_ 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         def combine_url(base_url, url):
  59             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  60
  61         checkout_url = unescapeHTML(self._search_regex(
  62             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/payment/checkout/.+?)\1',
  63             webpage, 'checkout url', group='url', default=None))
  64         if checkout_url:
  65             raise ExtractorError(
  66                 'Course %s is not free. You have to pay for it before you can download. '
  67                 'Use this URL to confirm purchase: %s'
  68                 % (course_id, combine_url(base_url, checkout_url)),
  69                 expected=True)
  70
  71         enroll_url = unescapeHTML(self._search_regex(
  72             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  73             webpage, 'enroll url', group='url', default=None))
  74         if enroll_url:
  75             webpage = self._download_webpage(
  76                 combine_url(base_url, enroll_url),
  77                 course_id, 'Enrolling in the course')
  78             if '>You have enrolled in' in webpage:
  79                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
  80
  81     def _download_lecture(self, course_id, lecture_id):
  82         return self._download_json(
  83             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?%s' % (
  84                 course_id, lecture_id, compat_urllib_parse_urlencode({
  85                     'fields[lecture]': 'title,description,view_html,asset',
  86                     'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,data',
  87                 })),
  88             lecture_id, 'Downloading lecture JSON')
  89
  90     def _handle_error(self, response):
  91         if not isinstance(response, dict):
  92             return
  93         error = response.get('error')
  94         if error:
  95             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
  96             error_data = error.get('data')
  97             if error_data:
  98                 error_str += ' - %s' % error_data.get('formErrors')
  99             raise ExtractorError(error_str, expected=True)
 100
 101     def _download_json(self, url_or_request, video_id, note='Downloading JSON metadata'):
 102         headers = {
 103             'X-Udemy-Snail-Case': 'true',
 104             'X-Requested-With': 'XMLHttpRequest',
 105         }
 106         for cookie in self._downloader.cookiejar:
 107             if cookie.name == 'client_id':
 108                 headers['X-Udemy-Client-Id'] = cookie.value
 109             elif cookie.name == 'access_token':
 110                 headers['X-Udemy-Bearer-Token'] = cookie.value
 111                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 112
 113         if isinstance(url_or_request, compat_urllib_request.Request):
 114             for header, value in headers.items():
 115                 url_or_request.add_header(header, value)
 116         else:
 117             url_or_request = sanitized_Request(url_or_request, headers=headers)
 118
 119         response = super(UdemyIE, self)._download_json(url_or_request, video_id, note)
 120         self._handle_error(response)
 121         return response
 122
 123     def _real_initialize(self):
 124         self._login()
 125
 126     def _login(self):
 127         (username, password) = self._get_login_info()
 128         if username is None:
 129             return
 130
 131         login_popup = self._download_webpage(
 132             self._LOGIN_URL, None, 'Downloading login popup')
 133
 134         def is_logged(webpage):
 135             return any(p in webpage for p in ['href="https://www.udemy.com/user/logout/', '>Logout<'])
 136
 137         # already logged in
 138         if is_logged(login_popup):
 139             return
 140
 141         login_form = self._form_hidden_inputs('login-form', login_popup)
 142
 143         login_form.update({
 144             'email': username.encode('utf-8'),
 145             'password': password.encode('utf-8'),
 146         })
 147
 148         request = sanitized_Request(
 149             self._LOGIN_URL, urlencode_postdata(login_form))
 150         request.add_header('Referer', self._ORIGIN_URL)
 151         request.add_header('Origin', self._ORIGIN_URL)
 152
 153         response = self._download_webpage(
 154             request, None, 'Logging in as %s' % username)
 155
 156         if not is_logged(response):
 157             error = self._html_search_regex(
 158                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 159                 response, 'error message', default=None)
 160             if error:
 161                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 162             raise ExtractorError('Unable to log in')
 163
 164     def _real_extract(self, url):
 165         lecture_id = self._match_id(url)
 166
 167         webpage = self._download_webpage(url, lecture_id)
 168
 169         course_id = self._search_regex(
 170             (r'data-course-id=["\'](\d+)', r'&quot;id&quot;\s*:\s*(\d+)'),
 171             webpage, 'course id')
 172
 173         try:
 174             lecture = self._download_lecture(course_id, lecture_id)
 175         except ExtractorError as e:
 176             # Error could possibly mean we are not enrolled in the course
 177             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 178                 self._enroll_course(url, webpage, course_id)
 179                 lecture = self._download_lecture(course_id, lecture_id)
 180             else:
 181                 raise
 182
 183         title = lecture['title']
 184         description = lecture.get('description')
 185
 186         asset = lecture['asset']
 187
 188         asset_type = asset.get('assetType') or asset.get('asset_type')
 189         if asset_type != 'Video':
 190             raise ExtractorError(
 191                 'Lecture %s is not a video' % lecture_id, expected=True)
 192
 193         stream_url = asset.get('streamUrl') or asset.get('stream_url')
 194         if stream_url:
 195             youtube_url = self._search_regex(
 196                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 197             if youtube_url:
 198                 return self.url_result(youtube_url, 'Youtube')
 199
 200         video_id = asset['id']
 201         thumbnail = asset.get('thumbnailUrl') or asset.get('thumbnail_url')
 202         duration = float_or_none(asset.get('data', {}).get('duration'))
 203
 204         formats = []
 205
 206         def extract_output_format(src):
 207             return {
 208                 'url': src['url'],
 209                 'format_id': '%sp' % (src.get('height') or format_id),
 210                 'width': int_or_none(src.get('width')),
 211                 'height': int_or_none(src.get('height')),
 212                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 213                 'vcodec': src.get('video_codec'),
 214                 'fps': int_or_none(src.get('frame_rate')),
 215                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 216                 'acodec': src.get('audio_codec'),
 217                 'asr': int_or_none(src.get('audio_sample_rate')),
 218                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 219                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 220             }
 221
 222         outputs = asset.get('data', {}).get('outputs')
 223         if not isinstance(outputs, dict):
 224             outputs = {}
 225
 226         def add_output_format_meta(f, key):
 227             output = outputs.get(key)
 228             if isinstance(output, dict):
 229                 output_format = extract_output_format(output)
 230                 output_format.update(f)
 231                 return output_format
 232             return f
 233
 234         download_urls = asset.get('download_urls')
 235         if isinstance(download_urls, dict):
 236             video = download_urls.get('Video')
 237             if isinstance(video, list):
 238                 for format_ in video:
 239                     video_url = format_.get('file')
 240                     if not video_url:
 241                         continue
 242                     format_id = format_.get('label')
 243                     f = {
 244                         'url': format_['file'],
 245                         'format_id': '%sp' % format_id,
 246                         'height': int_or_none(format_id),
 247                     }
 248                     if format_id:
 249                         # Some videos contain additional metadata (e.g.
 250                         # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 251                         f = add_output_format_meta(f, format_id)
 252                     formats.append(f)
 253
 254         view_html = lecture.get('view_html')
 255         if view_html:
 256             view_html_urls = set()
 257             for source in re.findall(r'<source[^>]+>', view_html):
 258                 attributes = extract_attributes(source)
 259                 src = attributes.get('src')
 260                 if not src:
 261                     continue
 262                 res = attributes.get('data-res')
 263                 height = int_or_none(res)
 264                 if src in view_html_urls:
 265                     continue
 266                 view_html_urls.add(src)
 267                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 268                     m3u8_formats = self._extract_m3u8_formats(
 269                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 270                         m3u8_id='hls', fatal=False)
 271                     for f in m3u8_formats:
 272                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 273                         if m:
 274                             if not f.get('height'):
 275                                 f['height'] = int(m.group('height'))
 276                             if not f.get('tbr'):
 277                                 f['tbr'] = int(m.group('tbr'))
 278                     formats.extend(m3u8_formats)
 279                 else:
 280                     formats.append(add_output_format_meta({
 281                         'url': src,
 282                         'format_id': '%dp' % height if height else None,
 283                         'height': height,
 284                     }, res))
 285
 286         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 287
 288         return {
 289             'id': video_id,
 290             'title': title,
 291             'description': description,
 292             'thumbnail': thumbnail,
 293             'duration': duration,
 294             'formats': formats
 295         }
 296
 297
 298 class UdemyCourseIE(UdemyIE):
 299     IE_NAME = 'udemy:course'
 300     _VALID_URL = r'https?://www\.udemy\.com/(?P<id>[\da-z-]+)'
 301     _TESTS = []
 302
 303     @classmethod
 304     def suitable(cls, url):
 305         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 306
 307     def _real_extract(self, url):
 308         course_path = self._match_id(url)
 309
 310         webpage = self._download_webpage(url, course_path)
 311
 312         response = self._download_json(
 313             'https://www.udemy.com/api-1.1/courses/%s' % course_path,
 314             course_path, 'Downloading course JSON')
 315
 316         course_id = response['id']
 317         course_title = response.get('title')
 318
 319         self._enroll_course(url, webpage, course_id)
 320
 321         response = self._download_json(
 322             'https://www.udemy.com/api-1.1/courses/%s/curriculum' % course_id,
 323             course_id, 'Downloading course curriculum')
 324
 325         entries = []
 326         chapter, chapter_number = None, None
 327         for asset in response:
 328             asset_type = asset.get('assetType') or asset.get('asset_type')
 329             if asset_type == 'Video':
 330                 asset_id = asset.get('id')
 331                 if asset_id:
 332                     entry = {
 333                         '_type': 'url_transparent',
 334                         'url': 'https://www.udemy.com/%s/#/lecture/%s' % (course_path, asset['id']),
 335                         'ie_key': UdemyIE.ie_key(),
 336                     }
 337                     if chapter_number:
 338                         entry['chapter_number'] = chapter_number
 339                     if chapter:
 340                         entry['chapter'] = chapter
 341                     entries.append(entry)
 342             elif asset.get('type') == 'chapter':
 343                 chapter_number = asset.get('index') or asset.get('object_index')
 344                 chapter = asset.get('title')
 345
 346         return self.playlist_result(entries, course_id, course_title)