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