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