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