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     try_get,
  22     unescapeHTML,
  23     url_or_none,
  24     urlencode_postdata,
  25 )
  26
  27
  28 class UdemyIE(InfoExtractor):
  29     IE_NAME = 'udemy'
  30     _VALID_URL = r'''(?x)
  31                     https?://
  32                         (?:[^/]+\.)?udemy\.com/
  33                         (?:
  34                             [^#]+\#/lecture/|
  35                             lecture/view/?\?lectureId=|
  36                             [^/]+/learn/v4/t/lecture/
  37                         )
  38                         (?P<id>\d+)
  39                     '''
  40     _LOGIN_URL = 'https://www.udemy.com/join/login-popup/?displayType=ajax&showSkipButton=1'
  41     _ORIGIN_URL = 'https://www.udemy.com'
  42     _NETRC_MACHINE = 'udemy'
  43
  44     _TESTS = [{
  45         'url': 'https://www.udemy.com/java-tutorial/#/lecture/172757',
  46         'md5': '98eda5b657e752cf945d8445e261b5c5',
  47         'info_dict': {
  48             'id': '160614',
  49             'ext': 'mp4',
  50             'title': 'Introduction and Installation',
  51             'description': 'md5:c0d51f6f21ef4ec65f091055a5eef876',
  52             'duration': 579.29,
  53         },
  54         'skip': 'Requires udemy account credentials',
  55     }, {
  56         # new URL schema
  57         'url': 'https://www.udemy.com/electric-bass-right-from-the-start/learn/v4/t/lecture/4580906',
  58         'only_matching': True,
  59     }, {
  60         # no url in outputs format entry
  61         'url': 'https://www.udemy.com/learn-web-development-complete-step-by-step-guide-to-success/learn/v4/t/lecture/4125812',
  62         'only_matching': True,
  63     }, {
  64         # only outputs rendition
  65         'url': 'https://www.udemy.com/how-you-can-help-your-local-community-5-amazing-examples/learn/v4/t/lecture/3225750?start=0',
  66         'only_matching': True,
  67     }, {
  68         'url': 'https://wipro.udemy.com/java-tutorial/#/lecture/172757',
  69         'only_matching': True,
  70     }]
  71
  72     def _extract_course_info(self, webpage, video_id):
  73         course = self._parse_json(
  74             unescapeHTML(self._search_regex(
  75                 r'ng-init=["\'].*\bcourse=({.+?})[;"\']',
  76                 webpage, 'course', default='{}')),
  77             video_id, fatal=False) or {}
  78         course_id = course.get('id') or self._search_regex(
  79             r'data-course-id=["\'](\d+)', webpage, 'course id')
  80         return course_id, course.get('title')
  81
  82     def _enroll_course(self, base_url, webpage, course_id):
  83         def combine_url(base_url, url):
  84             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  85
  86         checkout_url = unescapeHTML(self._search_regex(
  87             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/(?:payment|cart)/checkout/.+?)\1',
  88             webpage, 'checkout url', group='url', default=None))
  89         if checkout_url:
  90             raise ExtractorError(
  91                 'Course %s is not free. You have to pay for it before you can download. '
  92                 'Use this URL to confirm purchase: %s'
  93                 % (course_id, combine_url(base_url, checkout_url)),
  94                 expected=True)
  95
  96         enroll_url = unescapeHTML(self._search_regex(
  97             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  98             webpage, 'enroll url', group='url', default=None))
  99         if enroll_url:
 100             webpage = self._download_webpage(
 101                 combine_url(base_url, enroll_url),
 102                 course_id, 'Enrolling in the course',
 103                 headers={'Referer': base_url})
 104             if '>You have enrolled in' in webpage:
 105                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
 106
 107     def _download_lecture(self, course_id, lecture_id):
 108         return self._download_json(
 109             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
 110             % (course_id, lecture_id),
 111             lecture_id, 'Downloading lecture JSON', query={
 112                 'fields[lecture]': 'title,description,view_html,asset',
 113                 'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,stream_urls,captions,data',
 114             })
 115
 116     def _handle_error(self, response):
 117         if not isinstance(response, dict):
 118             return
 119         error = response.get('error')
 120         if error:
 121             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
 122             error_data = error.get('data')
 123             if error_data:
 124                 error_str += ' - %s' % error_data.get('formErrors')
 125             raise ExtractorError(error_str, expected=True)
 126
 127     def _download_webpage_handle(self, *args, **kwargs):
 128         headers = kwargs.get('headers', {}).copy()
 129         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'
 130         kwargs['headers'] = headers
 131         ret = super(UdemyIE, self)._download_webpage_handle(
 132             *args, **compat_kwargs(kwargs))
 133         if not ret:
 134             return ret
 135         webpage, _ = ret
 136         if any(p in webpage for p in (
 137                 '>Please verify you are a human',
 138                 'Access to this page has been denied because we believe you are using automation tools to browse the website',
 139                 '"_pxCaptcha"')):
 140             raise ExtractorError(
 141                 'Udemy asks you to solve a CAPTCHA. Login with browser, '
 142                 'solve CAPTCHA, then export cookies and pass cookie file to '
 143                 'youtube-dl with --cookies.', expected=True)
 144         return ret
 145
 146     def _download_json(self, url_or_request, *args, **kwargs):
 147         headers = {
 148             'X-Udemy-Snail-Case': 'true',
 149             'X-Requested-With': 'XMLHttpRequest',
 150         }
 151         for cookie in self._downloader.cookiejar:
 152             if cookie.name == 'client_id':
 153                 headers['X-Udemy-Client-Id'] = cookie.value
 154             elif cookie.name == 'access_token':
 155                 headers['X-Udemy-Bearer-Token'] = cookie.value
 156                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 157
 158         if isinstance(url_or_request, compat_urllib_request.Request):
 159             for header, value in headers.items():
 160                 url_or_request.add_header(header, value)
 161         else:
 162             url_or_request = sanitized_Request(url_or_request, headers=headers)
 163
 164         response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)
 165         self._handle_error(response)
 166         return response
 167
 168     def _real_initialize(self):
 169         self._login()
 170
 171     def _login(self):
 172         username, password = self._get_login_info()
 173         if username is None:
 174             return
 175
 176         login_popup = self._download_webpage(
 177             self._LOGIN_URL, None, 'Downloading login popup')
 178
 179         def is_logged(webpage):
 180             return any(re.search(p, webpage) for p in (
 181                 r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
 182                 r'>Logout<'))
 183
 184         # already logged in
 185         if is_logged(login_popup):
 186             return
 187
 188         login_form = self._form_hidden_inputs('login-form', login_popup)
 189
 190         login_form.update({
 191             'email': username,
 192             'password': password,
 193         })
 194
 195         response = self._download_webpage(
 196             self._LOGIN_URL, None, 'Logging in',
 197             data=urlencode_postdata(login_form),
 198             headers={
 199                 'Referer': self._ORIGIN_URL,
 200                 'Origin': self._ORIGIN_URL,
 201             })
 202
 203         if not is_logged(response):
 204             error = self._html_search_regex(
 205                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 206                 response, 'error message', default=None)
 207             if error:
 208                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 209             raise ExtractorError('Unable to log in')
 210
 211     def _real_extract(self, url):
 212         lecture_id = self._match_id(url)
 213
 214         webpage = self._download_webpage(url, lecture_id)
 215
 216         course_id, _ = self._extract_course_info(webpage, lecture_id)
 217
 218         try:
 219             lecture = self._download_lecture(course_id, lecture_id)
 220         except ExtractorError as e:
 221             # Error could possibly mean we are not enrolled in the course
 222             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 223                 self._enroll_course(url, webpage, course_id)
 224                 lecture = self._download_lecture(course_id, lecture_id)
 225             else:
 226                 raise
 227
 228         title = lecture['title']
 229         description = lecture.get('description')
 230
 231         asset = lecture['asset']
 232
 233         asset_type = asset.get('asset_type') or asset.get('assetType')
 234         if asset_type != 'Video':
 235             raise ExtractorError(
 236                 'Lecture %s is not a video' % lecture_id, expected=True)
 237
 238         stream_url = asset.get('stream_url') or asset.get('streamUrl')
 239         if stream_url:
 240             youtube_url = self._search_regex(
 241                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 242             if youtube_url:
 243                 return self.url_result(youtube_url, 'Youtube')
 244
 245         video_id = compat_str(asset['id'])
 246         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
 247         duration = float_or_none(asset.get('data', {}).get('duration'))
 248
 249         subtitles = {}
 250         automatic_captions = {}
 251
 252         formats = []
 253
 254         def extract_output_format(src, f_id):
 255             return {
 256                 'url': src.get('url'),
 257                 'format_id': '%sp' % (src.get('height') or f_id),
 258                 'width': int_or_none(src.get('width')),
 259                 'height': int_or_none(src.get('height')),
 260                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 261                 'vcodec': src.get('video_codec'),
 262                 'fps': int_or_none(src.get('frame_rate')),
 263                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 264                 'acodec': src.get('audio_codec'),
 265                 'asr': int_or_none(src.get('audio_sample_rate')),
 266                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 267                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 268             }
 269
 270         outputs = asset.get('data', {}).get('outputs')
 271         if not isinstance(outputs, dict):
 272             outputs = {}
 273
 274         def add_output_format_meta(f, key):
 275             output = outputs.get(key)
 276             if isinstance(output, dict):
 277                 output_format = extract_output_format(output, key)
 278                 output_format.update(f)
 279                 return output_format
 280             return f
 281
 282         def extract_formats(source_list):
 283             if not isinstance(source_list, list):
 284                 return
 285             for source in source_list:
 286                 video_url = url_or_none(source.get('file') or source.get('src'))
 287                 if not video_url:
 288                     continue
 289                 if source.get('type') == 'application/x-mpegURL' or determine_ext(video_url) == 'm3u8':
 290                     formats.extend(self._extract_m3u8_formats(
 291                         video_url, video_id, 'mp4', entry_protocol='m3u8_native',
 292                         m3u8_id='hls', fatal=False))
 293                     continue
 294                 format_id = source.get('label')
 295                 f = {
 296                     'url': video_url,
 297                     'format_id': '%sp' % format_id,
 298                     'height': int_or_none(format_id),
 299                 }
 300                 if format_id:
 301                     # Some videos contain additional metadata (e.g.
 302                     # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 303                     f = add_output_format_meta(f, format_id)
 304                 formats.append(f)
 305
 306         def extract_subtitles(track_list):
 307             if not isinstance(track_list, list):
 308                 return
 309             for track in track_list:
 310                 if not isinstance(track, dict):
 311                     continue
 312                 if track.get('kind') != 'captions':
 313                     continue
 314                 src = url_or_none(track.get('src'))
 315                 if not src:
 316                     continue
 317                 lang = track.get('language') or track.get(
 318                     'srclang') or track.get('label')
 319                 sub_dict = automatic_captions if track.get(
 320                     'autogenerated') is True else subtitles
 321                 sub_dict.setdefault(lang, []).append({
 322                     'url': src,
 323                 })
 324
 325         for url_kind in ('download', 'stream'):
 326             urls = asset.get('%s_urls' % url_kind)
 327             if isinstance(urls, dict):
 328                 extract_formats(urls.get('Video'))
 329
 330         captions = asset.get('captions')
 331         if isinstance(captions, list):
 332             for cc in captions:
 333                 if not isinstance(cc, dict):
 334                     continue
 335                 cc_url = url_or_none(cc.get('url'))
 336                 if not cc_url:
 337                     continue
 338                 lang = try_get(cc, lambda x: x['locale']['locale'], compat_str)
 339                 sub_dict = (automatic_captions if cc.get('source') == 'auto'
 340                             else subtitles)
 341                 sub_dict.setdefault(lang or 'en', []).append({
 342                     'url': cc_url,
 343                 })
 344
 345         view_html = lecture.get('view_html')
 346         if view_html:
 347             view_html_urls = set()
 348             for source in re.findall(r'<source[^>]+>', view_html):
 349                 attributes = extract_attributes(source)
 350                 src = attributes.get('src')
 351                 if not src:
 352                     continue
 353                 res = attributes.get('data-res')
 354                 height = int_or_none(res)
 355                 if src in view_html_urls:
 356                     continue
 357                 view_html_urls.add(src)
 358                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 359                     m3u8_formats = self._extract_m3u8_formats(
 360                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 361                         m3u8_id='hls', fatal=False)
 362                     for f in m3u8_formats:
 363                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 364                         if m:
 365                             if not f.get('height'):
 366                                 f['height'] = int(m.group('height'))
 367                             if not f.get('tbr'):
 368                                 f['tbr'] = int(m.group('tbr'))
 369                     formats.extend(m3u8_formats)
 370                 else:
 371                     formats.append(add_output_format_meta({
 372                         'url': src,
 373                         'format_id': '%dp' % height if height else None,
 374                         'height': height,
 375                     }, res))
 376
 377             # react rendition since 2017.04.15 (see
 378             # https://github.com/ytdl-org/youtube-dl/issues/12744)
 379             data = self._parse_json(
 380                 self._search_regex(
 381                     r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
 382                     'setup data', default='{}', group='data'), video_id,
 383                 transform_source=unescapeHTML, fatal=False)
 384             if data and isinstance(data, dict):
 385                 extract_formats(data.get('sources'))
 386                 if not duration:
 387                     duration = int_or_none(data.get('duration'))
 388                 extract_subtitles(data.get('tracks'))
 389
 390             if not subtitles and not automatic_captions:
 391                 text_tracks = self._parse_json(
 392                     self._search_regex(
 393                         r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
 394                         'text tracks', default='{}', group='data'), video_id,
 395                     transform_source=lambda s: js_to_json(unescapeHTML(s)),
 396                     fatal=False)
 397                 extract_subtitles(text_tracks)
 398
 399         if not formats and outputs:
 400             for format_id, output in outputs.items():
 401                 f = extract_output_format(output, format_id)
 402                 if f.get('url'):
 403                     formats.append(f)
 404
 405         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 406
 407         return {
 408             'id': video_id,
 409             'title': title,
 410             'description': description,
 411             'thumbnail': thumbnail,
 412             'duration': duration,
 413             'formats': formats,
 414             'subtitles': subtitles,
 415             'automatic_captions': automatic_captions,
 416         }
 417
 418
 419 class UdemyCourseIE(UdemyIE):
 420     IE_NAME = 'udemy:course'
 421     _VALID_URL = r'https?://(?:[^/]+\.)?udemy\.com/(?P<id>[^/?#&]+)'
 422     _TESTS = [{
 423         'url': 'https://www.udemy.com/java-tutorial/',
 424         'only_matching': True,
 425     }, {
 426         'url': 'https://wipro.udemy.com/java-tutorial/',
 427         'only_matching': True,
 428     }]
 429
 430     @classmethod
 431     def suitable(cls, url):
 432         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 433
 434     def _real_extract(self, url):
 435         course_path = self._match_id(url)
 436
 437         webpage = self._download_webpage(url, course_path)
 438
 439         course_id, title = self._extract_course_info(webpage, course_path)
 440
 441         self._enroll_course(url, webpage, course_id)
 442
 443         response = self._download_json(
 444             'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
 445             course_id, 'Downloading course curriculum', query={
 446                 'fields[chapter]': 'title,object_index',
 447                 'fields[lecture]': 'title,asset',
 448                 'page_size': '1000',
 449             })
 450
 451         entries = []
 452         chapter, chapter_number = [None] * 2
 453         for entry in response['results']:
 454             clazz = entry.get('_class')
 455             if clazz == 'lecture':
 456                 asset = entry.get('asset')
 457                 if isinstance(asset, dict):
 458                     asset_type = asset.get('asset_type') or asset.get('assetType')
 459                     if asset_type != 'Video':
 460                         continue
 461                 lecture_id = entry.get('id')
 462                 if lecture_id:
 463                     entry = {
 464                         '_type': 'url_transparent',
 465                         'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
 466                         'title': entry.get('title'),
 467                         'ie_key': UdemyIE.ie_key(),
 468                     }
 469                     if chapter_number:
 470                         entry['chapter_number'] = chapter_number
 471                     if chapter:
 472                         entry['chapter'] = chapter
 473                     entries.append(entry)
 474             elif clazz == 'chapter':
 475                 chapter_number = entry.get('object_index')
 476                 chapter = entry.get('title')
 477
 478         return self.playlist_result(entries, course_id, title)