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