_ Git - youtube-dl/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         response = self._download_json(
  48             action_url, None, note,
  49             data=urlencode_postdata(form_data),
  50             headers={
  51                 'Referer': referrer_url,
  52                 'X-Requested-With': 'XMLHttpRequest',
  53             }, expected_status=(418, 500, ))
  54
  55         self._check_error(response, ('email', 'password', 'ErrorMessage'))
  56
  57         return response, action_url
  58
  59     def _login(self):
  60         username, password = self._get_login_info()
  61         if username is None:
  62             return
  63
  64         # Step 1: download signin page
  65         signin_page = self._download_webpage(
  66             self._SIGNIN_URL, None, 'Downloading signin page')
  67
  68         # Already logged in
  69         if any(re.search(p, signin_page) for p in (
  70                 r'isLoggedIn\s*:\s*true', r'logout\.aspx', r'>Log out<')):
  71             return
  72
  73         # Step 2: submit email
  74         signin_form = self._search_regex(
  75             r'(?s)(<form[^>]+data-form-name=["\']signin["\'][^>]*>.+?</form>)',
  76             signin_page, 'signin form')
  77         signin_page, signin_url = self._login_step(
  78             signin_form, self._PASSWORD_URL, {'email': username},
  79             'Submitting email', self._SIGNIN_URL)
  80
  81         # Step 3: submit password
  82         password_form = signin_page['body']
  83         self._login_step(
  84             password_form, self._USER_URL, {'email': username, 'password': password},
  85             'Submitting password', signin_url)
  86
  87
  88 class LyndaIE(LyndaBaseIE):
  89     IE_NAME = 'lynda'
  90     IE_DESC = 'lynda.com videos'
  91     _VALID_URL = r'''(?x)
  92                     https?://
  93                         (?:www\.)?(?:lynda\.com|educourse\.ga)/
  94                         (?:
  95                             (?:[^/]+/){2,3}(?P<course_id>\d+)|
  96                             player/embed
  97                         )/
  98                         (?P<id>\d+)
  99                     '''
 100
 101     _TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
 102
 103     _TESTS = [{
 104         'url': 'https://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
 105         # md5 is unstable
 106         'info_dict': {
 107             'id': '114408',
 108             'ext': 'mp4',
 109             'title': 'Using the exercise files',
 110             'duration': 68
 111         }
 112     }, {
 113         'url': 'https://www.lynda.com/player/embed/133770?tr=foo=1;bar=g;fizz=rt&fs=0',
 114         'only_matching': True,
 115     }, {
 116         'url': 'https://educourse.ga/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
 117         'only_matching': True,
 118     }, {
 119         'url': 'https://www.lynda.com/de/Graphic-Design-tutorials/Willkommen-Grundlagen-guten-Gestaltung/393570/393572-4.html',
 120         'only_matching': True,
 121     }]
 122
 123     def _raise_unavailable(self, video_id):
 124         self.raise_login_required(
 125             'Video %s is only available for members' % video_id)
 126
 127     def _real_extract(self, url):
 128         mobj = re.match(self._VALID_URL, url)
 129         video_id = mobj.group('id')
 130         course_id = mobj.group('course_id')
 131
 132         query = {
 133             'videoId': video_id,
 134             'type': 'video',
 135         }
 136
 137         video = self._download_json(
 138             'https://www.lynda.com/ajax/player', video_id,
 139             'Downloading video JSON', fatal=False, query=query)
 140
 141         # Fallback scenario
 142         if not video:
 143             query['courseId'] = course_id
 144
 145             play = self._download_json(
 146                 'https://www.lynda.com/ajax/course/%s/%s/play'
 147                 % (course_id, video_id), video_id, 'Downloading play JSON')
 148
 149             if not play:
 150                 self._raise_unavailable(video_id)
 151
 152             formats = []
 153             for formats_dict in play:
 154                 urls = formats_dict.get('urls')
 155                 if not isinstance(urls, dict):
 156                     continue
 157                 cdn = formats_dict.get('name')
 158                 for format_id, format_url in urls.items():
 159                     if not format_url:
 160                         continue
 161                     formats.append({
 162                         'url': format_url,
 163                         'format_id': '%s-%s' % (cdn, format_id) if cdn else format_id,
 164                         'height': int_or_none(format_id),
 165                     })
 166             self._sort_formats(formats)
 167
 168             conviva = self._download_json(
 169                 'https://www.lynda.com/ajax/player/conviva', video_id,
 170                 'Downloading conviva JSON', query=query)
 171
 172             return {
 173                 'id': video_id,
 174                 'title': conviva['VideoTitle'],
 175                 'description': conviva.get('VideoDescription'),
 176                 'release_year': int_or_none(conviva.get('ReleaseYear')),
 177                 'duration': int_or_none(conviva.get('Duration')),
 178                 'creator': conviva.get('Author'),
 179                 'formats': formats,
 180             }
 181
 182         if 'Status' in video:
 183             raise ExtractorError(
 184                 'lynda returned error: %s' % video['Message'], expected=True)
 185
 186         if video.get('HasAccess') is False:
 187             self._raise_unavailable(video_id)
 188
 189         video_id = compat_str(video.get('ID') or video_id)
 190         duration = int_or_none(video.get('DurationInSeconds'))
 191         title = video['Title']
 192
 193         formats = []
 194
 195         fmts = video.get('Formats')
 196         if fmts:
 197             formats.extend([{
 198                 'url': f['Url'],
 199                 'ext': f.get('Extension'),
 200                 'width': int_or_none(f.get('Width')),
 201                 'height': int_or_none(f.get('Height')),
 202                 'filesize': int_or_none(f.get('FileSize')),
 203                 'format_id': compat_str(f.get('Resolution')) if f.get('Resolution') else None,
 204             } for f in fmts if f.get('Url')])
 205
 206         prioritized_streams = video.get('PrioritizedStreams')
 207         if prioritized_streams:
 208             for prioritized_stream_id, prioritized_stream in prioritized_streams.items():
 209                 formats.extend([{
 210                     'url': video_url,
 211                     'height': int_or_none(format_id),
 212                     'format_id': '%s-%s' % (prioritized_stream_id, format_id),
 213                 } for format_id, video_url in prioritized_stream.items()])
 214
 215         self._check_formats(formats, video_id)
 216         self._sort_formats(formats)
 217
 218         subtitles = self.extract_subtitles(video_id)
 219
 220         return {
 221             'id': video_id,
 222             'title': title,
 223             'duration': duration,
 224             'subtitles': subtitles,
 225             'formats': formats
 226         }
 227
 228     def _fix_subtitles(self, subs):
 229         srt = ''
 230         seq_counter = 0
 231         for pos in range(0, len(subs) - 1):
 232             seq_current = subs[pos]
 233             m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
 234             if m_current is None:
 235                 continue
 236             seq_next = subs[pos + 1]
 237             m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
 238             if m_next is None:
 239                 continue
 240             appear_time = m_current.group('timecode')
 241             disappear_time = m_next.group('timecode')
 242             text = seq_current['Caption'].strip()
 243             if text:
 244                 seq_counter += 1
 245                 srt += '%s\r\n%s --> %s\r\n%s\r\n\r\n' % (seq_counter, appear_time, disappear_time, text)
 246         if srt:
 247             return srt
 248
 249     def _get_subtitles(self, video_id):
 250         url = 'https://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
 251         subs = self._download_json(url, None, False)
 252         fixed_subs = self._fix_subtitles(subs)
 253         if fixed_subs:
 254             return {'en': [{'ext': 'srt', 'data': fixed_subs}]}
 255         else:
 256             return {}
 257
 258
 259 class LyndaCourseIE(LyndaBaseIE):
 260     IE_NAME = 'lynda:course'
 261     IE_DESC = 'lynda.com online courses'
 262
 263     # Course link equals to welcome/introduction video link of same course
 264     # We will recognize it as course link
 265     _VALID_URL = r'https?://(?:www|m)\.(?:lynda\.com|educourse\.ga)/(?P<coursepath>(?:[^/]+/){2,3}(?P<courseid>\d+))-2\.html'
 266
 267     _TESTS = [{
 268         'url': 'https://www.lynda.com/Graphic-Design-tutorials/Grundlagen-guten-Gestaltung/393570-2.html',
 269         'only_matching': True,
 270     }, {
 271         'url': 'https://www.lynda.com/de/Graphic-Design-tutorials/Grundlagen-guten-Gestaltung/393570-2.html',
 272         'only_matching': True,
 273     }]
 274
 275     def _real_extract(self, url):
 276         mobj = re.match(self._VALID_URL, url)
 277         course_path = mobj.group('coursepath')
 278         course_id = mobj.group('courseid')
 279
 280         item_template = 'https://www.lynda.com/%s/%%s-4.html' % course_path
 281
 282         course = self._download_json(
 283             'https://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
 284             course_id, 'Downloading course JSON', fatal=False)
 285
 286         if not course:
 287             webpage = self._download_webpage(url, course_id)
 288             entries = [
 289                 self.url_result(
 290                     item_template % video_id, ie=LyndaIE.ie_key(),
 291                     video_id=video_id)
 292                 for video_id in re.findall(
 293                     r'data-video-id=["\'](\d+)', webpage)]
 294             return self.playlist_result(
 295                 entries, course_id,
 296                 self._og_search_title(webpage, fatal=False),
 297                 self._og_search_description(webpage))
 298
 299         if course.get('Status') == 'NotFound':
 300             raise ExtractorError(
 301                 'Course %s does not exist' % course_id, expected=True)
 302
 303         unaccessible_videos = 0
 304         entries = []
 305
 306         # Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
 307         # by single video API anymore
 308
 309         for chapter in course['Chapters']:
 310             for video in chapter.get('Videos', []):
 311                 if video.get('HasAccess') is False:
 312                     unaccessible_videos += 1
 313                     continue
 314                 video_id = video.get('ID')
 315                 if video_id:
 316                     entries.append({
 317                         '_type': 'url_transparent',
 318                         'url': item_template % video_id,
 319                         'ie_key': LyndaIE.ie_key(),
 320                         'chapter': chapter.get('Title'),
 321                         'chapter_number': int_or_none(chapter.get('ChapterIndex')),
 322                         'chapter_id': compat_str(chapter.get('ID')),
 323                     })
 324
 325         if unaccessible_videos > 0:
 326             self._downloader.report_warning(
 327                 '%s videos are only available for members (or paid members) and will not be downloaded. '
 328                 % unaccessible_videos + self._ACCOUNT_CREDENTIALS_HINT)
 329
 330         course_title = course.get('Title')
 331         course_description = course.get('Description')
 332
 333         return self.playlist_result(entries, course_id, course_title, course_description)