yt_dlp/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_str,
   9     compat_urllib_request,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     determine_ext,
  14     extract_attributes,
  15     ExtractorError,
  16     float_or_none,
  17     int_or_none,
  18     js_to_json,
  19     sanitized_Request,
  20     try_get,
  21     unescapeHTML,
  22     url_or_none,
  23     urlencode_postdata,
  24 )
  25
  26
  27 class UdemyIE(InfoExtractor):
  28     IE_NAME = 'udemy'
  29     _VALID_URL = r'''(?x)
  30                     https?://
  31                         (?:[^/]+\.)?udemy\.com/
  32                         (?:
  33                             [^#]+\#/lecture/|
  34                             lecture/view/?\?lectureId=|
  35                             [^/]+/learn/v4/t/lecture/
  36                         )
  37                         (?P<id>\d+)
  38                     '''
  39     _LOGIN_URL = 'https://www.udemy.com/join/login-popup/?displayType=ajax&showSkipButton=1'
  40     _ORIGIN_URL = 'https://www.udemy.com'
  41     _NETRC_MACHINE = 'udemy'
  42
  43     _TESTS = [{
  44         'url': 'https://www.udemy.com/java-tutorial/#/lecture/172757',
  45         'md5': '98eda5b657e752cf945d8445e261b5c5',
  46         'info_dict': {
  47             'id': '160614',
  48             'ext': 'mp4',
  49             'title': 'Introduction and Installation',
  50             'description': 'md5:c0d51f6f21ef4ec65f091055a5eef876',
  51             'duration': 579.29,
  52         },
  53         'skip': 'Requires udemy account credentials',
  54     }, {
  55         # new URL schema
  56         'url': 'https://www.udemy.com/electric-bass-right-from-the-start/learn/v4/t/lecture/4580906',
  57         'only_matching': True,
  58     }, {
  59         # no url in outputs format entry
  60         'url': 'https://www.udemy.com/learn-web-development-complete-step-by-step-guide-to-success/learn/v4/t/lecture/4125812',
  61         'only_matching': True,
  62     }, {
  63         # only outputs rendition
  64         'url': 'https://www.udemy.com/how-you-can-help-your-local-community-5-amazing-examples/learn/v4/t/lecture/3225750?start=0',
  65         'only_matching': True,
  66     }, {
  67         'url': 'https://wipro.udemy.com/java-tutorial/#/lecture/172757',
  68         'only_matching': True,
  69     }]
  70
  71     def _extract_course_info(self, webpage, video_id):
  72         course = self._parse_json(
  73             unescapeHTML(self._search_regex(
  74                 r'ng-init=["\'].*\bcourse=({.+?})[;"\']',
  75                 webpage, 'course', default='{}')),
  76             video_id, fatal=False) or {}
  77         course_id = course.get('id') or self._search_regex(
  78             [
  79                 r'data-course-id=["\'](\d+)',
  80                 r'&quot;courseId&quot;\s*:\s*(\d+)'
  81             ], webpage, 'course id')
  82         return course_id, course.get('title')
  83
  84     def _enroll_course(self, base_url, webpage, course_id):
  85         def combine_url(base_url, url):
  86             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  87
  88         checkout_url = unescapeHTML(self._search_regex(
  89             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/(?:payment|cart)/checkout/.+?)\1',
  90             webpage, 'checkout url', group='url', default=None))
  91         if checkout_url:
  92             raise ExtractorError(
  93                 'Course %s is not free. You have to pay for it before you can download. '
  94                 'Use this URL to confirm purchase: %s'
  95                 % (course_id, combine_url(base_url, checkout_url)),
  96                 expected=True)
  97
  98         enroll_url = unescapeHTML(self._search_regex(
  99             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
 100             webpage, 'enroll url', group='url', default=None))
 101         if enroll_url:
 102             webpage = self._download_webpage(
 103                 combine_url(base_url, enroll_url),
 104                 course_id, 'Enrolling in the course',
 105                 headers={'Referer': base_url})
 106             if '>You have enrolled in' in webpage:
 107                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
 108
 109     def _download_lecture(self, course_id, lecture_id):
 110         return self._download_json(
 111             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
 112             % (course_id, lecture_id),
 113             lecture_id, 'Downloading lecture JSON', query={
 114                 'fields[lecture]': 'title,description,view_html,asset',
 115                 'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,stream_urls,captions,data',
 116             })
 117
 118     def _handle_error(self, response):
 119         if not isinstance(response, dict):
 120             return
 121         error = response.get('error')
 122         if error:
 123             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
 124             error_data = error.get('data')
 125             if error_data:
 126                 error_str += ' - %s' % error_data.get('formErrors')
 127             raise ExtractorError(error_str, expected=True)
 128
 129     def _download_webpage_handle(self, *args, **kwargs):
 130         headers = kwargs.get('headers', {}).copy()
 131         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'
 132         kwargs['headers'] = headers
 133         ret = super(UdemyIE, self)._download_webpage_handle(
 134             *args, **kwargs)
 135         if not ret:
 136             return ret
 137         webpage, _ = ret
 138         if any(p in webpage for p in (
 139                 '>Please verify you are a human',
 140                 'Access to this page has been denied because we believe you are using automation tools to browse the website',
 141                 '"_pxCaptcha"')):
 142             raise ExtractorError(
 143                 'Udemy asks you to solve a CAPTCHA. Login with browser, '
 144                 'solve CAPTCHA, then export cookies and pass cookie file to '
 145                 'yt-dlp with --cookies.', expected=True)
 146         return ret
 147
 148     def _download_json(self, url_or_request, *args, **kwargs):
 149         headers = {
 150             'X-Udemy-Snail-Case': 'true',
 151             'X-Requested-With': 'XMLHttpRequest',
 152         }
 153         for cookie in self._downloader.cookiejar:
 154             if cookie.name == 'client_id':
 155                 headers['X-Udemy-Client-Id'] = cookie.value
 156             elif cookie.name == 'access_token':
 157                 headers['X-Udemy-Bearer-Token'] = cookie.value
 158                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 159
 160         if isinstance(url_or_request, compat_urllib_request.Request):
 161             for header, value in headers.items():
 162                 url_or_request.add_header(header, value)
 163         else:
 164             url_or_request = sanitized_Request(url_or_request, headers=headers)
 165
 166         response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)
 167         self._handle_error(response)
 168         return response
 169
 170     def _perform_login(self, username, password):
 171         login_popup = self._download_webpage(
 172             self._LOGIN_URL, None, 'Downloading login popup')
 173
 174         def is_logged(webpage):
 175             return any(re.search(p, webpage) for p in (
 176                 r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
 177                 r'>Logout<'))
 178
 179         # already logged in
 180         if is_logged(login_popup):
 181             return
 182
 183         login_form = self._form_hidden_inputs('login-form', login_popup)
 184
 185         login_form.update({
 186             'email': username,
 187             'password': password,
 188         })
 189
 190         response = self._download_webpage(
 191             self._LOGIN_URL, None, 'Logging in',
 192             data=urlencode_postdata(login_form),
 193             headers={
 194                 'Referer': self._ORIGIN_URL,
 195                 'Origin': self._ORIGIN_URL,
 196             })
 197
 198         if not is_logged(response):
 199             error = self._html_search_regex(
 200                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 201                 response, 'error message', default=None)
 202             if error:
 203                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 204             raise ExtractorError('Unable to log in')
 205
 206     def _real_extract(self, url):
 207         lecture_id = self._match_id(url)
 208
 209         webpage = self._download_webpage(url, lecture_id)
 210
 211         course_id, _ = self._extract_course_info(webpage, lecture_id)
 212
 213         try:
 214             lecture = self._download_lecture(course_id, lecture_id)
 215         except ExtractorError as e:
 216             # Error could possibly mean we are not enrolled in the course
 217             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 218                 self._enroll_course(url, webpage, course_id)
 219                 lecture = self._download_lecture(course_id, lecture_id)
 220             else:
 221                 raise
 222
 223         title = lecture['title']
 224         description = lecture.get('description')
 225
 226         asset = lecture['asset']
 227
 228         asset_type = asset.get('asset_type') or asset.get('assetType')
 229         if asset_type != 'Video':
 230             raise ExtractorError(
 231                 'Lecture %s is not a video' % lecture_id, expected=True)
 232
 233         stream_url = asset.get('stream_url') or asset.get('streamUrl')
 234         if stream_url:
 235             youtube_url = self._search_regex(
 236                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 237             if youtube_url:
 238                 return self.url_result(youtube_url, 'Youtube')
 239
 240         video_id = compat_str(asset['id'])
 241         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
 242         duration = float_or_none(asset.get('data', {}).get('duration'))
 243
 244         subtitles = {}
 245         automatic_captions = {}
 246
 247         formats = []
 248
 249         def extract_output_format(src, f_id):
 250             return {
 251                 'url': src.get('url'),
 252                 'format_id': '%sp' % (src.get('height') or f_id),
 253                 'width': int_or_none(src.get('width')),
 254                 'height': int_or_none(src.get('height')),
 255                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 256                 'vcodec': src.get('video_codec'),
 257                 'fps': int_or_none(src.get('frame_rate')),
 258                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 259                 'acodec': src.get('audio_codec'),
 260                 'asr': int_or_none(src.get('audio_sample_rate')),
 261                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 262                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 263             }
 264
 265         outputs = asset.get('data', {}).get('outputs')
 266         if not isinstance(outputs, dict):
 267             outputs = {}
 268
 269         def add_output_format_meta(f, key):
 270             output = outputs.get(key)
 271             if isinstance(output, dict):
 272                 output_format = extract_output_format(output, key)
 273                 output_format.update(f)
 274                 return output_format
 275             return f
 276
 277         def extract_formats(source_list):
 278             if not isinstance(source_list, list):
 279                 return
 280             for source in source_list:
 281                 video_url = url_or_none(source.get('file') or source.get('src'))
 282                 if not video_url:
 283                     continue
 284                 if source.get('type') == 'application/x-mpegURL' or determine_ext(video_url) == 'm3u8':
 285                     formats.extend(self._extract_m3u8_formats(
 286                         video_url, video_id, 'mp4', entry_protocol='m3u8_native',
 287                         m3u8_id='hls', fatal=False))
 288                     continue
 289                 format_id = source.get('label')
 290                 f = {
 291                     'url': video_url,
 292                     'format_id': '%sp' % format_id,
 293                     'height': int_or_none(format_id),
 294                 }
 295                 if format_id:
 296                     # Some videos contain additional metadata (e.g.
 297                     # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 298                     f = add_output_format_meta(f, format_id)
 299                 formats.append(f)
 300
 301         def extract_subtitles(track_list):
 302             if not isinstance(track_list, list):
 303                 return
 304             for track in track_list:
 305                 if not isinstance(track, dict):
 306                     continue
 307                 if track.get('kind') != 'captions':
 308                     continue
 309                 src = url_or_none(track.get('src'))
 310                 if not src:
 311                     continue
 312                 lang = track.get('language') or track.get(
 313                     'srclang') or track.get('label')
 314                 sub_dict = automatic_captions if track.get(
 315                     'autogenerated') is True else subtitles
 316                 sub_dict.setdefault(lang, []).append({
 317                     'url': src,
 318                 })
 319
 320         for url_kind in ('download', 'stream'):
 321             urls = asset.get('%s_urls' % url_kind)
 322             if isinstance(urls, dict):
 323                 extract_formats(urls.get('Video'))
 324
 325         captions = asset.get('captions')
 326         if isinstance(captions, list):
 327             for cc in captions:
 328                 if not isinstance(cc, dict):
 329                     continue
 330                 cc_url = url_or_none(cc.get('url'))
 331                 if not cc_url:
 332                     continue
 333                 lang = try_get(cc, lambda x: x['locale']['locale'], compat_str)
 334                 sub_dict = (automatic_captions if cc.get('source') == 'auto'
 335                             else subtitles)
 336                 sub_dict.setdefault(lang or 'en', []).append({
 337                     'url': cc_url,
 338                 })
 339
 340         view_html = lecture.get('view_html')
 341         if view_html:
 342             view_html_urls = set()
 343             for source in re.findall(r'<source[^>]+>', view_html):
 344                 attributes = extract_attributes(source)
 345                 src = attributes.get('src')
 346                 if not src:
 347                     continue
 348                 res = attributes.get('data-res')
 349                 height = int_or_none(res)
 350                 if src in view_html_urls:
 351                     continue
 352                 view_html_urls.add(src)
 353                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 354                     m3u8_formats = self._extract_m3u8_formats(
 355                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 356                         m3u8_id='hls', fatal=False)
 357                     for f in m3u8_formats:
 358                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 359                         if m:
 360                             if not f.get('height'):
 361                                 f['height'] = int(m.group('height'))
 362                             if not f.get('tbr'):
 363                                 f['tbr'] = int(m.group('tbr'))
 364                     formats.extend(m3u8_formats)
 365                 else:
 366                     formats.append(add_output_format_meta({
 367                         'url': src,
 368                         'format_id': '%dp' % height if height else None,
 369                         'height': height,
 370                     }, res))
 371
 372             # react rendition since 2017.04.15 (see
 373             # https://github.com/ytdl-org/youtube-dl/issues/12744)
 374             data = self._parse_json(
 375                 self._search_regex(
 376                     r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
 377                     'setup data', default='{}', group='data'), video_id,
 378                 transform_source=unescapeHTML, fatal=False)
 379             if data and isinstance(data, dict):
 380                 extract_formats(data.get('sources'))
 381                 if not duration:
 382                     duration = int_or_none(data.get('duration'))
 383                 extract_subtitles(data.get('tracks'))
 384
 385             if not subtitles and not automatic_captions:
 386                 text_tracks = self._parse_json(
 387                     self._search_regex(
 388                         r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
 389                         'text tracks', default='{}', group='data'), video_id,
 390                     transform_source=lambda s: js_to_json(unescapeHTML(s)),
 391                     fatal=False)
 392                 extract_subtitles(text_tracks)
 393
 394         if not formats and outputs:
 395             for format_id, output in outputs.items():
 396                 f = extract_output_format(output, format_id)
 397                 if f.get('url'):
 398                     formats.append(f)
 399
 400         self._sort_formats(formats)
 401
 402         return {
 403             'id': video_id,
 404             'title': title,
 405             'description': description,
 406             'thumbnail': thumbnail,
 407             'duration': duration,
 408             'formats': formats,
 409             'subtitles': subtitles,
 410             'automatic_captions': automatic_captions,
 411         }
 412
 413
 414 class UdemyCourseIE(UdemyIE):
 415     IE_NAME = 'udemy:course'
 416     _VALID_URL = r'https?://(?:[^/]+\.)?udemy\.com/(?P<id>[^/?#&]+)'
 417     _TESTS = [{
 418         'url': 'https://www.udemy.com/java-tutorial/',
 419         'only_matching': True,
 420     }, {
 421         'url': 'https://wipro.udemy.com/java-tutorial/',
 422         'only_matching': True,
 423     }]
 424
 425     @classmethod
 426     def suitable(cls, url):
 427         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 428
 429     def _real_extract(self, url):
 430         course_path = self._match_id(url)
 431
 432         webpage = self._download_webpage(url, course_path)
 433
 434         course_id, title = self._extract_course_info(webpage, course_path)
 435
 436         self._enroll_course(url, webpage, course_id)
 437
 438         response = self._download_json(
 439             'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
 440             course_id, 'Downloading course curriculum', query={
 441                 'fields[chapter]': 'title,object_index',
 442                 'fields[lecture]': 'title,asset',
 443                 'page_size': '1000',
 444             })
 445
 446         entries = []
 447         chapter, chapter_number = [None] * 2
 448         for entry in response['results']:
 449             clazz = entry.get('_class')
 450             if clazz == 'lecture':
 451                 asset = entry.get('asset')
 452                 if isinstance(asset, dict):
 453                     asset_type = asset.get('asset_type') or asset.get('assetType')
 454                     if asset_type != 'Video':
 455                         continue
 456                 lecture_id = entry.get('id')
 457                 if lecture_id:
 458                     entry = {
 459                         '_type': 'url_transparent',
 460                         'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
 461                         'title': entry.get('title'),
 462                         'ie_key': UdemyIE.ie_key(),
 463                     }
 464                     if chapter_number:
 465                         entry['chapter_number'] = chapter_number
 466                     if chapter:
 467                         entry['chapter'] = chapter
 468                     entries.append(entry)
 469             elif clazz == 'chapter':
 470                 chapter_number = entry.get('object_index')
 471                 chapter = entry.get('title')
 472
 473         return self.playlist_result(entries, course_id, title)