Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/udemy.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_kwargs,
   9     compat_str,
  10     compat_urllib_request,
  11     compat_urlparse,
  12 )
  13 from ..utils import (
  14     determine_ext,
  15     extract_attributes,
  16     ExtractorError,
  17     float_or_none,
  18     int_or_none,
  19     js_to_json,
  20     sanitized_Request,
  21     unescapeHTML,
  22     urlencode_postdata,
  23 )
  24
  25
  26 class UdemyIE(InfoExtractor):
  27     IE_NAME = 'udemy'
  28     _VALID_URL = r'''(?x)
  29                     https?://
  30                         www\.udemy\.com/
  31                         (?:
  32                             [^#]+\#/lecture/|
  33                             lecture/view/?\?lectureId=|
  34                             [^/]+/learn/v4/t/lecture/
  35                         )
  36                         (?P<id>\d+)
  37                     '''
  38     _LOGIN_URL = 'https://www.udemy.com/join/login-popup/?displayType=ajax&showSkipButton=1'
  39     _ORIGIN_URL = 'https://www.udemy.com'
  40     _NETRC_MACHINE = 'udemy'
  41
  42     _TESTS = [{
  43         'url': 'https://www.udemy.com/java-tutorial/#/lecture/172757',
  44         'md5': '98eda5b657e752cf945d8445e261b5c5',
  45         'info_dict': {
  46             'id': '160614',
  47             'ext': 'mp4',
  48             'title': 'Introduction and Installation',
  49             'description': 'md5:c0d51f6f21ef4ec65f091055a5eef876',
  50             'duration': 579.29,
  51         },
  52         'skip': 'Requires udemy account credentials',
  53     }, {
  54         # new URL schema
  55         'url': 'https://www.udemy.com/electric-bass-right-from-the-start/learn/v4/t/lecture/4580906',
  56         'only_matching': True,
  57     }, {
  58         # no url in outputs format entry
  59         'url': 'https://www.udemy.com/learn-web-development-complete-step-by-step-guide-to-success/learn/v4/t/lecture/4125812',
  60         'only_matching': True,
  61     }]
  62
  63     def _extract_course_info(self, webpage, video_id):
  64         course = self._parse_json(
  65             unescapeHTML(self._search_regex(
  66                 r'ng-init=["\'].*\bcourse=({.+?})[;"\']',
  67                 webpage, 'course', default='{}')),
  68             video_id, fatal=False) or {}
  69         course_id = course.get('id') or self._search_regex(
  70             r'data-course-id=["\'](\d+)', webpage, 'course id')
  71         return course_id, course.get('title')
  72
  73     def _enroll_course(self, base_url, webpage, course_id):
  74         def combine_url(base_url, url):
  75             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  76
  77         checkout_url = unescapeHTML(self._search_regex(
  78             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/(?:payment|cart)/checkout/.+?)\1',
  79             webpage, 'checkout url', group='url', default=None))
  80         if checkout_url:
  81             raise ExtractorError(
  82                 'Course %s is not free. You have to pay for it before you can download. '
  83                 'Use this URL to confirm purchase: %s'
  84                 % (course_id, combine_url(base_url, checkout_url)),
  85                 expected=True)
  86
  87         enroll_url = unescapeHTML(self._search_regex(
  88             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  89             webpage, 'enroll url', group='url', default=None))
  90         if enroll_url:
  91             webpage = self._download_webpage(
  92                 combine_url(base_url, enroll_url),
  93                 course_id, 'Enrolling in the course',
  94                 headers={'Referer': base_url})
  95             if '>You have enrolled in' in webpage:
  96                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
  97
  98     def _download_lecture(self, course_id, lecture_id):
  99         return self._download_json(
 100             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
 101             % (course_id, lecture_id),
 102             lecture_id, 'Downloading lecture JSON', query={
 103                 'fields[lecture]': 'title,description,view_html,asset',
 104                 'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,data',
 105             })
 106
 107     def _handle_error(self, response):
 108         if not isinstance(response, dict):
 109             return
 110         error = response.get('error')
 111         if error:
 112             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
 113             error_data = error.get('data')
 114             if error_data:
 115                 error_str += ' - %s' % error_data.get('formErrors')
 116             raise ExtractorError(error_str, expected=True)
 117
 118     def _download_webpage(self, *args, **kwargs):
 119         kwargs.setdefault('headers', {})['User-Agent'] = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_5) AppleWebKit/603.2.4 (KHTML, like Gecko) Version/10.1.1 Safari/603.2.4'
 120         return super(UdemyIE, self)._download_webpage(
 121             *args, **compat_kwargs(kwargs))
 122
 123     def _download_json(self, url_or_request, *args, **kwargs):
 124         headers = {
 125             'X-Udemy-Snail-Case': 'true',
 126             'X-Requested-With': 'XMLHttpRequest',
 127         }
 128         for cookie in self._downloader.cookiejar:
 129             if cookie.name == 'client_id':
 130                 headers['X-Udemy-Client-Id'] = cookie.value
 131             elif cookie.name == 'access_token':
 132                 headers['X-Udemy-Bearer-Token'] = cookie.value
 133                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 134
 135         if isinstance(url_or_request, compat_urllib_request.Request):
 136             for header, value in headers.items():
 137                 url_or_request.add_header(header, value)
 138         else:
 139             url_or_request = sanitized_Request(url_or_request, headers=headers)
 140
 141         response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)
 142         self._handle_error(response)
 143         return response
 144
 145     def _real_initialize(self):
 146         self._login()
 147
 148     def _login(self):
 149         (username, password) = self._get_login_info()
 150         if username is None:
 151             return
 152
 153         login_popup = self._download_webpage(
 154             self._LOGIN_URL, None, 'Downloading login popup')
 155
 156         def is_logged(webpage):
 157             return any(re.search(p, webpage) for p in (
 158                 r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
 159                 r'>Logout<'))
 160
 161         # already logged in
 162         if is_logged(login_popup):
 163             return
 164
 165         login_form = self._form_hidden_inputs('login-form', login_popup)
 166
 167         login_form.update({
 168             'email': username,
 169             'password': password,
 170         })
 171
 172         response = self._download_webpage(
 173             self._LOGIN_URL, None, 'Logging in',
 174             data=urlencode_postdata(login_form),
 175             headers={
 176                 'Referer': self._ORIGIN_URL,
 177                 'Origin': self._ORIGIN_URL,
 178             })
 179
 180         if not is_logged(response):
 181             error = self._html_search_regex(
 182                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 183                 response, 'error message', default=None)
 184             if error:
 185                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 186             raise ExtractorError('Unable to log in')
 187
 188     def _real_extract(self, url):
 189         lecture_id = self._match_id(url)
 190
 191         webpage = self._download_webpage(url, lecture_id)
 192
 193         course_id, _ = self._extract_course_info(webpage, lecture_id)
 194
 195         try:
 196             lecture = self._download_lecture(course_id, lecture_id)
 197         except ExtractorError as e:
 198             # Error could possibly mean we are not enrolled in the course
 199             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 200                 self._enroll_course(url, webpage, course_id)
 201                 lecture = self._download_lecture(course_id, lecture_id)
 202             else:
 203                 raise
 204
 205         title = lecture['title']
 206         description = lecture.get('description')
 207
 208         asset = lecture['asset']
 209
 210         asset_type = asset.get('asset_type') or asset.get('assetType')
 211         if asset_type != 'Video':
 212             raise ExtractorError(
 213                 'Lecture %s is not a video' % lecture_id, expected=True)
 214
 215         stream_url = asset.get('stream_url') or asset.get('streamUrl')
 216         if stream_url:
 217             youtube_url = self._search_regex(
 218                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 219             if youtube_url:
 220                 return self.url_result(youtube_url, 'Youtube')
 221
 222         video_id = compat_str(asset['id'])
 223         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
 224         duration = float_or_none(asset.get('data', {}).get('duration'))
 225
 226         subtitles = {}
 227         automatic_captions = {}
 228
 229         formats = []
 230
 231         def extract_output_format(src, f_id):
 232             return {
 233                 'url': src.get('url'),
 234                 'format_id': '%sp' % (src.get('height') or f_id),
 235                 'width': int_or_none(src.get('width')),
 236                 'height': int_or_none(src.get('height')),
 237                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 238                 'vcodec': src.get('video_codec'),
 239                 'fps': int_or_none(src.get('frame_rate')),
 240                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 241                 'acodec': src.get('audio_codec'),
 242                 'asr': int_or_none(src.get('audio_sample_rate')),
 243                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 244                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 245             }
 246
 247         outputs = asset.get('data', {}).get('outputs')
 248         if not isinstance(outputs, dict):
 249             outputs = {}
 250
 251         def add_output_format_meta(f, key):
 252             output = outputs.get(key)
 253             if isinstance(output, dict):
 254                 output_format = extract_output_format(output, key)
 255                 output_format.update(f)
 256                 return output_format
 257             return f
 258
 259         def extract_formats(source_list):
 260             if not isinstance(source_list, list):
 261                 return
 262             for source in source_list:
 263                 video_url = source.get('file') or source.get('src')
 264                 if not video_url or not isinstance(video_url, compat_str):
 265                     continue
 266                 if source.get('type') == 'application/x-mpegURL' or determine_ext(video_url) == 'm3u8':
 267                     formats.extend(self._extract_m3u8_formats(
 268                         video_url, video_id, 'mp4', entry_protocol='m3u8_native',
 269                         m3u8_id='hls', fatal=False))
 270                     continue
 271                 format_id = source.get('label')
 272                 f = {
 273                     'url': video_url,
 274                     'format_id': '%sp' % format_id,
 275                     'height': int_or_none(format_id),
 276                 }
 277                 if format_id:
 278                     # Some videos contain additional metadata (e.g.
 279                     # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 280                     f = add_output_format_meta(f, format_id)
 281                 formats.append(f)
 282
 283         def extract_subtitles(track_list):
 284             if not isinstance(track_list, list):
 285                 return
 286             for track in track_list:
 287                 if not isinstance(track, dict):
 288                     continue
 289                 if track.get('kind') != 'captions':
 290                     continue
 291                 src = track.get('src')
 292                 if not src or not isinstance(src, compat_str):
 293                     continue
 294                 lang = track.get('language') or track.get(
 295                     'srclang') or track.get('label')
 296                 sub_dict = automatic_captions if track.get(
 297                     'autogenerated') is True else subtitles
 298                 sub_dict.setdefault(lang, []).append({
 299                     'url': src,
 300                 })
 301
 302         download_urls = asset.get('download_urls')
 303         if isinstance(download_urls, dict):
 304             extract_formats(download_urls.get('Video'))
 305
 306         view_html = lecture.get('view_html')
 307         if view_html:
 308             view_html_urls = set()
 309             for source in re.findall(r'<source[^>]+>', view_html):
 310                 attributes = extract_attributes(source)
 311                 src = attributes.get('src')
 312                 if not src:
 313                     continue
 314                 res = attributes.get('data-res')
 315                 height = int_or_none(res)
 316                 if src in view_html_urls:
 317                     continue
 318                 view_html_urls.add(src)
 319                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 320                     m3u8_formats = self._extract_m3u8_formats(
 321                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 322                         m3u8_id='hls', fatal=False)
 323                     for f in m3u8_formats:
 324                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 325                         if m:
 326                             if not f.get('height'):
 327                                 f['height'] = int(m.group('height'))
 328                             if not f.get('tbr'):
 329                                 f['tbr'] = int(m.group('tbr'))
 330                     formats.extend(m3u8_formats)
 331                 else:
 332                     formats.append(add_output_format_meta({
 333                         'url': src,
 334                         'format_id': '%dp' % height if height else None,
 335                         'height': height,
 336                     }, res))
 337
 338             # react rendition since 2017.04.15 (see
 339             # https://github.com/rg3/youtube-dl/issues/12744)
 340             data = self._parse_json(
 341                 self._search_regex(
 342                     r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
 343                     'setup data', default='{}', group='data'), video_id,
 344                 transform_source=unescapeHTML, fatal=False)
 345             if data and isinstance(data, dict):
 346                 extract_formats(data.get('sources'))
 347                 if not duration:
 348                     duration = int_or_none(data.get('duration'))
 349                 extract_subtitles(data.get('tracks'))
 350
 351             if not subtitles and not automatic_captions:
 352                 text_tracks = self._parse_json(
 353                     self._search_regex(
 354                         r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
 355                         'text tracks', default='{}', group='data'), video_id,
 356                     transform_source=lambda s: js_to_json(unescapeHTML(s)),
 357                     fatal=False)
 358                 extract_subtitles(text_tracks)
 359
 360         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 361
 362         return {
 363             'id': video_id,
 364             'title': title,
 365             'description': description,
 366             'thumbnail': thumbnail,
 367             'duration': duration,
 368             'formats': formats,
 369             'subtitles': subtitles,
 370             'automatic_captions': automatic_captions,
 371         }
 372
 373
 374 class UdemyCourseIE(UdemyIE):
 375     IE_NAME = 'udemy:course'
 376     _VALID_URL = r'https?://(?:www\.)?udemy\.com/(?P<id>[^/?#&]+)'
 377     _TESTS = []
 378
 379     @classmethod
 380     def suitable(cls, url):
 381         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 382
 383     def _real_extract(self, url):
 384         course_path = self._match_id(url)
 385
 386         webpage = self._download_webpage(url, course_path)
 387
 388         course_id, title = self._extract_course_info(webpage, course_path)
 389
 390         self._enroll_course(url, webpage, course_id)
 391
 392         response = self._download_json(
 393             'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
 394             course_id, 'Downloading course curriculum', query={
 395                 'fields[chapter]': 'title,object_index',
 396                 'fields[lecture]': 'title,asset',
 397                 'page_size': '1000',
 398             })
 399
 400         entries = []
 401         chapter, chapter_number = [None] * 2
 402         for entry in response['results']:
 403             clazz = entry.get('_class')
 404             if clazz == 'lecture':
 405                 asset = entry.get('asset')
 406                 if isinstance(asset, dict):
 407                     asset_type = asset.get('asset_type') or asset.get('assetType')
 408                     if asset_type != 'Video':
 409                         continue
 410                 lecture_id = entry.get('id')
 411                 if lecture_id:
 412                     entry = {
 413                         '_type': 'url_transparent',
 414                         'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
 415                         'title': entry.get('title'),
 416                         'ie_key': UdemyIE.ie_key(),
 417                     }
 418                     if chapter_number:
 419                         entry['chapter_number'] = chapter_number
 420                     if chapter:
 421                         entry['chapter'] = chapter
 422                     entries.append(entry)
 423             elif clazz == 'chapter':
 424                 chapter_number = entry.get('object_index')
 425                 chapter = entry.get('title')
 426
 427         return self.playlist_result(entries, course_id, title)