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