yt_dlp/extractor/udemy.py

   1 import re
   2 import urllib.request
   3
   4 from .common import InfoExtractor
   5 from ..compat import compat_HTTPError, compat_str, compat_urlparse
   6 from ..utils import (
   7     ExtractorError,
   8     determine_ext,
   9     extract_attributes,
  10     float_or_none,
  11     int_or_none,
  12     js_to_json,
  13     sanitized_Request,
  14     try_get,
  15     unescapeHTML,
  16     url_or_none,
  17     urlencode_postdata,
  18 )
  19
  20
  21 class UdemyIE(InfoExtractor):
  22     IE_NAME = 'udemy'
  23     _VALID_URL = r'''(?x)
  24                     https?://
  25                         (?:[^/]+\.)?udemy\.com/
  26                         (?:
  27                             [^#]+\#/lecture/|
  28                             lecture/view/?\?lectureId=|
  29                             [^/]+/learn/v4/t/lecture/
  30                         )
  31                         (?P<id>\d+)
  32                     '''
  33     _LOGIN_URL = 'https://www.udemy.com/join/login-popup/?displayType=ajax&showSkipButton=1'
  34     _ORIGIN_URL = 'https://www.udemy.com'
  35     _NETRC_MACHINE = 'udemy'
  36
  37     _TESTS = [{
  38         'url': 'https://www.udemy.com/java-tutorial/#/lecture/172757',
  39         'md5': '98eda5b657e752cf945d8445e261b5c5',
  40         'info_dict': {
  41             'id': '160614',
  42             'ext': 'mp4',
  43             'title': 'Introduction and Installation',
  44             'description': 'md5:c0d51f6f21ef4ec65f091055a5eef876',
  45             'duration': 579.29,
  46         },
  47         'skip': 'Requires udemy account credentials',
  48     }, {
  49         # new URL schema
  50         'url': 'https://www.udemy.com/electric-bass-right-from-the-start/learn/v4/t/lecture/4580906',
  51         'only_matching': True,
  52     }, {
  53         # no url in outputs format entry
  54         'url': 'https://www.udemy.com/learn-web-development-complete-step-by-step-guide-to-success/learn/v4/t/lecture/4125812',
  55         'only_matching': True,
  56     }, {
  57         # only outputs rendition
  58         'url': 'https://www.udemy.com/how-you-can-help-your-local-community-5-amazing-examples/learn/v4/t/lecture/3225750?start=0',
  59         'only_matching': True,
  60     }, {
  61         'url': 'https://wipro.udemy.com/java-tutorial/#/lecture/172757',
  62         'only_matching': True,
  63     }]
  64
  65     def _extract_course_info(self, webpage, video_id):
  66         course = self._parse_json(
  67             unescapeHTML(self._search_regex(
  68                 r'ng-init=["\'].*\bcourse=({.+?})[;"\']',
  69                 webpage, 'course', default='{}')),
  70             video_id, fatal=False) or {}
  71         course_id = course.get('id') or self._search_regex(
  72             [
  73                 r'data-course-id=["\'](\d+)',
  74                 r'&quot;courseId&quot;\s*:\s*(\d+)'
  75             ], webpage, 'course id')
  76         return course_id, course.get('title')
  77
  78     def _enroll_course(self, base_url, webpage, course_id):
  79         def combine_url(base_url, url):
  80             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  81
  82         checkout_url = unescapeHTML(self._search_regex(
  83             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/(?:payment|cart)/checkout/.+?)\1',
  84             webpage, 'checkout url', group='url', default=None))
  85         if checkout_url:
  86             raise ExtractorError(
  87                 'Course %s is not free. You have to pay for it before you can download. '
  88                 'Use this URL to confirm purchase: %s'
  89                 % (course_id, combine_url(base_url, checkout_url)),
  90                 expected=True)
  91
  92         enroll_url = unescapeHTML(self._search_regex(
  93             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  94             webpage, 'enroll url', group='url', default=None))
  95         if enroll_url:
  96             webpage = self._download_webpage(
  97                 combine_url(base_url, enroll_url),
  98                 course_id, 'Enrolling in the course',
  99                 headers={'Referer': base_url})
 100             if '>You have enrolled in' in webpage:
 101                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
 102
 103     def _download_lecture(self, course_id, lecture_id):
 104         return self._download_json(
 105             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
 106             % (course_id, lecture_id),
 107             lecture_id, 'Downloading lecture JSON', query={
 108                 'fields[lecture]': 'title,description,view_html,asset',
 109                 'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,stream_urls,captions,data',
 110             })
 111
 112     def _handle_error(self, response):
 113         if not isinstance(response, dict):
 114             return
 115         error = response.get('error')
 116         if error:
 117             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
 118             error_data = error.get('data')
 119             if error_data:
 120                 error_str += ' - %s' % error_data.get('formErrors')
 121             raise ExtractorError(error_str, expected=True)
 122
 123     def _download_webpage_handle(self, *args, **kwargs):
 124         headers = kwargs.get('headers', {}).copy()
 125         headers['User-Agent'] = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/72.0.3626.109 Safari/537.36'
 126         kwargs['headers'] = headers
 127         ret = super(UdemyIE, self)._download_webpage_handle(
 128             *args, **kwargs)
 129         if not ret:
 130             return ret
 131         webpage, _ = ret
 132         if any(p in webpage for p in (
 133                 '>Please verify you are a human',
 134                 'Access to this page has been denied because we believe you are using automation tools to browse the website',
 135                 '"_pxCaptcha"')):
 136             raise ExtractorError(
 137                 'Udemy asks you to solve a CAPTCHA. Login with browser, '
 138                 'solve CAPTCHA, then export cookies and pass cookie file to '
 139                 'yt-dlp with --cookies.', expected=True)
 140         return ret
 141
 142     def _download_json(self, url_or_request, *args, **kwargs):
 143         headers = {
 144             'X-Udemy-Snail-Case': 'true',
 145             'X-Requested-With': 'XMLHttpRequest',
 146         }
 147         for cookie in self.cookiejar:
 148             if cookie.name == 'client_id':
 149                 headers['X-Udemy-Client-Id'] = cookie.value
 150             elif cookie.name == 'access_token':
 151                 headers['X-Udemy-Bearer-Token'] = cookie.value
 152                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 153
 154         if isinstance(url_or_request, urllib.request.Request):
 155             for header, value in headers.items():
 156                 url_or_request.add_header(header, value)
 157         else:
 158             url_or_request = sanitized_Request(url_or_request, headers=headers)
 159
 160         response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)
 161         self._handle_error(response)
 162         return response
 163
 164     def _perform_login(self, username, password):
 165         login_popup = self._download_webpage(
 166             self._LOGIN_URL, None, 'Downloading login popup')
 167
 168         def is_logged(webpage):
 169             return any(re.search(p, webpage) for p in (
 170                 r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
 171                 r'>Logout<'))
 172
 173         # already logged in
 174         if is_logged(login_popup):
 175             return
 176
 177         login_form = self._form_hidden_inputs('login-form', login_popup)
 178
 179         login_form.update({
 180             'email': username,
 181             'password': password,
 182         })
 183
 184         response = self._download_webpage(
 185             self._LOGIN_URL, None, 'Logging in',
 186             data=urlencode_postdata(login_form),
 187             headers={
 188                 'Referer': self._ORIGIN_URL,
 189                 'Origin': self._ORIGIN_URL,
 190             })
 191
 192         if not is_logged(response):
 193             error = self._html_search_regex(
 194                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 195                 response, 'error message', default=None)
 196             if error:
 197                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 198             raise ExtractorError('Unable to log in')
 199
 200     def _real_extract(self, url):
 201         lecture_id = self._match_id(url)
 202
 203         webpage = self._download_webpage(url, lecture_id)
 204
 205         course_id, _ = self._extract_course_info(webpage, lecture_id)
 206
 207         try:
 208             lecture = self._download_lecture(course_id, lecture_id)
 209         except ExtractorError as e:
 210             # Error could possibly mean we are not enrolled in the course
 211             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 212                 self._enroll_course(url, webpage, course_id)
 213                 lecture = self._download_lecture(course_id, lecture_id)
 214             else:
 215                 raise
 216
 217         title = lecture['title']
 218         description = lecture.get('description')
 219
 220         asset = lecture['asset']
 221
 222         asset_type = asset.get('asset_type') or asset.get('assetType')
 223         if asset_type != 'Video':
 224             raise ExtractorError(
 225                 'Lecture %s is not a video' % lecture_id, expected=True)
 226
 227         stream_url = asset.get('stream_url') or asset.get('streamUrl')
 228         if stream_url:
 229             youtube_url = self._search_regex(
 230                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 231             if youtube_url:
 232                 return self.url_result(youtube_url, 'Youtube')
 233
 234         video_id = compat_str(asset['id'])
 235         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
 236         duration = float_or_none(asset.get('data', {}).get('duration'))
 237
 238         subtitles = {}
 239         automatic_captions = {}
 240
 241         formats = []
 242
 243         def extract_output_format(src, f_id):
 244             return {
 245                 'url': src.get('url'),
 246                 'format_id': '%sp' % (src.get('height') or f_id),
 247                 'width': int_or_none(src.get('width')),
 248                 'height': int_or_none(src.get('height')),
 249                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 250                 'vcodec': src.get('video_codec'),
 251                 'fps': int_or_none(src.get('frame_rate')),
 252                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 253                 'acodec': src.get('audio_codec'),
 254                 'asr': int_or_none(src.get('audio_sample_rate')),
 255                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 256                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 257             }
 258
 259         outputs = asset.get('data', {}).get('outputs')
 260         if not isinstance(outputs, dict):
 261             outputs = {}
 262
 263         def add_output_format_meta(f, key):
 264             output = outputs.get(key)
 265             if isinstance(output, dict):
 266                 output_format = extract_output_format(output, key)
 267                 output_format.update(f)
 268                 return output_format
 269             return f
 270
 271         def extract_formats(source_list):
 272             if not isinstance(source_list, list):
 273                 return
 274             for source in source_list:
 275                 video_url = url_or_none(source.get('file') or source.get('src'))
 276                 if not video_url:
 277                     continue
 278                 if source.get('type') == 'application/x-mpegURL' or determine_ext(video_url) == 'm3u8':
 279                     formats.extend(self._extract_m3u8_formats(
 280                         video_url, video_id, 'mp4', entry_protocol='m3u8_native',
 281                         m3u8_id='hls', fatal=False))
 282                     continue
 283                 format_id = source.get('label')
 284                 f = {
 285                     'url': video_url,
 286                     'format_id': '%sp' % format_id,
 287                     'height': int_or_none(format_id),
 288                 }
 289                 if format_id:
 290                     # Some videos contain additional metadata (e.g.
 291                     # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 292                     f = add_output_format_meta(f, format_id)
 293                 formats.append(f)
 294
 295         def extract_subtitles(track_list):
 296             if not isinstance(track_list, list):
 297                 return
 298             for track in track_list:
 299                 if not isinstance(track, dict):
 300                     continue
 301                 if track.get('kind') != 'captions':
 302                     continue
 303                 src = url_or_none(track.get('src'))
 304                 if not src:
 305                     continue
 306                 lang = track.get('language') or track.get(
 307                     'srclang') or track.get('label')
 308                 sub_dict = automatic_captions if track.get(
 309                     'autogenerated') is True else subtitles
 310                 sub_dict.setdefault(lang, []).append({
 311                     'url': src,
 312                 })
 313
 314         for url_kind in ('download', 'stream'):
 315             urls = asset.get('%s_urls' % url_kind)
 316             if isinstance(urls, dict):
 317                 extract_formats(urls.get('Video'))
 318
 319         captions = asset.get('captions')
 320         if isinstance(captions, list):
 321             for cc in captions:
 322                 if not isinstance(cc, dict):
 323                     continue
 324                 cc_url = url_or_none(cc.get('url'))
 325                 if not cc_url:
 326                     continue
 327                 lang = try_get(cc, lambda x: x['locale']['locale'], compat_str)
 328                 sub_dict = (automatic_captions if cc.get('source') == 'auto'
 329                             else subtitles)
 330                 sub_dict.setdefault(lang or 'en', []).append({
 331                     'url': cc_url,
 332                 })
 333
 334         view_html = lecture.get('view_html')
 335         if view_html:
 336             view_html_urls = set()
 337             for source in re.findall(r'<source[^>]+>', view_html):
 338                 attributes = extract_attributes(source)
 339                 src = attributes.get('src')
 340                 if not src:
 341                     continue
 342                 res = attributes.get('data-res')
 343                 height = int_or_none(res)
 344                 if src in view_html_urls:
 345                     continue
 346                 view_html_urls.add(src)
 347                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 348                     m3u8_formats = self._extract_m3u8_formats(
 349                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 350                         m3u8_id='hls', fatal=False)
 351                     for f in m3u8_formats:
 352                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 353                         if m:
 354                             if not f.get('height'):
 355                                 f['height'] = int(m.group('height'))
 356                             if not f.get('tbr'):
 357                                 f['tbr'] = int(m.group('tbr'))
 358                     formats.extend(m3u8_formats)
 359                 else:
 360                     formats.append(add_output_format_meta({
 361                         'url': src,
 362                         'format_id': '%dp' % height if height else None,
 363                         'height': height,
 364                     }, res))
 365
 366             # react rendition since 2017.04.15 (see
 367             # https://github.com/ytdl-org/youtube-dl/issues/12744)
 368             data = self._parse_json(
 369                 self._search_regex(
 370                     r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
 371                     'setup data', default='{}', group='data'), video_id,
 372                 transform_source=unescapeHTML, fatal=False)
 373             if data and isinstance(data, dict):
 374                 extract_formats(data.get('sources'))
 375                 if not duration:
 376                     duration = int_or_none(data.get('duration'))
 377                 extract_subtitles(data.get('tracks'))
 378
 379             if not subtitles and not automatic_captions:
 380                 text_tracks = self._parse_json(
 381                     self._search_regex(
 382                         r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
 383                         'text tracks', default='{}', group='data'), video_id,
 384                     transform_source=lambda s: js_to_json(unescapeHTML(s)),
 385                     fatal=False)
 386                 extract_subtitles(text_tracks)
 387
 388         if not formats and outputs:
 389             for format_id, output in outputs.items():
 390                 f = extract_output_format(output, format_id)
 391                 if f.get('url'):
 392                     formats.append(f)
 393
 394         self._sort_formats(formats)
 395
 396         return {
 397             'id': video_id,
 398             'title': title,
 399             'description': description,
 400             'thumbnail': thumbnail,
 401             'duration': duration,
 402             'formats': formats,
 403             'subtitles': subtitles,
 404             'automatic_captions': automatic_captions,
 405         }
 406
 407
 408 class UdemyCourseIE(UdemyIE):
 409     IE_NAME = 'udemy:course'
 410     _VALID_URL = r'https?://(?:[^/]+\.)?udemy\.com/(?P<id>[^/?#&]+)'
 411     _TESTS = [{
 412         'url': 'https://www.udemy.com/java-tutorial/',
 413         'only_matching': True,
 414     }, {
 415         'url': 'https://wipro.udemy.com/java-tutorial/',
 416         'only_matching': True,
 417     }]
 418
 419     @classmethod
 420     def suitable(cls, url):
 421         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 422
 423     def _real_extract(self, url):
 424         course_path = self._match_id(url)
 425
 426         webpage = self._download_webpage(url, course_path)
 427
 428         course_id, title = self._extract_course_info(webpage, course_path)
 429
 430         self._enroll_course(url, webpage, course_id)
 431
 432         response = self._download_json(
 433             'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
 434             course_id, 'Downloading course curriculum', query={
 435                 'fields[chapter]': 'title,object_index',
 436                 'fields[lecture]': 'title,asset',
 437                 'page_size': '1000',
 438             })
 439
 440         entries = []
 441         chapter, chapter_number = [None] * 2
 442         for entry in response['results']:
 443             clazz = entry.get('_class')
 444             if clazz == 'lecture':
 445                 asset = entry.get('asset')
 446                 if isinstance(asset, dict):
 447                     asset_type = asset.get('asset_type') or asset.get('assetType')
 448                     if asset_type != 'Video':
 449                         continue
 450                 lecture_id = entry.get('id')
 451                 if lecture_id:
 452                     entry = {
 453                         '_type': 'url_transparent',
 454                         'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
 455                         'title': entry.get('title'),
 456                         'ie_key': UdemyIE.ie_key(),
 457                     }
 458                     if chapter_number:
 459                         entry['chapter_number'] = chapter_number
 460                     if chapter:
 461                         entry['chapter'] = chapter
 462                     entries.append(entry)
 463             elif clazz == 'chapter':
 464                 chapter_number = entry.get('object_index')
 465                 chapter = entry.get('title')
 466
 467         return self.playlist_result(entries, course_id, title)