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     urlencode_postdata,
  24 )
  25
  26
  27 class UdemyIE(InfoExtractor):
  28     IE_NAME = 'udemy'
  29     _VALID_URL = r'''(?x)
  30                     https?://
  31                         www\.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
  68     def _extract_course_info(self, webpage, video_id):
  69         course = self._parse_json(
  70             unescapeHTML(self._search_regex(
  71                 r'ng-init=["\'].*\bcourse=({.+?})[;"\']',
  72                 webpage, 'course', default='{}')),
  73             video_id, fatal=False) or {}
  74         course_id = course.get('id') or self._search_regex(
  75             r'data-course-id=["\'](\d+)', webpage, 'course id')
  76         return course_id, course.get('title')
  77
  78     def _enroll_course(self, base_url, webpage, course_id):
  79         def combine_url(base_url, url):
  80             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  81
  82         checkout_url = unescapeHTML(self._search_regex(
  83             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/(?:payment|cart)/checkout/.+?)\1',
  84             webpage, 'checkout url', group='url', default=None))
  85         if checkout_url:
  86             raise ExtractorError(
  87                 'Course %s is not free. You have to pay for it before you can download. '
  88                 'Use this URL to confirm purchase: %s'
  89                 % (course_id, combine_url(base_url, checkout_url)),
  90                 expected=True)
  91
  92         enroll_url = unescapeHTML(self._search_regex(
  93             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  94             webpage, 'enroll url', group='url', default=None))
  95         if enroll_url:
  96             webpage = self._download_webpage(
  97                 combine_url(base_url, enroll_url),
  98                 course_id, 'Enrolling in the course',
  99                 headers={'Referer': base_url})
 100             if '>You have enrolled in' in webpage:
 101                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
 102
 103     def _download_lecture(self, course_id, lecture_id):
 104         return self._download_json(
 105             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
 106             % (course_id, lecture_id),
 107             lecture_id, 'Downloading lecture JSON', query={
 108                 'fields[lecture]': 'title,description,view_html,asset',
 109                 'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,stream_urls,captions,data',
 110             })
 111
 112     def _handle_error(self, response):
 113         if not isinstance(response, dict):
 114             return
 115         error = response.get('error')
 116         if error:
 117             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
 118             error_data = error.get('data')
 119             if error_data:
 120                 error_str += ' - %s' % error_data.get('formErrors')
 121             raise ExtractorError(error_str, expected=True)
 122
 123     def _download_webpage_handle(self, *args, **kwargs):
 124         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'
 125         return super(UdemyIE, self)._download_webpage_handle(
 126             *args, **compat_kwargs(kwargs))
 127
 128     def _download_json(self, url_or_request, *args, **kwargs):
 129         headers = {
 130             'X-Udemy-Snail-Case': 'true',
 131             'X-Requested-With': 'XMLHttpRequest',
 132         }
 133         for cookie in self._downloader.cookiejar:
 134             if cookie.name == 'client_id':
 135                 headers['X-Udemy-Client-Id'] = cookie.value
 136             elif cookie.name == 'access_token':
 137                 headers['X-Udemy-Bearer-Token'] = cookie.value
 138                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 139
 140         if isinstance(url_or_request, compat_urllib_request.Request):
 141             for header, value in headers.items():
 142                 url_or_request.add_header(header, value)
 143         else:
 144             url_or_request = sanitized_Request(url_or_request, headers=headers)
 145
 146         response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)
 147         self._handle_error(response)
 148         return response
 149
 150     def _real_initialize(self):
 151         self._login()
 152
 153     def _login(self):
 154         username, password = self._get_login_info()
 155         if username is None:
 156             return
 157
 158         login_popup = self._download_webpage(
 159             self._LOGIN_URL, None, 'Downloading login popup')
 160
 161         def is_logged(webpage):
 162             return any(re.search(p, webpage) for p in (
 163                 r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
 164                 r'>Logout<'))
 165
 166         # already logged in
 167         if is_logged(login_popup):
 168             return
 169
 170         login_form = self._form_hidden_inputs('login-form', login_popup)
 171
 172         login_form.update({
 173             'email': username,
 174             'password': password,
 175         })
 176
 177         response = self._download_webpage(
 178             self._LOGIN_URL, None, 'Logging in',
 179             data=urlencode_postdata(login_form),
 180             headers={
 181                 'Referer': self._ORIGIN_URL,
 182                 'Origin': self._ORIGIN_URL,
 183             })
 184
 185         if not is_logged(response):
 186             error = self._html_search_regex(
 187                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 188                 response, 'error message', default=None)
 189             if error:
 190                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 191             raise ExtractorError('Unable to log in')
 192
 193     def _real_extract(self, url):
 194         lecture_id = self._match_id(url)
 195
 196         webpage = self._download_webpage(url, lecture_id)
 197
 198         course_id, _ = self._extract_course_info(webpage, lecture_id)
 199
 200         try:
 201             lecture = self._download_lecture(course_id, lecture_id)
 202         except ExtractorError as e:
 203             # Error could possibly mean we are not enrolled in the course
 204             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 205                 self._enroll_course(url, webpage, course_id)
 206                 lecture = self._download_lecture(course_id, lecture_id)
 207             else:
 208                 raise
 209
 210         title = lecture['title']
 211         description = lecture.get('description')
 212
 213         asset = lecture['asset']
 214
 215         asset_type = asset.get('asset_type') or asset.get('assetType')
 216         if asset_type != 'Video':
 217             raise ExtractorError(
 218                 'Lecture %s is not a video' % lecture_id, expected=True)
 219
 220         stream_url = asset.get('stream_url') or asset.get('streamUrl')
 221         if stream_url:
 222             youtube_url = self._search_regex(
 223                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 224             if youtube_url:
 225                 return self.url_result(youtube_url, 'Youtube')
 226
 227         video_id = compat_str(asset['id'])
 228         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
 229         duration = float_or_none(asset.get('data', {}).get('duration'))
 230
 231         subtitles = {}
 232         automatic_captions = {}
 233
 234         formats = []
 235
 236         def extract_output_format(src, f_id):
 237             return {
 238                 'url': src.get('url'),
 239                 'format_id': '%sp' % (src.get('height') or f_id),
 240                 'width': int_or_none(src.get('width')),
 241                 'height': int_or_none(src.get('height')),
 242                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 243                 'vcodec': src.get('video_codec'),
 244                 'fps': int_or_none(src.get('frame_rate')),
 245                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 246                 'acodec': src.get('audio_codec'),
 247                 'asr': int_or_none(src.get('audio_sample_rate')),
 248                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 249                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 250             }
 251
 252         outputs = asset.get('data', {}).get('outputs')
 253         if not isinstance(outputs, dict):
 254             outputs = {}
 255
 256         def add_output_format_meta(f, key):
 257             output = outputs.get(key)
 258             if isinstance(output, dict):
 259                 output_format = extract_output_format(output, key)
 260                 output_format.update(f)
 261                 return output_format
 262             return f
 263
 264         def extract_formats(source_list):
 265             if not isinstance(source_list, list):
 266                 return
 267             for source in source_list:
 268                 video_url = source.get('file') or source.get('src')
 269                 if not video_url or not isinstance(video_url, compat_str):
 270                     continue
 271                 if source.get('type') == 'application/x-mpegURL' or determine_ext(video_url) == 'm3u8':
 272                     formats.extend(self._extract_m3u8_formats(
 273                         video_url, video_id, 'mp4', entry_protocol='m3u8_native',
 274                         m3u8_id='hls', fatal=False))
 275                     continue
 276                 format_id = source.get('label')
 277                 f = {
 278                     'url': video_url,
 279                     'format_id': '%sp' % format_id,
 280                     'height': int_or_none(format_id),
 281                 }
 282                 if format_id:
 283                     # Some videos contain additional metadata (e.g.
 284                     # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 285                     f = add_output_format_meta(f, format_id)
 286                 formats.append(f)
 287
 288         def extract_subtitles(track_list):
 289             if not isinstance(track_list, list):
 290                 return
 291             for track in track_list:
 292                 if not isinstance(track, dict):
 293                     continue
 294                 if track.get('kind') != 'captions':
 295                     continue
 296                 src = track.get('src')
 297                 if not src or not isinstance(src, compat_str):
 298                     continue
 299                 lang = track.get('language') or track.get(
 300                     'srclang') or track.get('label')
 301                 sub_dict = automatic_captions if track.get(
 302                     'autogenerated') is True else subtitles
 303                 sub_dict.setdefault(lang, []).append({
 304                     'url': src,
 305                 })
 306
 307         for url_kind in ('download', 'stream'):
 308             urls = asset.get('%s_urls' % url_kind)
 309             if isinstance(urls, dict):
 310                 extract_formats(urls.get('Video'))
 311
 312         captions = asset.get('captions')
 313         if isinstance(captions, list):
 314             for cc in captions:
 315                 if not isinstance(cc, dict):
 316                     continue
 317                 cc_url = cc.get('url')
 318                 if not cc_url or not isinstance(cc_url, compat_str):
 319                     continue
 320                 lang = try_get(cc, lambda x: x['locale']['locale'], compat_str)
 321                 sub_dict = (automatic_captions if cc.get('source') == 'auto'
 322                             else subtitles)
 323                 sub_dict.setdefault(lang or 'en', []).append({
 324                     'url': cc_url,
 325                 })
 326
 327         view_html = lecture.get('view_html')
 328         if view_html:
 329             view_html_urls = set()
 330             for source in re.findall(r'<source[^>]+>', view_html):
 331                 attributes = extract_attributes(source)
 332                 src = attributes.get('src')
 333                 if not src:
 334                     continue
 335                 res = attributes.get('data-res')
 336                 height = int_or_none(res)
 337                 if src in view_html_urls:
 338                     continue
 339                 view_html_urls.add(src)
 340                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 341                     m3u8_formats = self._extract_m3u8_formats(
 342                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 343                         m3u8_id='hls', fatal=False)
 344                     for f in m3u8_formats:
 345                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 346                         if m:
 347                             if not f.get('height'):
 348                                 f['height'] = int(m.group('height'))
 349                             if not f.get('tbr'):
 350                                 f['tbr'] = int(m.group('tbr'))
 351                     formats.extend(m3u8_formats)
 352                 else:
 353                     formats.append(add_output_format_meta({
 354                         'url': src,
 355                         'format_id': '%dp' % height if height else None,
 356                         'height': height,
 357                     }, res))
 358
 359             # react rendition since 2017.04.15 (see
 360             # https://github.com/rg3/youtube-dl/issues/12744)
 361             data = self._parse_json(
 362                 self._search_regex(
 363                     r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
 364                     'setup data', default='{}', group='data'), video_id,
 365                 transform_source=unescapeHTML, fatal=False)
 366             if data and isinstance(data, dict):
 367                 extract_formats(data.get('sources'))
 368                 if not duration:
 369                     duration = int_or_none(data.get('duration'))
 370                 extract_subtitles(data.get('tracks'))
 371
 372             if not subtitles and not automatic_captions:
 373                 text_tracks = self._parse_json(
 374                     self._search_regex(
 375                         r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
 376                         'text tracks', default='{}', group='data'), video_id,
 377                     transform_source=lambda s: js_to_json(unescapeHTML(s)),
 378                     fatal=False)
 379                 extract_subtitles(text_tracks)
 380
 381         if not formats and outputs:
 382             for format_id, output in outputs.items():
 383                 f = extract_output_format(output, format_id)
 384                 if f.get('url'):
 385                     formats.append(f)
 386
 387         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 388
 389         return {
 390             'id': video_id,
 391             'title': title,
 392             'description': description,
 393             'thumbnail': thumbnail,
 394             'duration': duration,
 395             'formats': formats,
 396             'subtitles': subtitles,
 397             'automatic_captions': automatic_captions,
 398         }
 399
 400
 401 class UdemyCourseIE(UdemyIE):
 402     IE_NAME = 'udemy:course'
 403     _VALID_URL = r'https?://(?:www\.)?udemy\.com/(?P<id>[^/?#&]+)'
 404     _TESTS = []
 405
 406     @classmethod
 407     def suitable(cls, url):
 408         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 409
 410     def _real_extract(self, url):
 411         course_path = self._match_id(url)
 412
 413         webpage = self._download_webpage(url, course_path)
 414
 415         course_id, title = self._extract_course_info(webpage, course_path)
 416
 417         self._enroll_course(url, webpage, course_id)
 418
 419         response = self._download_json(
 420             'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
 421             course_id, 'Downloading course curriculum', query={
 422                 'fields[chapter]': 'title,object_index',
 423                 'fields[lecture]': 'title,asset',
 424                 'page_size': '1000',
 425             })
 426
 427         entries = []
 428         chapter, chapter_number = [None] * 2
 429         for entry in response['results']:
 430             clazz = entry.get('_class')
 431             if clazz == 'lecture':
 432                 asset = entry.get('asset')
 433                 if isinstance(asset, dict):
 434                     asset_type = asset.get('asset_type') or asset.get('assetType')
 435                     if asset_type != 'Video':
 436                         continue
 437                 lecture_id = entry.get('id')
 438                 if lecture_id:
 439                     entry = {
 440                         '_type': 'url_transparent',
 441                         'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
 442                         'title': entry.get('title'),
 443                         'ie_key': UdemyIE.ie_key(),
 444                     }
 445                     if chapter_number:
 446                         entry['chapter_number'] = chapter_number
 447                     if chapter:
 448                         entry['chapter'] = chapter
 449                     entries.append(entry)
 450             elif clazz == 'chapter':
 451                 chapter_number = entry.get('object_index')
 452                 chapter = entry.get('title')
 453
 454         return self.playlist_result(entries, course_id, title)