_ 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_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 _enroll_course(self, base_url, webpage, course_id):
  58         checkout_url = unescapeHTML(self._search_regex(
  59             r'href=(["\'])(?P<url>https?://(?:www\.)?udemy\.com/payment/checkout/.+?)\1',
  60             webpage, 'checkout url', group='url', default=None))
  61         if checkout_url:
  62             raise ExtractorError(
  63                 'Course %s is not free. You have to pay for it before you can download. '
  64                 'Use this URL to confirm purchase: %s' % (course_id, checkout_url), expected=True)
  65
  66         enroll_url = unescapeHTML(self._search_regex(
  67             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  68             webpage, 'enroll url', group='url', default=None))
  69         if enroll_url:
  70             if not enroll_url.startswith('http'):
  71                 enroll_url = compat_urlparse.urljoin(base_url, enroll_url)
  72             webpage = self._download_webpage(enroll_url, course_id, 'Enrolling in the course')
  73             if '>You have enrolled in' in webpage:
  74                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
  75
  76     def _download_lecture(self, course_id, lecture_id):
  77         return self._download_json(
  78             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?%s' % (
  79                 course_id, lecture_id, compat_urllib_parse_urlencode({
  80                     'fields[lecture]': 'title,description,view_html,asset',
  81                     'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,data',
  82                 })),
  83             lecture_id, 'Downloading lecture JSON')
  84
  85     def _handle_error(self, response):
  86         if not isinstance(response, dict):
  87             return
  88         error = response.get('error')
  89         if error:
  90             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
  91             error_data = error.get('data')
  92             if error_data:
  93                 error_str += ' - %s' % error_data.get('formErrors')
  94             raise ExtractorError(error_str, expected=True)
  95
  96     def _download_json(self, url_or_request, video_id, note='Downloading JSON metadata'):
  97         headers = {
  98             'X-Udemy-Snail-Case': 'true',
  99             'X-Requested-With': 'XMLHttpRequest',
 100         }
 101         for cookie in self._downloader.cookiejar:
 102             if cookie.name == 'client_id':
 103                 headers['X-Udemy-Client-Id'] = cookie.value
 104             elif cookie.name == 'access_token':
 105                 headers['X-Udemy-Bearer-Token'] = cookie.value
 106                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 107
 108         if isinstance(url_or_request, compat_urllib_request.Request):
 109             for header, value in headers.items():
 110                 url_or_request.add_header(header, value)
 111         else:
 112             url_or_request = sanitized_Request(url_or_request, headers=headers)
 113
 114         response = super(UdemyIE, self)._download_json(url_or_request, video_id, note)
 115         self._handle_error(response)
 116         return response
 117
 118     def _real_initialize(self):
 119         self._login()
 120
 121     def _login(self):
 122         (username, password) = self._get_login_info()
 123         if username is None:
 124             return
 125
 126         login_popup = self._download_webpage(
 127             self._LOGIN_URL, None, 'Downloading login popup')
 128
 129         def is_logged(webpage):
 130             return any(p in webpage for p in ['href="https://www.udemy.com/user/logout/', '>Logout<'])
 131
 132         # already logged in
 133         if is_logged(login_popup):
 134             return
 135
 136         login_form = self._form_hidden_inputs('login-form', login_popup)
 137
 138         login_form.update({
 139             'email': username.encode('utf-8'),
 140             'password': password.encode('utf-8'),
 141         })
 142
 143         request = sanitized_Request(
 144             self._LOGIN_URL, urlencode_postdata(login_form))
 145         request.add_header('Referer', self._ORIGIN_URL)
 146         request.add_header('Origin', self._ORIGIN_URL)
 147
 148         response = self._download_webpage(
 149             request, None, 'Logging in as %s' % username)
 150
 151         if not is_logged(response):
 152             error = self._html_search_regex(
 153                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 154                 response, 'error message', default=None)
 155             if error:
 156                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 157             raise ExtractorError('Unable to log in')
 158
 159     def _real_extract(self, url):
 160         lecture_id = self._match_id(url)
 161
 162         webpage = self._download_webpage(url, lecture_id)
 163
 164         course_id = self._search_regex(
 165             (r'data-course-id=["\'](\d+)', r'&quot;id&quot;\s*:\s*(\d+)'),
 166             webpage, 'course id')
 167
 168         try:
 169             lecture = self._download_lecture(course_id, lecture_id)
 170         except ExtractorError as e:
 171             # Error could possibly mean we are not enrolled in the course
 172             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 173                 self._enroll_course(url, webpage, course_id)
 174                 lecture = self._download_lecture(course_id, lecture_id)
 175             else:
 176                 raise
 177
 178         title = lecture['title']
 179         description = lecture.get('description')
 180
 181         asset = lecture['asset']
 182
 183         asset_type = asset.get('assetType') or asset.get('asset_type')
 184         if asset_type != 'Video':
 185             raise ExtractorError(
 186                 'Lecture %s is not a video' % lecture_id, expected=True)
 187
 188         stream_url = asset.get('streamUrl') or asset.get('stream_url')
 189         if stream_url:
 190             youtube_url = self._search_regex(
 191                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 192             if youtube_url:
 193                 return self.url_result(youtube_url, 'Youtube')
 194
 195         video_id = asset['id']
 196         thumbnail = asset.get('thumbnailUrl') or asset.get('thumbnail_url')
 197         duration = float_or_none(asset.get('data', {}).get('duration'))
 198
 199         formats = []
 200
 201         def extract_output_format(src):
 202             return {
 203                 'url': src['url'],
 204                 'format_id': '%sp' % (src.get('height') or format_id),
 205                 'width': int_or_none(src.get('width')),
 206                 'height': int_or_none(src.get('height')),
 207                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 208                 'vcodec': src.get('video_codec'),
 209                 'fps': int_or_none(src.get('frame_rate')),
 210                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 211                 'acodec': src.get('audio_codec'),
 212                 'asr': int_or_none(src.get('audio_sample_rate')),
 213                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 214                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 215             }
 216
 217         outputs = asset.get('data', {}).get('outputs')
 218         if not isinstance(outputs, dict):
 219             outputs = {}
 220
 221         def add_output_format_meta(f, key, format_id):
 222             output = outputs.get(key)
 223             if isinstance(output, dict):
 224                 output_format = extract_output_format(output)
 225                 output_format.update(f)
 226                 return output_format
 227             else:
 228                 f['format_id'] = format_id
 229                 return f
 230
 231         download_urls = asset.get('download_urls')
 232         if isinstance(download_urls, dict):
 233             video = download_urls.get('Video')
 234             if isinstance(video, list):
 235                 for format_ in video:
 236                     video_url = format_.get('file')
 237                     if not video_url:
 238                         continue
 239                     format_id = format_.get('label')
 240                     f = {
 241                         'url': format_['file'],
 242                         'height': int_or_none(format_id),
 243                     }
 244                     if format_id:
 245                         # Some videos contain additional metadata (e.g.
 246                         # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 247                         f = add_output_format_meta(f, format_id, '%sp' % format_id)
 248                     formats.append(f)
 249
 250         view_html = lecture.get('view_html')
 251         if view_html:
 252             view_html_urls = set()
 253             for source in re.findall(r'<source[^>]+>', view_html):
 254                 attributes = extract_attributes(source)
 255                 src = attributes.get('src')
 256                 if not src:
 257                     continue
 258                 res = attributes.get('data-res')
 259                 height = int_or_none(res)
 260                 if src in view_html_urls:
 261                     continue
 262                 view_html_urls.add(src)
 263                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 264                     m3u8_formats = self._extract_m3u8_formats(
 265                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 266                         m3u8_id='hls', fatal=False)
 267                     for f in m3u8_formats:
 268                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 269                         if m:
 270                             if not f.get('height'):
 271                                 f['height'] = int(m.group('height'))
 272                             if not f.get('tbr'):
 273                                 f['tbr'] = int(m.group('tbr'))
 274                     formats.extend(m3u8_formats)
 275                 else:
 276                     formats.append(add_output_format_meta({
 277                         'url': src,
 278                         'height': height,
 279                     }, res, '%dp' % height if height else None))
 280
 281         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 282
 283         return {
 284             'id': video_id,
 285             'title': title,
 286             'description': description,
 287             'thumbnail': thumbnail,
 288             'duration': duration,
 289             'formats': formats
 290         }
 291
 292
 293 class UdemyCourseIE(UdemyIE):
 294     IE_NAME = 'udemy:course'
 295     _VALID_URL = r'https?://www\.udemy\.com/(?P<id>[\da-z-]+)'
 296     _TESTS = []
 297
 298     @classmethod
 299     def suitable(cls, url):
 300         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 301
 302     def _real_extract(self, url):
 303         course_path = self._match_id(url)
 304
 305         webpage = self._download_webpage(url, course_path)
 306
 307         response = self._download_json(
 308             'https://www.udemy.com/api-1.1/courses/%s' % course_path,
 309             course_path, 'Downloading course JSON')
 310
 311         course_id = response['id']
 312         course_title = response.get('title')
 313
 314         self._enroll_course(url, webpage, course_id)
 315
 316         response = self._download_json(
 317             'https://www.udemy.com/api-1.1/courses/%s/curriculum' % course_id,
 318             course_id, 'Downloading course curriculum')
 319
 320         entries = []
 321         chapter, chapter_number = None, None
 322         for asset in response:
 323             asset_type = asset.get('assetType') or asset.get('asset_type')
 324             if asset_type == 'Video':
 325                 asset_id = asset.get('id')
 326                 if asset_id:
 327                     entry = {
 328                         '_type': 'url_transparent',
 329                         'url': 'https://www.udemy.com/%s/#/lecture/%s' % (course_path, asset['id']),
 330                         'ie_key': UdemyIE.ie_key(),
 331                     }
 332                     if chapter_number:
 333                         entry['chapter_number'] = chapter_number
 334                     if chapter:
 335                         entry['chapter'] = chapter
 336                     entries.append(entry)
 337             elif asset.get('type') == 'chapter':
 338                 chapter_number = asset.get('index') or asset.get('object_index')
 339                 chapter = asset.get('title')
 340
 341         return self.playlist_result(entries, course_id, course_title)