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