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         # only outputs rendition
  63         'url': 'https://www.udemy.com/how-you-can-help-your-local-community-5-amazing-examples/learn/v4/t/lecture/3225750?start=0',
  64         'only_matching': True,
  65     }]
  66
  67     def _extract_course_info(self, webpage, video_id):
  68         course = self._parse_json(
  69             unescapeHTML(self._search_regex(
  70                 r'ng-init=["\'].*\bcourse=({.+?})[;"\']',
  71                 webpage, 'course', default='{}')),
  72             video_id, fatal=False) or {}
  73         course_id = course.get('id') or self._search_regex(
  74             r'data-course-id=["\'](\d+)', webpage, 'course id')
  75         return course_id, course.get('title')
  76
  77     def _enroll_course(self, base_url, webpage, course_id):
  78         def combine_url(base_url, url):
  79             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  80
  81         checkout_url = unescapeHTML(self._search_regex(
  82             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/(?:payment|cart)/checkout/.+?)\1',
  83             webpage, 'checkout url', group='url', default=None))
  84         if checkout_url:
  85             raise ExtractorError(
  86                 'Course %s is not free. You have to pay for it before you can download. '
  87                 'Use this URL to confirm purchase: %s'
  88                 % (course_id, combine_url(base_url, checkout_url)),
  89                 expected=True)
  90
  91         enroll_url = unescapeHTML(self._search_regex(
  92             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  93             webpage, 'enroll url', group='url', default=None))
  94         if enroll_url:
  95             webpage = self._download_webpage(
  96                 combine_url(base_url, enroll_url),
  97                 course_id, 'Enrolling in the course',
  98                 headers={'Referer': base_url})
  99             if '>You have enrolled in' in webpage:
 100                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
 101
 102     def _download_lecture(self, course_id, lecture_id):
 103         return self._download_json(
 104             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
 105             % (course_id, lecture_id),
 106             lecture_id, 'Downloading lecture JSON', query={
 107                 'fields[lecture]': 'title,description,view_html,asset',
 108                 'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,stream_urls,data',
 109             })
 110
 111     def _handle_error(self, response):
 112         if not isinstance(response, dict):
 113             return
 114         error = response.get('error')
 115         if error:
 116             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
 117             error_data = error.get('data')
 118             if error_data:
 119                 error_str += ' - %s' % error_data.get('formErrors')
 120             raise ExtractorError(error_str, expected=True)
 121
 122     def _download_webpage_handle(self, *args, **kwargs):
 123         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'
 124         return super(UdemyIE, self)._download_webpage_handle(
 125             *args, **compat_kwargs(kwargs))
 126
 127     def _download_json(self, url_or_request, *args, **kwargs):
 128         headers = {
 129             'X-Udemy-Snail-Case': 'true',
 130             'X-Requested-With': 'XMLHttpRequest',
 131         }
 132         for cookie in self._downloader.cookiejar:
 133             if cookie.name == 'client_id':
 134                 headers['X-Udemy-Client-Id'] = cookie.value
 135             elif cookie.name == 'access_token':
 136                 headers['X-Udemy-Bearer-Token'] = cookie.value
 137                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 138
 139         if isinstance(url_or_request, compat_urllib_request.Request):
 140             for header, value in headers.items():
 141                 url_or_request.add_header(header, value)
 142         else:
 143             url_or_request = sanitized_Request(url_or_request, headers=headers)
 144
 145         response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)
 146         self._handle_error(response)
 147         return response
 148
 149     def _real_initialize(self):
 150         self._login()
 151
 152     def _login(self):
 153         (username, password) = self._get_login_info()
 154         if username is None:
 155             return
 156
 157         login_popup = self._download_webpage(
 158             self._LOGIN_URL, None, 'Downloading login popup')
 159
 160         def is_logged(webpage):
 161             return any(re.search(p, webpage) for p in (
 162                 r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
 163                 r'>Logout<'))
 164
 165         # already logged in
 166         if is_logged(login_popup):
 167             return
 168
 169         login_form = self._form_hidden_inputs('login-form', login_popup)
 170
 171         login_form.update({
 172             'email': username,
 173             'password': password,
 174         })
 175
 176         response = self._download_webpage(
 177             self._LOGIN_URL, None, 'Logging in',
 178             data=urlencode_postdata(login_form),
 179             headers={
 180                 'Referer': self._ORIGIN_URL,
 181                 'Origin': self._ORIGIN_URL,
 182             })
 183
 184         if not is_logged(response):
 185             error = self._html_search_regex(
 186                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 187                 response, 'error message', default=None)
 188             if error:
 189                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 190             raise ExtractorError('Unable to log in')
 191
 192     def _real_extract(self, url):
 193         lecture_id = self._match_id(url)
 194
 195         webpage = self._download_webpage(url, lecture_id)
 196
 197         course_id, _ = self._extract_course_info(webpage, lecture_id)
 198
 199         try:
 200             lecture = self._download_lecture(course_id, lecture_id)
 201         except ExtractorError as e:
 202             # Error could possibly mean we are not enrolled in the course
 203             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 204                 self._enroll_course(url, webpage, course_id)
 205                 lecture = self._download_lecture(course_id, lecture_id)
 206             else:
 207                 raise
 208
 209         title = lecture['title']
 210         description = lecture.get('description')
 211
 212         asset = lecture['asset']
 213
 214         asset_type = asset.get('asset_type') or asset.get('assetType')
 215         if asset_type != 'Video':
 216             raise ExtractorError(
 217                 'Lecture %s is not a video' % lecture_id, expected=True)
 218
 219         stream_url = asset.get('stream_url') or asset.get('streamUrl')
 220         if stream_url:
 221             youtube_url = self._search_regex(
 222                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 223             if youtube_url:
 224                 return self.url_result(youtube_url, 'Youtube')
 225
 226         video_id = compat_str(asset['id'])
 227         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
 228         duration = float_or_none(asset.get('data', {}).get('duration'))
 229
 230         subtitles = {}
 231         automatic_captions = {}
 232
 233         formats = []
 234
 235         def extract_output_format(src, f_id):
 236             return {
 237                 'url': src.get('url'),
 238                 'format_id': '%sp' % (src.get('height') or f_id),
 239                 'width': int_or_none(src.get('width')),
 240                 'height': int_or_none(src.get('height')),
 241                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 242                 'vcodec': src.get('video_codec'),
 243                 'fps': int_or_none(src.get('frame_rate')),
 244                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 245                 'acodec': src.get('audio_codec'),
 246                 'asr': int_or_none(src.get('audio_sample_rate')),
 247                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 248                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 249             }
 250
 251         outputs = asset.get('data', {}).get('outputs')
 252         if not isinstance(outputs, dict):
 253             outputs = {}
 254
 255         def add_output_format_meta(f, key):
 256             output = outputs.get(key)
 257             if isinstance(output, dict):
 258                 output_format = extract_output_format(output, key)
 259                 output_format.update(f)
 260                 return output_format
 261             return f
 262
 263         def extract_formats(source_list):
 264             if not isinstance(source_list, list):
 265                 return
 266             for source in source_list:
 267                 video_url = source.get('file') or source.get('src')
 268                 if not video_url or not isinstance(video_url, compat_str):
 269                     continue
 270                 if source.get('type') == 'application/x-mpegURL' or determine_ext(video_url) == 'm3u8':
 271                     formats.extend(self._extract_m3u8_formats(
 272                         video_url, video_id, 'mp4', entry_protocol='m3u8_native',
 273                         m3u8_id='hls', fatal=False))
 274                     continue
 275                 format_id = source.get('label')
 276                 f = {
 277                     'url': video_url,
 278                     'format_id': '%sp' % format_id,
 279                     'height': int_or_none(format_id),
 280                 }
 281                 if format_id:
 282                     # Some videos contain additional metadata (e.g.
 283                     # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 284                     f = add_output_format_meta(f, format_id)
 285                 formats.append(f)
 286
 287         def extract_subtitles(track_list):
 288             if not isinstance(track_list, list):
 289                 return
 290             for track in track_list:
 291                 if not isinstance(track, dict):
 292                     continue
 293                 if track.get('kind') != 'captions':
 294                     continue
 295                 src = track.get('src')
 296                 if not src or not isinstance(src, compat_str):
 297                     continue
 298                 lang = track.get('language') or track.get(
 299                     'srclang') or track.get('label')
 300                 sub_dict = automatic_captions if track.get(
 301                     'autogenerated') is True else subtitles
 302                 sub_dict.setdefault(lang, []).append({
 303                     'url': src,
 304                 })
 305
 306         for url_kind in ('download', 'stream'):
 307             urls = asset.get('%s_urls' % url_kind)
 308             if isinstance(urls, dict):
 309                 extract_formats(urls.get('Video'))
 310
 311         view_html = lecture.get('view_html')
 312         if view_html:
 313             view_html_urls = set()
 314             for source in re.findall(r'<source[^>]+>', view_html):
 315                 attributes = extract_attributes(source)
 316                 src = attributes.get('src')
 317                 if not src:
 318                     continue
 319                 res = attributes.get('data-res')
 320                 height = int_or_none(res)
 321                 if src in view_html_urls:
 322                     continue
 323                 view_html_urls.add(src)
 324                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 325                     m3u8_formats = self._extract_m3u8_formats(
 326                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 327                         m3u8_id='hls', fatal=False)
 328                     for f in m3u8_formats:
 329                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 330                         if m:
 331                             if not f.get('height'):
 332                                 f['height'] = int(m.group('height'))
 333                             if not f.get('tbr'):
 334                                 f['tbr'] = int(m.group('tbr'))
 335                     formats.extend(m3u8_formats)
 336                 else:
 337                     formats.append(add_output_format_meta({
 338                         'url': src,
 339                         'format_id': '%dp' % height if height else None,
 340                         'height': height,
 341                     }, res))
 342
 343             # react rendition since 2017.04.15 (see
 344             # https://github.com/rg3/youtube-dl/issues/12744)
 345             data = self._parse_json(
 346                 self._search_regex(
 347                     r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
 348                     'setup data', default='{}', group='data'), video_id,
 349                 transform_source=unescapeHTML, fatal=False)
 350             if data and isinstance(data, dict):
 351                 extract_formats(data.get('sources'))
 352                 if not duration:
 353                     duration = int_or_none(data.get('duration'))
 354                 extract_subtitles(data.get('tracks'))
 355
 356             if not subtitles and not automatic_captions:
 357                 text_tracks = self._parse_json(
 358                     self._search_regex(
 359                         r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
 360                         'text tracks', default='{}', group='data'), video_id,
 361                     transform_source=lambda s: js_to_json(unescapeHTML(s)),
 362                     fatal=False)
 363                 extract_subtitles(text_tracks)
 364
 365         if not formats and outputs:
 366             for format_id, output in outputs.items():
 367                 f = extract_output_format(output, format_id)
 368                 if f.get('url'):
 369                     formats.append(f)
 370
 371         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 372
 373         return {
 374             'id': video_id,
 375             'title': title,
 376             'description': description,
 377             'thumbnail': thumbnail,
 378             'duration': duration,
 379             'formats': formats,
 380             'subtitles': subtitles,
 381             'automatic_captions': automatic_captions,
 382         }
 383
 384
 385 class UdemyCourseIE(UdemyIE):
 386     IE_NAME = 'udemy:course'
 387     _VALID_URL = r'https?://(?:www\.)?udemy\.com/(?P<id>[^/?#&]+)'
 388     _TESTS = []
 389
 390     @classmethod
 391     def suitable(cls, url):
 392         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 393
 394     def _real_extract(self, url):
 395         course_path = self._match_id(url)
 396
 397         webpage = self._download_webpage(url, course_path)
 398
 399         course_id, title = self._extract_course_info(webpage, course_path)
 400
 401         self._enroll_course(url, webpage, course_id)
 402
 403         response = self._download_json(
 404             'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
 405             course_id, 'Downloading course curriculum', query={
 406                 'fields[chapter]': 'title,object_index',
 407                 'fields[lecture]': 'title,asset',
 408                 'page_size': '1000',
 409             })
 410
 411         entries = []
 412         chapter, chapter_number = [None] * 2
 413         for entry in response['results']:
 414             clazz = entry.get('_class')
 415             if clazz == 'lecture':
 416                 asset = entry.get('asset')
 417                 if isinstance(asset, dict):
 418                     asset_type = asset.get('asset_type') or asset.get('assetType')
 419                     if asset_type != 'Video':
 420                         continue
 421                 lecture_id = entry.get('id')
 422                 if lecture_id:
 423                     entry = {
 424                         '_type': 'url_transparent',
 425                         'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
 426                         'title': entry.get('title'),
 427                         'ie_key': UdemyIE.ie_key(),
 428                     }
 429                     if chapter_number:
 430                         entry['chapter_number'] = chapter_number
 431                     if chapter:
 432                         entry['chapter'] = chapter
 433                     entries.append(entry)
 434             elif clazz == 'chapter':
 435                 chapter_number = entry.get('object_index')
 436                 chapter = entry.get('title')
 437
 438         return self.playlist_result(entries, course_id, title)