
 206     def _real_extract(self, url):
 207         mobj = re.match(self._VALID_URL, url)
 208         course_path = mobj.group('coursepath')
 209         course_id = mobj.group('courseid')
 210
 211         course = self._download_json(
 212             'http://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
 213             course_id, 'Downloading course JSON')
 214
 215         self._logout()
 216
 217         if course.get('Status') == 'NotFound':
 218             raise ExtractorError(
 219                 'Course %s does not exist' % course_id, expected=True)
 220
 221         unaccessible_videos = 0
 222         entries = []
 223
 224         # Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
 225         # by single video API anymore
 226
 227         for chapter in course['Chapters']:
 228             for video in chapter.get('Videos', []):
 229                 if video.get('HasAccess') is False:
 230                     unaccessible_videos += 1
 231                     continue
 232                 video_id = video.get('ID')
 233                 if video_id:
 234                     entries.append({
 235                         '_type': 'url_transparent',
 236                         'url': 'http://www.lynda.com/%s/%s-4.html' % (course_path, video_id),
 237                         'ie_key': LyndaIE.ie_key(),
 238                         'chapter': chapter.get('Title'),
 239                         'chapter_number': int_or_none(chapter.get('ChapterIndex')),
 240                         'chapter_id': compat_str(chapter.get('ID')),
 241                     })
 242
 243         if unaccessible_videos > 0:
 244             self._downloader.report_warning(
 245                 '%s videos are only available for members (or paid members) and will not be downloaded. '
 246                 % unaccessible_videos + self._ACCOUNT_CREDENTIALS_HINT)
 247
 248         course_title = course.get('Title')
 249
 250         return self.playlist_result(entries, course_id, course_title)