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                 r'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'''(?x)
  98                     https?://
  99                         (?:www\.)?(?:lynda\.com|educourse\.ga)/
 100                         (?:
 101                             (?:[^/]+/){2,3}(?P<course_id>\d+)|
 102                             player/embed
 103                         )/
 104                         (?P<id>\d+)
 105                     '''
 106
 107     _TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
 108
 109     _TESTS = [{
 110         'url': 'https://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
 111         # md5 is unstable
 112         'info_dict': {
 113             'id': '114408',
 114             'ext': 'mp4',
 115             'title': 'Using the exercise files',
 116             'duration': 68
 117         }
 118     }, {
 119         'url': 'https://www.lynda.com/player/embed/133770?tr=foo=1;bar=g;fizz=rt&fs=0',
 120         'only_matching': True,
 121     }, {
 122         'url': 'https://educourse.ga/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
 123         'only_matching': True,
 124     }, {
 125         'url': 'https://www.lynda.com/de/Graphic-Design-tutorials/Willkommen-Grundlagen-guten-Gestaltung/393570/393572-4.html',
 126         'only_matching': True,
 127     }]
 128
 129     def _raise_unavailable(self, video_id):
 130         self.raise_login_required(
 131             'Video %s is only available for members' % video_id)
 132
 133     def _real_extract(self, url):
 134         mobj = re.match(self._VALID_URL, url)
 135         video_id = mobj.group('id')
 136         course_id = mobj.group('course_id')
 137
 138         query = {
 139             'videoId': video_id,
 140             'type': 'video',
 141         }
 142
 143         video = self._download_json(
 144             'https://www.lynda.com/ajax/player', video_id,
 145             'Downloading video JSON', fatal=False, query=query)
 146
 147         # Fallback scenario
 148         if not video:
 149             query['courseId'] = course_id
 150
 151             play = self._download_json(
 152                 'https://www.lynda.com/ajax/course/%s/%s/play'
 153                 % (course_id, video_id), video_id, 'Downloading play JSON')
 154
 155             if not play:
 156                 self._raise_unavailable(video_id)
 157
 158             formats = []
 159             for formats_dict in play:
 160                 urls = formats_dict.get('urls')
 161                 if not isinstance(urls, dict):
 162                     continue
 163                 cdn = formats_dict.get('name')
 164                 for format_id, format_url in urls.items():
 165                     if not format_url:
 166                         continue
 167                     formats.append({
 168                         'url': format_url,
 169                         'format_id': '%s-%s' % (cdn, format_id) if cdn else format_id,
 170                         'height': int_or_none(format_id),
 171                     })
 172             self._sort_formats(formats)
 173
 174             conviva = self._download_json(
 175                 'https://www.lynda.com/ajax/player/conviva', video_id,
 176                 'Downloading conviva JSON', query=query)
 177
 178             return {
 179                 'id': video_id,
 180                 'title': conviva['VideoTitle'],
 181                 'description': conviva.get('VideoDescription'),
 182                 'release_year': int_or_none(conviva.get('ReleaseYear')),
 183                 'duration': int_or_none(conviva.get('Duration')),
 184                 'creator': conviva.get('Author'),
 185                 'formats': formats,
 186             }
 187
 188         if 'Status' in video:
 189             raise ExtractorError(
 190                 'lynda returned error: %s' % video['Message'], expected=True)
 191
 192         if video.get('HasAccess') is False:
 193             self._raise_unavailable(video_id)
 194
 195         video_id = compat_str(video.get('ID') or video_id)
 196         duration = int_or_none(video.get('DurationInSeconds'))
 197         title = video['Title']
 198
 199         formats = []
 200
 201         fmts = video.get('Formats')
 202         if fmts:
 203             formats.extend([{
 204                 'url': f['Url'],
 205                 'ext': f.get('Extension'),
 206                 'width': int_or_none(f.get('Width')),
 207                 'height': int_or_none(f.get('Height')),
 208                 'filesize': int_or_none(f.get('FileSize')),
 209                 'format_id': compat_str(f.get('Resolution')) if f.get('Resolution') else None,
 210             } for f in fmts if f.get('Url')])
 211
 212         prioritized_streams = video.get('PrioritizedStreams')
 213         if prioritized_streams:
 214             for prioritized_stream_id, prioritized_stream in prioritized_streams.items():
 215                 formats.extend([{
 216                     'url': video_url,
 217                     'height': int_or_none(format_id),
 218                     'format_id': '%s-%s' % (prioritized_stream_id, format_id),
 219                 } for format_id, video_url in prioritized_stream.items()])
 220
 221         self._check_formats(formats, video_id)
 222         self._sort_formats(formats)
 223
 224         subtitles = self.extract_subtitles(video_id)
 225
 226         return {
 227             'id': video_id,
 228             'title': title,
 229             'duration': duration,
 230             'subtitles': subtitles,
 231             'formats': formats
 232         }
 233
 234     def _fix_subtitles(self, subs):
 235         srt = ''
 236         seq_counter = 0
 237         for pos in range(0, len(subs) - 1):
 238             seq_current = subs[pos]
 239             m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
 240             if m_current is None:
 241                 continue
 242             seq_next = subs[pos + 1]
 243             m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
 244             if m_next is None:
 245                 continue
 246             appear_time = m_current.group('timecode')
 247             disappear_time = m_next.group('timecode')
 248             text = seq_current['Caption'].strip()
 249             if text:
 250                 seq_counter += 1
 251                 srt += '%s\r\n%s --> %s\r\n%s\r\n\r\n' % (seq_counter, appear_time, disappear_time, text)
 252         if srt:
 253             return srt
 254
 255     def _get_subtitles(self, video_id):
 256         url = 'https://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
 257         subs = self._download_json(url, None, False)
 258         fixed_subs = self._fix_subtitles(subs)
 259         if fixed_subs:
 260             return {'en': [{'ext': 'srt', 'data': fixed_subs}]}
 261         else:
 262             return {}
 263
 264
 265 class LyndaCourseIE(LyndaBaseIE):
 266     IE_NAME = 'lynda:course'
 267     IE_DESC = 'lynda.com online courses'
 268
 269     # Course link equals to welcome/introduction video link of same course
 270     # We will recognize it as course link
 271     _VALID_URL = r'https?://(?:www|m)\.(?:lynda\.com|educourse\.ga)/(?P<coursepath>(?:[^/]+/){2,3}(?P<courseid>\d+))-2\.html'
 272
 273     _TESTS = [{
 274         'url': 'https://www.lynda.com/Graphic-Design-tutorials/Grundlagen-guten-Gestaltung/393570-2.html',
 275         'only_matching': True,
 276     }, {
 277         'url': 'https://www.lynda.com/de/Graphic-Design-tutorials/Grundlagen-guten-Gestaltung/393570-2.html',
 278         'only_matching': True,
 279     }]
 280
 281     def _real_extract(self, url):
 282         mobj = re.match(self._VALID_URL, url)
 283         course_path = mobj.group('coursepath')
 284         course_id = mobj.group('courseid')
 285
 286         item_template = 'https://www.lynda.com/%s/%%s-4.html' % course_path
 287
 288         course = self._download_json(
 289             'https://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
 290             course_id, 'Downloading course JSON', fatal=False)
 291
 292         if not course:
 293             webpage = self._download_webpage(url, course_id)
 294             entries = [
 295                 self.url_result(
 296                     item_template % video_id, ie=LyndaIE.ie_key(),
 297                     video_id=video_id)
 298                 for video_id in re.findall(
 299                     r'data-video-id=["\'](\d+)', webpage)]
 300             return self.playlist_result(
 301                 entries, course_id,
 302                 self._og_search_title(webpage, fatal=False),
 303                 self._og_search_description(webpage))
 304
 305         if course.get('Status') == 'NotFound':
 306             raise ExtractorError(
 307                 'Course %s does not exist' % course_id, expected=True)
 308
 309         unaccessible_videos = 0
 310         entries = []
 311
 312         # Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
 313         # by single video API anymore
 314
 315         for chapter in course['Chapters']:
 316             for video in chapter.get('Videos', []):
 317                 if video.get('HasAccess') is False:
 318                     unaccessible_videos += 1
 319                     continue
 320                 video_id = video.get('ID')
 321                 if video_id:
 322                     entries.append({
 323                         '_type': 'url_transparent',
 324                         'url': item_template % video_id,
 325                         'ie_key': LyndaIE.ie_key(),
 326                         'chapter': chapter.get('Title'),
 327                         'chapter_number': int_or_none(chapter.get('ChapterIndex')),
 328                         'chapter_id': compat_str(chapter.get('ID')),
 329                     })
 330
 331         if unaccessible_videos > 0:
 332             self._downloader.report_warning(
 333                 '%s videos are only available for members (or paid members) and will not be downloaded. '
 334                 % unaccessible_videos + self._ACCOUNT_CREDENTIALS_HINT)
 335
 336         course_title = course.get('Title')
 337         course_description = course.get('Description')
 338
 339         return self.playlist_result(entries, course_id, course_title, course_description)