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

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..compat import (
   7     compat_HTTPError,
   8     compat_str,
   9     compat_urlparse,
  10 )
  11 from ..utils import (
  12     ExtractorError,
  13     int_or_none,
  14     urlencode_postdata,
  15 )
  16
  17
  18 class LyndaBaseIE(InfoExtractor):
  19     _SIGNIN_URL = 'https://www.lynda.com/signin'
  20     _PASSWORD_URL = 'https://www.lynda.com/signin/password'
  21     _USER_URL = 'https://www.lynda.com/signin/user'
  22     _ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
  23     _NETRC_MACHINE = 'lynda'
  24
  25     def _real_initialize(self):
  26         self._login()
  27
  28     @staticmethod
  29     def _check_error(json_string, key_or_keys):
  30         keys = [key_or_keys] if isinstance(key_or_keys, compat_str) else key_or_keys
  31         for key in keys:
  32             error = json_string.get(key)
  33             if error:
  34                 raise ExtractorError('Unable to login: %s' % error, expected=True)
  35
  36     def _login_step(self, form_html, fallback_action_url, extra_form_data, note, referrer_url):
  37         action_url = self._search_regex(
  38             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', form_html,
  39             'post url', default=fallback_action_url, group='url')
  40
  41         if not action_url.startswith('http'):
  42             action_url = compat_urlparse.urljoin(self._SIGNIN_URL, action_url)
  43
  44         form_data = self._hidden_inputs(form_html)
  45         form_data.update(extra_form_data)
  46
  47         try:
  48             response = self._download_json(
  49                 action_url, None, note,
  50                 data=urlencode_postdata(form_data),
  51                 headers={
  52                     'Referer': referrer_url,
  53                     'X-Requested-With': 'XMLHttpRequest',
  54                 })
  55         except ExtractorError as e:
  56             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 500:
  57                 response = self._parse_json(e.cause.read().decode('utf-8'), None)
  58                 self._check_error(response, ('email', 'password'))
  59             raise
  60
  61         self._check_error(response, 'ErrorMessage')
  62
  63         return response, action_url
  64
  65     def _login(self):
  66         username, password = self._get_login_info()
  67         if username is None:
  68             return
  69
  70         # Step 1: download signin page
  71         signin_page = self._download_webpage(
  72             self._SIGNIN_URL, None, 'Downloading signin page')
  73
  74         # Already logged in
  75         if any(re.search(p, signin_page) for p in (
  76                 'isLoggedIn\s*:\s*true', r'logout\.aspx', r'>Log out<')):
  77             return
  78
  79         # Step 2: submit email
  80         signin_form = self._search_regex(
  81             r'(?s)(<form[^>]+data-form-name=["\']signin["\'][^>]*>.+?</form>)',
  82             signin_page, 'signin form')
  83         signin_page, signin_url = self._login_step(
  84             signin_form, self._PASSWORD_URL, {'email': username},
  85             'Submitting email', self._SIGNIN_URL)
  86
  87         # Step 3: submit password
  88         password_form = signin_page['body']
  89         self._login_step(
  90             password_form, self._USER_URL, {'email': username, 'password': password},
  91             'Submitting password', signin_url)
  92
  93
  94 class LyndaIE(LyndaBaseIE):
  95     IE_NAME = 'lynda'
  96     IE_DESC = 'lynda.com videos'
  97     _VALID_URL = r'https?://(?:www\.)?lynda\.com/(?:[^/]+/[^/]+/(?P<course_id>\d+)|player/embed)/(?P<id>\d+)'
  98
  99     _TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
 100
 101     _TESTS = [{
 102         'url': 'https://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
 103         # md5 is unstable
 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 _raise_unavailable(self, video_id):
 116         self.raise_login_required(
 117             'Video %s is only available for members' % video_id)
 118
 119     def _real_extract(self, url):
 120         mobj = re.match(self._VALID_URL, url)
 121         video_id = mobj.group('id')
 122         course_id = mobj.group('course_id')
 123
 124         query = {
 125             'videoId': video_id,
 126             'type': 'video',
 127         }
 128
 129         video = self._download_json(
 130             'https://www.lynda.com/ajax/player', video_id,
 131             'Downloading video JSON', fatal=False, query=query)
 132
 133         # Fallback scenario
 134         if not video:
 135             query['courseId'] = course_id
 136
 137             play = self._download_json(
 138                 'https://www.lynda.com/ajax/course/%s/%s/play'
 139                 % (course_id, video_id), video_id, 'Downloading play JSON')
 140
 141             if not play:
 142                 self._raise_unavailable(video_id)
 143
 144             formats = []
 145             for formats_dict in play:
 146                 urls = formats_dict.get('urls')
 147                 if not isinstance(urls, dict):
 148                     continue
 149                 cdn = formats_dict.get('name')
 150                 for format_id, format_url in urls.items():
 151                     if not format_url:
 152                         continue
 153                     formats.append({
 154                         'url': format_url,
 155                         'format_id': '%s-%s' % (cdn, format_id) if cdn else format_id,
 156                         'height': int_or_none(format_id),
 157                     })
 158             self._sort_formats(formats)
 159
 160             conviva = self._download_json(
 161                 'https://www.lynda.com/ajax/player/conviva', video_id,
 162                 'Downloading conviva JSON', query=query)
 163
 164             return {
 165                 'id': video_id,
 166                 'title': conviva['VideoTitle'],
 167                 'description': conviva.get('VideoDescription'),
 168                 'release_year': int_or_none(conviva.get('ReleaseYear')),
 169                 'duration': int_or_none(conviva.get('Duration')),
 170                 'creator': conviva.get('Author'),
 171                 'formats': formats,
 172             }
 173
 174         if 'Status' in video:
 175             raise ExtractorError(
 176                 'lynda returned error: %s' % video['Message'], expected=True)
 177
 178         if video.get('HasAccess') is False:
 179             self._raise_unavailable(video_id)
 180
 181         video_id = compat_str(video.get('ID') or video_id)
 182         duration = int_or_none(video.get('DurationInSeconds'))
 183         title = video['Title']
 184
 185         formats = []
 186
 187         fmts = video.get('Formats')
 188         if fmts:
 189             formats.extend([{
 190                 'url': f['Url'],
 191                 'ext': f.get('Extension'),
 192                 'width': int_or_none(f.get('Width')),
 193                 'height': int_or_none(f.get('Height')),
 194                 'filesize': int_or_none(f.get('FileSize')),
 195                 'format_id': compat_str(f.get('Resolution')) if f.get('Resolution') else None,
 196             } for f in fmts if f.get('Url')])
 197
 198         prioritized_streams = video.get('PrioritizedStreams')
 199         if prioritized_streams:
 200             for prioritized_stream_id, prioritized_stream in prioritized_streams.items():
 201                 formats.extend([{
 202                     'url': video_url,
 203                     'height': int_or_none(format_id),
 204                     'format_id': '%s-%s' % (prioritized_stream_id, format_id),
 205                 } for format_id, video_url in prioritized_stream.items()])
 206
 207         self._check_formats(formats, video_id)
 208         self._sort_formats(formats)
 209
 210         subtitles = self.extract_subtitles(video_id)
 211
 212         return {
 213             'id': video_id,
 214             'title': title,
 215             'duration': duration,
 216             'subtitles': subtitles,
 217             'formats': formats
 218         }
 219
 220     def _fix_subtitles(self, subs):
 221         srt = ''
 222         seq_counter = 0
 223         for pos in range(0, len(subs) - 1):
 224             seq_current = subs[pos]
 225             m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
 226             if m_current is None:
 227                 continue
 228             seq_next = subs[pos + 1]
 229             m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
 230             if m_next is None:
 231                 continue
 232             appear_time = m_current.group('timecode')
 233             disappear_time = m_next.group('timecode')
 234             text = seq_current['Caption'].strip()
 235             if text:
 236                 seq_counter += 1
 237                 srt += '%s\r\n%s --> %s\r\n%s\r\n\r\n' % (seq_counter, appear_time, disappear_time, text)
 238         if srt:
 239             return srt
 240
 241     def _get_subtitles(self, video_id):
 242         url = 'https://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
 243         subs = self._download_json(url, None, False)
 244         if subs:
 245             return {'en': [{'ext': 'srt', 'data': self._fix_subtitles(subs)}]}
 246         else:
 247             return {}
 248
 249
 250 class LyndaCourseIE(LyndaBaseIE):
 251     IE_NAME = 'lynda:course'
 252     IE_DESC = 'lynda.com online courses'
 253
 254     # Course link equals to welcome/introduction video link of same course
 255     # We will recognize it as course link
 256     _VALID_URL = r'https?://(?:www|m)\.lynda\.com/(?P<coursepath>[^/]+/[^/]+/(?P<courseid>\d+))-\d\.html'
 257
 258     def _real_extract(self, url):
 259         mobj = re.match(self._VALID_URL, url)
 260         course_path = mobj.group('coursepath')
 261         course_id = mobj.group('courseid')
 262
 263         course = self._download_json(
 264             'https://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
 265             course_id, 'Downloading course JSON')
 266
 267         if course.get('Status') == 'NotFound':
 268             raise ExtractorError(
 269                 'Course %s does not exist' % course_id, expected=True)
 270
 271         unaccessible_videos = 0
 272         entries = []
 273
 274         # Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
 275         # by single video API anymore
 276
 277         for chapter in course['Chapters']:
 278             for video in chapter.get('Videos', []):
 279                 if video.get('HasAccess') is False:
 280                     unaccessible_videos += 1
 281                     continue
 282                 video_id = video.get('ID')
 283                 if video_id:
 284                     entries.append({
 285                         '_type': 'url_transparent',
 286                         'url': 'https://www.lynda.com/%s/%s-4.html' % (course_path, video_id),
 287                         'ie_key': LyndaIE.ie_key(),
 288                         'chapter': chapter.get('Title'),
 289                         'chapter_number': int_or_none(chapter.get('ChapterIndex')),
 290                         'chapter_id': compat_str(chapter.get('ID')),
 291                     })
 292
 293         if unaccessible_videos > 0:
 294             self._downloader.report_warning(
 295                 '%s videos are only available for members (or paid members) and will not be downloaded. '
 296                 % unaccessible_videos + self._ACCOUNT_CREDENTIALS_HINT)
 297
 298         course_title = course.get('Title')
 299         course_description = course.get('Description')
 300
 301         return self.playlist_result(entries, course_id, course_title, course_description)