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_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                         www\.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
  69     def _extract_course_info(self, webpage, video_id):
  70         course = self._parse_json(
  71             unescapeHTML(self._search_regex(
  72                 r'ng-init=["\'].*\bcourse=({.+?})[;"\']',
  73                 webpage, 'course', default='{}')),
  74             video_id, fatal=False) or {}
  75         course_id = course.get('id') or self._search_regex(
  76             r'data-course-id=["\'](\d+)', webpage, 'course id')
  77         return course_id, course.get('title')
  78
  79     def _enroll_course(self, base_url, webpage, course_id):
  80         def combine_url(base_url, url):
  81             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  82
  83         checkout_url = unescapeHTML(self._search_regex(
  84             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/(?:payment|cart)/checkout/.+?)\1',
  85             webpage, 'checkout url', group='url', default=None))
  86         if checkout_url:
  87             raise ExtractorError(
  88                 'Course %s is not free. You have to pay for it before you can download. '
  89                 'Use this URL to confirm purchase: %s'
  90                 % (course_id, combine_url(base_url, checkout_url)),
  91                 expected=True)
  92
  93         enroll_url = unescapeHTML(self._search_regex(
  94             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  95             webpage, 'enroll url', group='url', default=None))
  96         if enroll_url:
  97             webpage = self._download_webpage(
  98                 combine_url(base_url, enroll_url),
  99                 course_id, 'Enrolling in the course',
 100                 headers={'Referer': base_url})
 101             if '>You have enrolled in' in webpage:
 102                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
 103
 104     def _download_lecture(self, course_id, lecture_id):
 105         return self._download_json(
 106             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
 107             % (course_id, lecture_id),
 108             lecture_id, 'Downloading lecture JSON', query={
 109                 'fields[lecture]': 'title,description,view_html,asset',
 110                 'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,stream_urls,captions,data',
 111             })
 112
 113     def _handle_error(self, response):
 114         if not isinstance(response, dict):
 115             return
 116         error = response.get('error')
 117         if error:
 118             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
 119             error_data = error.get('data')
 120             if error_data:
 121                 error_str += ' - %s' % error_data.get('formErrors')
 122             raise ExtractorError(error_str, expected=True)
 123
 124     def _download_webpage_handle(self, *args, **kwargs):
 125         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'
 126         return super(UdemyIE, self)._download_webpage_handle(
 127             *args, **compat_kwargs(kwargs))
 128
 129     def _download_json(self, url_or_request, *args, **kwargs):
 130         headers = {
 131             'X-Udemy-Snail-Case': 'true',
 132             'X-Requested-With': 'XMLHttpRequest',
 133         }
 134         for cookie in self._downloader.cookiejar:
 135             if cookie.name == 'client_id':
 136                 headers['X-Udemy-Client-Id'] = cookie.value
 137             elif cookie.name == 'access_token':
 138                 headers['X-Udemy-Bearer-Token'] = cookie.value
 139                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 140
 141         if isinstance(url_or_request, compat_urllib_request.Request):
 142             for header, value in headers.items():
 143                 url_or_request.add_header(header, value)
 144         else:
 145             url_or_request = sanitized_Request(url_or_request, headers=headers)
 146
 147         response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)
 148         self._handle_error(response)
 149         return response
 150
 151     def _real_initialize(self):
 152         self._login()
 153
 154     def _login(self):
 155         username, password = self._get_login_info()
 156         if username is None:
 157             return
 158
 159         login_popup = self._download_webpage(
 160             self._LOGIN_URL, None, 'Downloading login popup')
 161
 162         def is_logged(webpage):
 163             return any(re.search(p, webpage) for p in (
 164                 r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
 165                 r'>Logout<'))
 166
 167         # already logged in
 168         if is_logged(login_popup):
 169             return
 170
 171         login_form = self._form_hidden_inputs('login-form', login_popup)
 172
 173         login_form.update({
 174             'email': username,
 175             'password': password,
 176         })
 177
 178         response = self._download_webpage(
 179             self._LOGIN_URL, None, 'Logging in',
 180             data=urlencode_postdata(login_form),
 181             headers={
 182                 'Referer': self._ORIGIN_URL,
 183                 'Origin': self._ORIGIN_URL,
 184             })
 185
 186         if not is_logged(response):
 187             error = self._html_search_regex(
 188                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 189                 response, 'error message', default=None)
 190             if error:
 191                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 192             raise ExtractorError('Unable to log in')
 193
 194     def _real_extract(self, url):
 195         lecture_id = self._match_id(url)
 196
 197         webpage = self._download_webpage(url, lecture_id)
 198
 199         course_id, _ = self._extract_course_info(webpage, lecture_id)
 200
 201         try:
 202             lecture = self._download_lecture(course_id, lecture_id)
 203         except ExtractorError as e:
 204             # Error could possibly mean we are not enrolled in the course
 205             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 206                 self._enroll_course(url, webpage, course_id)
 207                 lecture = self._download_lecture(course_id, lecture_id)
 208             else:
 209                 raise
 210
 211         title = lecture['title']
 212         description = lecture.get('description')
 213
 214         asset = lecture['asset']
 215
 216         asset_type = asset.get('asset_type') or asset.get('assetType')
 217         if asset_type != 'Video':
 218             raise ExtractorError(
 219                 'Lecture %s is not a video' % lecture_id, expected=True)
 220
 221         stream_url = asset.get('stream_url') or asset.get('streamUrl')
 222         if stream_url:
 223             youtube_url = self._search_regex(
 224                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 225             if youtube_url:
 226                 return self.url_result(youtube_url, 'Youtube')
 227
 228         video_id = compat_str(asset['id'])
 229         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
 230         duration = float_or_none(asset.get('data', {}).get('duration'))
 231
 232         subtitles = {}
 233         automatic_captions = {}
 234
 235         formats = []
 236
 237         def extract_output_format(src, f_id):
 238             return {
 239                 'url': src.get('url'),
 240                 'format_id': '%sp' % (src.get('height') or f_id),
 241                 'width': int_or_none(src.get('width')),
 242                 'height': int_or_none(src.get('height')),
 243                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 244                 'vcodec': src.get('video_codec'),
 245                 'fps': int_or_none(src.get('frame_rate')),
 246                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 247                 'acodec': src.get('audio_codec'),
 248                 'asr': int_or_none(src.get('audio_sample_rate')),
 249                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 250                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 251             }
 252
 253         outputs = asset.get('data', {}).get('outputs')
 254         if not isinstance(outputs, dict):
 255             outputs = {}
 256
 257         def add_output_format_meta(f, key):
 258             output = outputs.get(key)
 259             if isinstance(output, dict):
 260                 output_format = extract_output_format(output, key)
 261                 output_format.update(f)
 262                 return output_format
 263             return f
 264
 265         def extract_formats(source_list):
 266             if not isinstance(source_list, list):
 267                 return
 268             for source in source_list:
 269                 video_url = url_or_none(source.get('file') or source.get('src'))
 270                 if not video_url:
 271                     continue
 272                 if source.get('type') == 'application/x-mpegURL' or determine_ext(video_url) == 'm3u8':
 273                     formats.extend(self._extract_m3u8_formats(
 274                         video_url, video_id, 'mp4', entry_protocol='m3u8_native',
 275                         m3u8_id='hls', fatal=False))
 276                     continue
 277                 format_id = source.get('label')
 278                 f = {
 279                     'url': video_url,
 280                     'format_id': '%sp' % format_id,
 281                     'height': int_or_none(format_id),
 282                 }
 283                 if format_id:
 284                     # Some videos contain additional metadata (e.g.
 285                     # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 286                     f = add_output_format_meta(f, format_id)
 287                 formats.append(f)
 288
 289         def extract_subtitles(track_list):
 290             if not isinstance(track_list, list):
 291                 return
 292             for track in track_list:
 293                 if not isinstance(track, dict):
 294                     continue
 295                 if track.get('kind') != 'captions':
 296                     continue
 297                 src = url_or_none(track.get('src'))
 298                 if not src:
 299                     continue
 300                 lang = track.get('language') or track.get(
 301                     'srclang') or track.get('label')
 302                 sub_dict = automatic_captions if track.get(
 303                     'autogenerated') is True else subtitles
 304                 sub_dict.setdefault(lang, []).append({
 305                     'url': src,
 306                 })
 307
 308         for url_kind in ('download', 'stream'):
 309             urls = asset.get('%s_urls' % url_kind)
 310             if isinstance(urls, dict):
 311                 extract_formats(urls.get('Video'))
 312
 313         captions = asset.get('captions')
 314         if isinstance(captions, list):
 315             for cc in captions:
 316                 if not isinstance(cc, dict):
 317                     continue
 318                 cc_url = url_or_none(cc.get('url'))
 319                 if not cc_url:
 320                     continue
 321                 lang = try_get(cc, lambda x: x['locale']['locale'], compat_str)
 322                 sub_dict = (automatic_captions if cc.get('source') == 'auto'
 323                             else subtitles)
 324                 sub_dict.setdefault(lang or 'en', []).append({
 325                     'url': cc_url,
 326                 })
 327
 328         view_html = lecture.get('view_html')
 329         if view_html:
 330             view_html_urls = set()
 331             for source in re.findall(r'<source[^>]+>', view_html):
 332                 attributes = extract_attributes(source)
 333                 src = attributes.get('src')
 334                 if not src:
 335                     continue
 336                 res = attributes.get('data-res')
 337                 height = int_or_none(res)
 338                 if src in view_html_urls:
 339                     continue
 340                 view_html_urls.add(src)
 341                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 342                     m3u8_formats = self._extract_m3u8_formats(
 343                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 344                         m3u8_id='hls', fatal=False)
 345                     for f in m3u8_formats:
 346                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 347                         if m:
 348                             if not f.get('height'):
 349                                 f['height'] = int(m.group('height'))
 350                             if not f.get('tbr'):
 351                                 f['tbr'] = int(m.group('tbr'))
 352                     formats.extend(m3u8_formats)
 353                 else:
 354                     formats.append(add_output_format_meta({
 355                         'url': src,
 356                         'format_id': '%dp' % height if height else None,
 357                         'height': height,
 358                     }, res))
 359
 360             # react rendition since 2017.04.15 (see
 361             # https://github.com/rg3/youtube-dl/issues/12744)
 362             data = self._parse_json(
 363                 self._search_regex(
 364                     r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
 365                     'setup data', default='{}', group='data'), video_id,
 366                 transform_source=unescapeHTML, fatal=False)
 367             if data and isinstance(data, dict):
 368                 extract_formats(data.get('sources'))
 369                 if not duration:
 370                     duration = int_or_none(data.get('duration'))
 371                 extract_subtitles(data.get('tracks'))
 372
 373             if not subtitles and not automatic_captions:
 374                 text_tracks = self._parse_json(
 375                     self._search_regex(
 376                         r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
 377                         'text tracks', default='{}', group='data'), video_id,
 378                     transform_source=lambda s: js_to_json(unescapeHTML(s)),
 379                     fatal=False)
 380                 extract_subtitles(text_tracks)
 381
 382         if not formats and outputs:
 383             for format_id, output in outputs.items():
 384                 f = extract_output_format(output, format_id)
 385                 if f.get('url'):
 386                     formats.append(f)
 387
 388         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 389
 390         return {
 391             'id': video_id,
 392             'title': title,
 393             'description': description,
 394             'thumbnail': thumbnail,
 395             'duration': duration,
 396             'formats': formats,
 397             'subtitles': subtitles,
 398             'automatic_captions': automatic_captions,
 399         }
 400
 401
 402 class UdemyCourseIE(UdemyIE):
 403     IE_NAME = 'udemy:course'
 404     _VALID_URL = r'https?://(?:www\.)?udemy\.com/(?P<id>[^/?#&]+)'
 405     _TESTS = []
 406
 407     @classmethod
 408     def suitable(cls, url):
 409         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 410
 411     def _real_extract(self, url):
 412         course_path = self._match_id(url)
 413
 414         webpage = self._download_webpage(url, course_path)
 415
 416         course_id, title = self._extract_course_info(webpage, course_path)
 417
 418         self._enroll_course(url, webpage, course_id)
 419
 420         response = self._download_json(
 421             'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
 422             course_id, 'Downloading course curriculum', query={
 423                 'fields[chapter]': 'title,object_index',
 424                 'fields[lecture]': 'title,asset',
 425                 'page_size': '1000',
 426             })
 427
 428         entries = []
 429         chapter, chapter_number = [None] * 2
 430         for entry in response['results']:
 431             clazz = entry.get('_class')
 432             if clazz == 'lecture':
 433                 asset = entry.get('asset')
 434                 if isinstance(asset, dict):
 435                     asset_type = asset.get('asset_type') or asset.get('assetType')
 436                     if asset_type != 'Video':
 437                         continue
 438                 lecture_id = entry.get('id')
 439                 if lecture_id:
 440                     entry = {
 441                         '_type': 'url_transparent',
 442                         'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
 443                         'title': entry.get('title'),
 444                         'ie_key': UdemyIE.ie_key(),
 445                     }
 446                     if chapter_number:
 447                         entry['chapter_number'] = chapter_number
 448                     if chapter:
 449                         entry['chapter'] = chapter
 450                     entries.append(entry)
 451             elif clazz == 'chapter':
 452                 chapter_number = entry.get('object_index')
 453                 chapter = entry.get('title')
 454
 455         return self.playlist_result(entries, course_id, title)