_ Git - youtube-dl/blob - youtube_dl/extractor/udemy.py

   1 from __future__ import unicode_literals
   2
   3 from .common import InfoExtractor
   4 from ..compat import (
   5     compat_HTTPError,
   6     compat_urllib_parse,
   7     compat_urllib_request,
   8     compat_urlparse,
   9 )
  10 from ..utils import (
  11     ExtractorError,
  12     float_or_none,
  13     int_or_none,
  14     sanitized_Request,
  15     unescapeHTML,
  16 )
  17
  18
  19 class UdemyIE(InfoExtractor):
  20     IE_NAME = 'udemy'
  21     _VALID_URL = r'''(?x)
  22                     https?://
  23                         www\.udemy\.com/
  24                         (?:
  25                             [^#]+\#/lecture/|
  26                             lecture/view/?\?lectureId=|
  27                             [^/]+/learn/v4/t/lecture/
  28                         )
  29                         (?P<id>\d+)
  30                     '''
  31     _LOGIN_URL = 'https://www.udemy.com/join/login-popup/?displayType=ajax&showSkipButton=1'
  32     _ORIGIN_URL = 'https://www.udemy.com'
  33     _NETRC_MACHINE = 'udemy'
  34
  35     _TESTS = [{
  36         'url': 'https://www.udemy.com/java-tutorial/#/lecture/172757',
  37         'md5': '98eda5b657e752cf945d8445e261b5c5',
  38         'info_dict': {
  39             'id': '160614',
  40             'ext': 'mp4',
  41             'title': 'Introduction and Installation',
  42             'description': 'md5:c0d51f6f21ef4ec65f091055a5eef876',
  43             'duration': 579.29,
  44         },
  45         'skip': 'Requires udemy account credentials',
  46     }, {
  47         # new URL schema
  48         'url': 'https://www.udemy.com/electric-bass-right-from-the-start/learn/v4/t/lecture/4580906',
  49         'only_matching': True,
  50     }]
  51
  52     def _enroll_course(self, base_url, webpage, course_id):
  53         checkout_url = unescapeHTML(self._search_regex(
  54             r'href=(["\'])(?P<url>https?://(?:www\.)?udemy\.com/payment/checkout/.+?)\1',
  55             webpage, 'checkout url', group='url', default=None))
  56         if checkout_url:
  57             raise ExtractorError(
  58                 'Course %s is not free. You have to pay for it before you can download. '
  59                 'Use this URL to confirm purchase: %s' % (course_id, checkout_url), expected=True)
  60
  61         enroll_url = unescapeHTML(self._search_regex(
  62             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  63             webpage, 'enroll url', group='url', default=None))
  64         if enroll_url:
  65             if not enroll_url.startswith('http'):
  66                 enroll_url = compat_urlparse.urljoin(base_url, enroll_url)
  67             webpage = self._download_webpage(enroll_url, course_id, 'Enrolling in the course')
  68             if '>You have enrolled in' in webpage:
  69                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
  70
  71     def _download_lecture(self, course_id, lecture_id):
  72         return self._download_json(
  73             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?%s' % (
  74                 course_id, lecture_id, compat_urllib_parse.urlencode({
  75                     'video_only': '',
  76                     'auto_play': '',
  77                     'fields[lecture]': 'title,description,asset',
  78                     'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,data',
  79                     'instructorPreviewMode': 'False',
  80                 })),
  81             lecture_id, 'Downloading lecture JSON')
  82
  83     def _handle_error(self, response):
  84         if not isinstance(response, dict):
  85             return
  86         error = response.get('error')
  87         if error:
  88             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
  89             error_data = error.get('data')
  90             if error_data:
  91                 error_str += ' - %s' % error_data.get('formErrors')
  92             raise ExtractorError(error_str, expected=True)
  93
  94     def _download_json(self, url_or_request, video_id, note='Downloading JSON metadata'):
  95         headers = {
  96             'X-Udemy-Snail-Case': 'true',
  97             'X-Requested-With': 'XMLHttpRequest',
  98         }
  99         for cookie in self._downloader.cookiejar:
 100             if cookie.name == 'client_id':
 101                 headers['X-Udemy-Client-Id'] = cookie.value
 102             elif cookie.name == 'access_token':
 103                 headers['X-Udemy-Bearer-Token'] = cookie.value
 104                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 105
 106         if isinstance(url_or_request, compat_urllib_request.Request):
 107             for header, value in headers.items():
 108                 url_or_request.add_header(header, value)
 109         else:
 110             url_or_request = sanitized_Request(url_or_request, headers=headers)
 111
 112         response = super(UdemyIE, self)._download_json(url_or_request, video_id, note)
 113         self._handle_error(response)
 114         return response
 115
 116     def _real_initialize(self):
 117         self._login()
 118
 119     def _login(self):
 120         (username, password) = self._get_login_info()
 121         if username is None:
 122             return
 123
 124         login_popup = self._download_webpage(
 125             self._LOGIN_URL, None, 'Downloading login popup')
 126
 127         def is_logged(webpage):
 128             return any(p in webpage for p in ['href="https://www.udemy.com/user/logout/', '>Logout<'])
 129
 130         # already logged in
 131         if is_logged(login_popup):
 132             return
 133
 134         login_form = self._form_hidden_inputs('login-form', login_popup)
 135
 136         login_form.update({
 137             'email': username.encode('utf-8'),
 138             'password': password.encode('utf-8'),
 139         })
 140
 141         request = sanitized_Request(
 142             self._LOGIN_URL, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
 143         request.add_header('Referer', self._ORIGIN_URL)
 144         request.add_header('Origin', self._ORIGIN_URL)
 145
 146         response = self._download_webpage(
 147             request, None, 'Logging in as %s' % username)
 148
 149         if not is_logged(response):
 150             error = self._html_search_regex(
 151                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 152                 response, 'error message', default=None)
 153             if error:
 154                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 155             raise ExtractorError('Unable to log in')
 156
 157     def _real_extract(self, url):
 158         lecture_id = self._match_id(url)
 159
 160         webpage = self._download_webpage(url, lecture_id)
 161
 162         course_id = self._search_regex(
 163             (r'data-course-id=["\'](\d+)', r'&quot;id&quot;\s*:\s*(\d+)'),
 164             webpage, 'course id')
 165
 166         try:
 167             lecture = self._download_lecture(course_id, lecture_id)
 168         except ExtractorError as e:
 169             # Error could possibly mean we are not enrolled in the course
 170             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 171                 self._enroll_course(url, webpage, course_id)
 172                 lecture = self._download_lecture(course_id, lecture_id)
 173             else:
 174                 raise
 175
 176         title = lecture['title']
 177         description = lecture.get('description')
 178
 179         asset = lecture['asset']
 180
 181         asset_type = asset.get('assetType') or asset.get('asset_type')
 182         if asset_type != 'Video':
 183             raise ExtractorError(
 184                 'Lecture %s is not a video' % lecture_id, expected=True)
 185
 186         stream_url = asset.get('streamUrl') or asset.get('stream_url')
 187         if stream_url:
 188             youtube_url = self._search_regex(
 189                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 190             if youtube_url:
 191                 return self.url_result(youtube_url, 'Youtube')
 192
 193         video_id = asset['id']
 194         thumbnail = asset.get('thumbnailUrl') or asset.get('thumbnail_url')
 195         duration = float_or_none(asset.get('data', {}).get('duration'))
 196
 197         formats = []
 198
 199         def extract_output_format(src):
 200             return {
 201                 'url': src['url'],
 202                 'format_id': '%sp' % (src.get('label') or format_id),
 203                 'width': int_or_none(src.get('width')),
 204                 'height': int_or_none(src.get('height')),
 205                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 206                 'vcodec': src.get('video_codec'),
 207                 'fps': int_or_none(src.get('frame_rate')),
 208                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 209                 'acodec': src.get('audio_codec'),
 210                 'asr': int_or_none(src.get('audio_sample_rate')),
 211                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 212                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 213             }
 214
 215         outputs = asset.get('data', {}).get('outputs')
 216         if not isinstance(outputs, dict):
 217             outputs = {}
 218
 219         for format_id, output in outputs.items():
 220             if isinstance(output, dict) and output.get('url'):
 221                 formats.append(extract_output_format(output))
 222
 223         download_urls = asset.get('download_urls')
 224         if isinstance(download_urls, dict):
 225             video = download_urls.get('Video')
 226             if isinstance(video, list):
 227                 for format_ in video:
 228                     video_url = format_.get('file')
 229                     if not video_url:
 230                         continue
 231                     format_id = format_.get('label')
 232                     f = {
 233                         'url': format_['file'],
 234                         'height': int_or_none(format_id),
 235                     }
 236                     if format_id:
 237                         # Some videos contain additional metadata (e.g.
 238                         # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 239                         output = outputs.get(format_id)
 240                         if isinstance(output, dict):
 241                             output_format = extract_output_format(output)
 242                             output_format.update(f)
 243                             f = output_format
 244                         else:
 245                             f['format_id'] = '%sp' % format_id
 246                     formats.append(f)
 247
 248         self._sort_formats(formats)
 249
 250         return {
 251             'id': video_id,
 252             'title': title,
 253             'description': description,
 254             'thumbnail': thumbnail,
 255             'duration': duration,
 256             'formats': formats
 257         }
 258
 259
 260 class UdemyCourseIE(UdemyIE):
 261     IE_NAME = 'udemy:course'
 262     _VALID_URL = r'https?://www\.udemy\.com/(?P<id>[\da-z-]+)'
 263     _TESTS = []
 264
 265     @classmethod
 266     def suitable(cls, url):
 267         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 268
 269     def _real_extract(self, url):
 270         course_path = self._match_id(url)
 271
 272         webpage = self._download_webpage(url, course_path)
 273
 274         response = self._download_json(
 275             'https://www.udemy.com/api-1.1/courses/%s' % course_path,
 276             course_path, 'Downloading course JSON')
 277
 278         course_id = response['id']
 279         course_title = response.get('title')
 280
 281         self._enroll_course(url, webpage, course_id)
 282
 283         response = self._download_json(
 284             'https://www.udemy.com/api-1.1/courses/%s/curriculum' % course_id,
 285             course_id, 'Downloading course curriculum')
 286
 287         entries = []
 288         chapter, chapter_number = None, None
 289         for asset in response:
 290             asset_type = asset.get('assetType') or asset.get('asset_type')
 291             if asset_type == 'Video':
 292                 asset_id = asset.get('id')
 293                 if asset_id:
 294                     entry = {
 295                         '_type': 'url_transparent',
 296                         'url': 'https://www.udemy.com/%s/#/lecture/%s' % (course_path, asset['id']),
 297                         'ie_key': UdemyIE.ie_key(),
 298                     }
 299                     if chapter_number:
 300                         entry['chapter_number'] = chapter_number
 301                     if chapter:
 302                         entry['chapter'] = chapter
 303                     entries.append(entry)
 304             elif asset.get('type') == 'chapter':
 305                 chapter_number = asset.get('index') or asset.get('object_index')
 306                 chapter = asset.get('title')
 307
 308         return self.playlist_result(entries, course_id, course_title)