Raphaël G. Git Repositories - youtubedl/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_request,
   9     compat_urlparse,
  10 )
  11 from ..utils import (
  12     determine_ext,
  13     extract_attributes,
  14     ExtractorError,
  15     float_or_none,
  16     int_or_none,
  17     sanitized_Request,
  18     unescapeHTML,
  19     urlencode_postdata,
  20 )
  21
  22
  23 class UdemyIE(InfoExtractor):
  24     IE_NAME = 'udemy'
  25     _VALID_URL = r'''(?x)
  26                     https?://
  27                         www\.udemy\.com/
  28                         (?:
  29                             [^#]+\#/lecture/|
  30                             lecture/view/?\?lectureId=|
  31                             [^/]+/learn/v4/t/lecture/
  32                         )
  33                         (?P<id>\d+)
  34                     '''
  35     _LOGIN_URL = 'https://www.udemy.com/join/login-popup/?displayType=ajax&showSkipButton=1'
  36     _ORIGIN_URL = 'https://www.udemy.com'
  37     _NETRC_MACHINE = 'udemy'
  38
  39     _TESTS = [{
  40         'url': 'https://www.udemy.com/java-tutorial/#/lecture/172757',
  41         'md5': '98eda5b657e752cf945d8445e261b5c5',
  42         'info_dict': {
  43             'id': '160614',
  44             'ext': 'mp4',
  45             'title': 'Introduction and Installation',
  46             'description': 'md5:c0d51f6f21ef4ec65f091055a5eef876',
  47             'duration': 579.29,
  48         },
  49         'skip': 'Requires udemy account credentials',
  50     }, {
  51         # new URL schema
  52         'url': 'https://www.udemy.com/electric-bass-right-from-the-start/learn/v4/t/lecture/4580906',
  53         'only_matching': True,
  54     }]
  55
  56     def _extract_course_info(self, webpage, video_id):
  57         course = self._parse_json(
  58             unescapeHTML(self._search_regex(
  59                 r'ng-init=["\'].*\bcourse=({.+?});', webpage, 'course', default='{}')),
  60             video_id, fatal=False) or {}
  61         course_id = course.get('id') or self._search_regex(
  62             (r'&quot;id&quot;\s*:\s*(\d+)', r'data-course-id=["\'](\d+)'),
  63             webpage, 'course id')
  64         return course_id, course.get('title')
  65
  66     def _enroll_course(self, base_url, webpage, course_id):
  67         def combine_url(base_url, url):
  68             return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
  69
  70         checkout_url = unescapeHTML(self._search_regex(
  71             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/payment/checkout/.+?)\1',
  72             webpage, 'checkout url', group='url', default=None))
  73         if checkout_url:
  74             raise ExtractorError(
  75                 'Course %s is not free. You have to pay for it before you can download. '
  76                 'Use this URL to confirm purchase: %s'
  77                 % (course_id, combine_url(base_url, checkout_url)),
  78                 expected=True)
  79
  80         enroll_url = unescapeHTML(self._search_regex(
  81             r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',
  82             webpage, 'enroll url', group='url', default=None))
  83         if enroll_url:
  84             webpage = self._download_webpage(
  85                 combine_url(base_url, enroll_url),
  86                 course_id, 'Enrolling in the course',
  87                 headers={'Referer': base_url})
  88             if '>You have enrolled in' in webpage:
  89                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
  90
  91     def _download_lecture(self, course_id, lecture_id):
  92         return self._download_json(
  93             'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
  94             % (course_id, lecture_id),
  95             lecture_id, 'Downloading lecture JSON', query={
  96                 'fields[lecture]': 'title,description,view_html,asset',
  97                 'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,data',
  98             })
  99
 100     def _handle_error(self, response):
 101         if not isinstance(response, dict):
 102             return
 103         error = response.get('error')
 104         if error:
 105             error_str = 'Udemy returned error #%s: %s' % (error.get('code'), error.get('message'))
 106             error_data = error.get('data')
 107             if error_data:
 108                 error_str += ' - %s' % error_data.get('formErrors')
 109             raise ExtractorError(error_str, expected=True)
 110
 111     def _download_json(self, url_or_request, *args, **kwargs):
 112         headers = {
 113             'X-Udemy-Snail-Case': 'true',
 114             'X-Requested-With': 'XMLHttpRequest',
 115         }
 116         for cookie in self._downloader.cookiejar:
 117             if cookie.name == 'client_id':
 118                 headers['X-Udemy-Client-Id'] = cookie.value
 119             elif cookie.name == 'access_token':
 120                 headers['X-Udemy-Bearer-Token'] = cookie.value
 121                 headers['X-Udemy-Authorization'] = 'Bearer %s' % cookie.value
 122
 123         if isinstance(url_or_request, compat_urllib_request.Request):
 124             for header, value in headers.items():
 125                 url_or_request.add_header(header, value)
 126         else:
 127             url_or_request = sanitized_Request(url_or_request, headers=headers)
 128
 129         response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)
 130         self._handle_error(response)
 131         return response
 132
 133     def _real_initialize(self):
 134         self._login()
 135
 136     def _login(self):
 137         (username, password) = self._get_login_info()
 138         if username is None:
 139             return
 140
 141         login_popup = self._download_webpage(
 142             self._LOGIN_URL, None, 'Downloading login popup')
 143
 144         def is_logged(webpage):
 145             return any(re.search(p, webpage) for p in (
 146                 r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
 147                 r'>Logout<'))
 148
 149         # already logged in
 150         if is_logged(login_popup):
 151             return
 152
 153         login_form = self._form_hidden_inputs('login-form', login_popup)
 154
 155         login_form.update({
 156             'email': username,
 157             'password': password,
 158         })
 159
 160         response = self._download_webpage(
 161             self._LOGIN_URL, None, 'Logging in as %s' % username,
 162             data=urlencode_postdata(login_form),
 163             headers={
 164                 'Referer': self._ORIGIN_URL,
 165                 'Origin': self._ORIGIN_URL,
 166             })
 167
 168         if not is_logged(response):
 169             error = self._html_search_regex(
 170                 r'(?s)<div[^>]+class="form-errors[^"]*">(.+?)</div>',
 171                 response, 'error message', default=None)
 172             if error:
 173                 raise ExtractorError('Unable to login: %s' % error, expected=True)
 174             raise ExtractorError('Unable to log in')
 175
 176     def _real_extract(self, url):
 177         lecture_id = self._match_id(url)
 178
 179         webpage = self._download_webpage(url, lecture_id)
 180
 181         course_id, _ = self._extract_course_info(webpage, lecture_id)
 182
 183         try:
 184             lecture = self._download_lecture(course_id, lecture_id)
 185         except ExtractorError as e:
 186             # Error could possibly mean we are not enrolled in the course
 187             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 188                 self._enroll_course(url, webpage, course_id)
 189                 lecture = self._download_lecture(course_id, lecture_id)
 190             else:
 191                 raise
 192
 193         title = lecture['title']
 194         description = lecture.get('description')
 195
 196         asset = lecture['asset']
 197
 198         asset_type = asset.get('asset_type') or asset.get('assetType')
 199         if asset_type != 'Video':
 200             raise ExtractorError(
 201                 'Lecture %s is not a video' % lecture_id, expected=True)
 202
 203         stream_url = asset.get('stream_url') or asset.get('streamUrl')
 204         if stream_url:
 205             youtube_url = self._search_regex(
 206                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
 207             if youtube_url:
 208                 return self.url_result(youtube_url, 'Youtube')
 209
 210         video_id = asset['id']
 211         thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
 212         duration = float_or_none(asset.get('data', {}).get('duration'))
 213
 214         formats = []
 215
 216         def extract_output_format(src):
 217             return {
 218                 'url': src['url'],
 219                 'format_id': '%sp' % (src.get('height') or format_id),
 220                 'width': int_or_none(src.get('width')),
 221                 'height': int_or_none(src.get('height')),
 222                 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
 223                 'vcodec': src.get('video_codec'),
 224                 'fps': int_or_none(src.get('frame_rate')),
 225                 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
 226                 'acodec': src.get('audio_codec'),
 227                 'asr': int_or_none(src.get('audio_sample_rate')),
 228                 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
 229                 'filesize': int_or_none(src.get('file_size_in_bytes')),
 230             }
 231
 232         outputs = asset.get('data', {}).get('outputs')
 233         if not isinstance(outputs, dict):
 234             outputs = {}
 235
 236         def add_output_format_meta(f, key):
 237             output = outputs.get(key)
 238             if isinstance(output, dict):
 239                 output_format = extract_output_format(output)
 240                 output_format.update(f)
 241                 return output_format
 242             return f
 243
 244         download_urls = asset.get('download_urls')
 245         if isinstance(download_urls, dict):
 246             video = download_urls.get('Video')
 247             if isinstance(video, list):
 248                 for format_ in video:
 249                     video_url = format_.get('file')
 250                     if not video_url:
 251                         continue
 252                     format_id = format_.get('label')
 253                     f = {
 254                         'url': format_['file'],
 255                         'format_id': '%sp' % format_id,
 256                         'height': int_or_none(format_id),
 257                     }
 258                     if format_id:
 259                         # Some videos contain additional metadata (e.g.
 260                         # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
 261                         f = add_output_format_meta(f, format_id)
 262                     formats.append(f)
 263
 264         view_html = lecture.get('view_html')
 265         if view_html:
 266             view_html_urls = set()
 267             for source in re.findall(r'<source[^>]+>', view_html):
 268                 attributes = extract_attributes(source)
 269                 src = attributes.get('src')
 270                 if not src:
 271                     continue
 272                 res = attributes.get('data-res')
 273                 height = int_or_none(res)
 274                 if src in view_html_urls:
 275                     continue
 276                 view_html_urls.add(src)
 277                 if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
 278                     m3u8_formats = self._extract_m3u8_formats(
 279                         src, video_id, 'mp4', entry_protocol='m3u8_native',
 280                         m3u8_id='hls', fatal=False)
 281                     for f in m3u8_formats:
 282                         m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
 283                         if m:
 284                             if not f.get('height'):
 285                                 f['height'] = int(m.group('height'))
 286                             if not f.get('tbr'):
 287                                 f['tbr'] = int(m.group('tbr'))
 288                     formats.extend(m3u8_formats)
 289                 else:
 290                     formats.append(add_output_format_meta({
 291                         'url': src,
 292                         'format_id': '%dp' % height if height else None,
 293                         'height': height,
 294                     }, res))
 295
 296         self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 297
 298         return {
 299             'id': video_id,
 300             'title': title,
 301             'description': description,
 302             'thumbnail': thumbnail,
 303             'duration': duration,
 304             'formats': formats
 305         }
 306
 307
 308 class UdemyCourseIE(UdemyIE):
 309     IE_NAME = 'udemy:course'
 310     _VALID_URL = r'https?://www\.udemy\.com/(?P<id>[^/?#&]+)'
 311     _TESTS = []
 312
 313     @classmethod
 314     def suitable(cls, url):
 315         return False if UdemyIE.suitable(url) else super(UdemyCourseIE, cls).suitable(url)
 316
 317     def _real_extract(self, url):
 318         course_path = self._match_id(url)
 319
 320         webpage = self._download_webpage(url, course_path)
 321
 322         course_id, title = self._extract_course_info(webpage, course_path)
 323
 324         self._enroll_course(url, webpage, course_id)
 325
 326         response = self._download_json(
 327             'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
 328             course_id, 'Downloading course curriculum', query={
 329                 'fields[chapter]': 'title,object_index',
 330                 'fields[lecture]': 'title,asset',
 331                 'page_size': '1000',
 332             })
 333
 334         entries = []
 335         chapter, chapter_number = [None] * 2
 336         for entry in response['results']:
 337             clazz = entry.get('_class')
 338             if clazz == 'lecture':
 339                 asset = entry.get('asset')
 340                 if isinstance(asset, dict):
 341                     asset_type = asset.get('asset_type') or asset.get('assetType')
 342                     if asset_type != 'Video':
 343                         continue
 344                 lecture_id = entry.get('id')
 345                 if lecture_id:
 346                     entry = {
 347                         '_type': 'url_transparent',
 348                         'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
 349                         'title': entry.get('title'),
 350                         'ie_key': UdemyIE.ie_key(),
 351                     }
 352                     if chapter_number:
 353                         entry['chapter_number'] = chapter_number
 354                     if chapter:
 355                         entry['chapter'] = chapter
 356                     entries.append(entry)
 357             elif clazz == 'chapter':
 358                 chapter_number = entry.get('object_index')
 359                 chapter = entry.get('title')
 360
 361         return self.playlist_result(entries, course_id, title)