]> gitweb @ CieloNegro.org - youtube-dl.git/blob - youtube_dl/extractor/lynda.py
[cbs] extract smpte and vtt subtitles
[youtube-dl.git] / youtube_dl / extractor / lynda.py
1 from __future__ import unicode_literals
2
3 import re
4
5 from .common import InfoExtractor
6 from ..compat import (
7     compat_str,
8     compat_urlparse,
9 )
10 from ..utils import (
11     ExtractorError,
12     int_or_none,
13     urlencode_postdata,
14 )
15
16
17 class LyndaBaseIE(InfoExtractor):
18     _SIGNIN_URL = 'https://www.lynda.com/signin/lynda'
19     _PASSWORD_URL = 'https://www.lynda.com/signin/password'
20     _USER_URL = 'https://www.lynda.com/signin/user'
21     _ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
22     _NETRC_MACHINE = 'lynda'
23
24     def _real_initialize(self):
25         self._login()
26
27     @staticmethod
28     def _check_error(json_string, key_or_keys):
29         keys = [key_or_keys] if isinstance(key_or_keys, compat_str) else key_or_keys
30         for key in keys:
31             error = json_string.get(key)
32             if error:
33                 raise ExtractorError('Unable to login: %s' % error, expected=True)
34
35     def _login_step(self, form_html, fallback_action_url, extra_form_data, note, referrer_url):
36         action_url = self._search_regex(
37             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', form_html,
38             'post url', default=fallback_action_url, group='url')
39
40         if not action_url.startswith('http'):
41             action_url = compat_urlparse.urljoin(self._SIGNIN_URL, action_url)
42
43         form_data = self._hidden_inputs(form_html)
44         form_data.update(extra_form_data)
45
46         response = self._download_json(
47             action_url, None, note,
48             data=urlencode_postdata(form_data),
49             headers={
50                 'Referer': referrer_url,
51                 'X-Requested-With': 'XMLHttpRequest',
52             }, expected_status=(418, 500, ))
53
54         self._check_error(response, ('email', 'password', 'ErrorMessage'))
55
56         return response, action_url
57
58     def _login(self):
59         username, password = self._get_login_info()
60         if username is None:
61             return
62
63         # Step 1: download signin page
64         signin_page = self._download_webpage(
65             self._SIGNIN_URL, None, 'Downloading signin page')
66
67         # Already logged in
68         if any(re.search(p, signin_page) for p in (
69                 r'isLoggedIn\s*:\s*true', r'logout\.aspx', r'>Log out<')):
70             return
71
72         # Step 2: submit email
73         signin_form = self._search_regex(
74             r'(?s)(<form[^>]+data-form-name=["\']signin["\'][^>]*>.+?</form>)',
75             signin_page, 'signin form')
76         signin_page, signin_url = self._login_step(
77             signin_form, self._PASSWORD_URL, {'email': username},
78             'Submitting email', self._SIGNIN_URL)
79
80         # Step 3: submit password
81         password_form = signin_page['body']
82         self._login_step(
83             password_form, self._USER_URL, {'email': username, 'password': password},
84             'Submitting password', signin_url)
85
86
87 class LyndaIE(LyndaBaseIE):
88     IE_NAME = 'lynda'
89     IE_DESC = 'lynda.com videos'
90     _VALID_URL = r'''(?x)
91                     https?://
92                         (?:www\.)?(?:lynda\.com|educourse\.ga)/
93                         (?:
94                             (?:[^/]+/){2,3}(?P<course_id>\d+)|
95                             player/embed
96                         )/
97                         (?P<id>\d+)
98                     '''
99
100     _TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
101
102     _TESTS = [{
103         'url': 'https://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
104         # md5 is unstable
105         'info_dict': {
106             'id': '114408',
107             'ext': 'mp4',
108             'title': 'Using the exercise files',
109             'duration': 68
110         }
111     }, {
112         'url': 'https://www.lynda.com/player/embed/133770?tr=foo=1;bar=g;fizz=rt&fs=0',
113         'only_matching': True,
114     }, {
115         'url': 'https://educourse.ga/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
116         'only_matching': True,
117     }, {
118         'url': 'https://www.lynda.com/de/Graphic-Design-tutorials/Willkommen-Grundlagen-guten-Gestaltung/393570/393572-4.html',
119         'only_matching': True,
120     }]
121
122     def _raise_unavailable(self, video_id):
123         self.raise_login_required(
124             'Video %s is only available for members' % video_id)
125
126     def _real_extract(self, url):
127         mobj = re.match(self._VALID_URL, url)
128         video_id = mobj.group('id')
129         course_id = mobj.group('course_id')
130
131         query = {
132             'videoId': video_id,
133             'type': 'video',
134         }
135
136         video = self._download_json(
137             'https://www.lynda.com/ajax/player', video_id,
138             'Downloading video JSON', fatal=False, query=query)
139
140         # Fallback scenario
141         if not video:
142             query['courseId'] = course_id
143
144             play = self._download_json(
145                 'https://www.lynda.com/ajax/course/%s/%s/play'
146                 % (course_id, video_id), video_id, 'Downloading play JSON')
147
148             if not play:
149                 self._raise_unavailable(video_id)
150
151             formats = []
152             for formats_dict in play:
153                 urls = formats_dict.get('urls')
154                 if not isinstance(urls, dict):
155                     continue
156                 cdn = formats_dict.get('name')
157                 for format_id, format_url in urls.items():
158                     if not format_url:
159                         continue
160                     formats.append({
161                         'url': format_url,
162                         'format_id': '%s-%s' % (cdn, format_id) if cdn else format_id,
163                         'height': int_or_none(format_id),
164                     })
165             self._sort_formats(formats)
166
167             conviva = self._download_json(
168                 'https://www.lynda.com/ajax/player/conviva', video_id,
169                 'Downloading conviva JSON', query=query)
170
171             return {
172                 'id': video_id,
173                 'title': conviva['VideoTitle'],
174                 'description': conviva.get('VideoDescription'),
175                 'release_year': int_or_none(conviva.get('ReleaseYear')),
176                 'duration': int_or_none(conviva.get('Duration')),
177                 'creator': conviva.get('Author'),
178                 'formats': formats,
179             }
180
181         if 'Status' in video:
182             raise ExtractorError(
183                 'lynda returned error: %s' % video['Message'], expected=True)
184
185         if video.get('HasAccess') is False:
186             self._raise_unavailable(video_id)
187
188         video_id = compat_str(video.get('ID') or video_id)
189         duration = int_or_none(video.get('DurationInSeconds'))
190         title = video['Title']
191
192         formats = []
193
194         fmts = video.get('Formats')
195         if fmts:
196             formats.extend([{
197                 'url': f['Url'],
198                 'ext': f.get('Extension'),
199                 'width': int_or_none(f.get('Width')),
200                 'height': int_or_none(f.get('Height')),
201                 'filesize': int_or_none(f.get('FileSize')),
202                 'format_id': compat_str(f.get('Resolution')) if f.get('Resolution') else None,
203             } for f in fmts if f.get('Url')])
204
205         prioritized_streams = video.get('PrioritizedStreams')
206         if prioritized_streams:
207             for prioritized_stream_id, prioritized_stream in prioritized_streams.items():
208                 formats.extend([{
209                     'url': video_url,
210                     'height': int_or_none(format_id),
211                     'format_id': '%s-%s' % (prioritized_stream_id, format_id),
212                 } for format_id, video_url in prioritized_stream.items()])
213
214         self._check_formats(formats, video_id)
215         self._sort_formats(formats)
216
217         subtitles = self.extract_subtitles(video_id)
218
219         return {
220             'id': video_id,
221             'title': title,
222             'duration': duration,
223             'subtitles': subtitles,
224             'formats': formats
225         }
226
227     def _fix_subtitles(self, subs):
228         srt = ''
229         seq_counter = 0
230         for pos in range(0, len(subs) - 1):
231             seq_current = subs[pos]
232             m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
233             if m_current is None:
234                 continue
235             seq_next = subs[pos + 1]
236             m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
237             if m_next is None:
238                 continue
239             appear_time = m_current.group('timecode')
240             disappear_time = m_next.group('timecode')
241             text = seq_current['Caption'].strip()
242             if text:
243                 seq_counter += 1
244                 srt += '%s\r\n%s --> %s\r\n%s\r\n\r\n' % (seq_counter, appear_time, disappear_time, text)
245         if srt:
246             return srt
247
248     def _get_subtitles(self, video_id):
249         url = 'https://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
250         subs = self._download_json(url, None, False)
251         fixed_subs = self._fix_subtitles(subs)
252         if fixed_subs:
253             return {'en': [{'ext': 'srt', 'data': fixed_subs}]}
254         else:
255             return {}
256
257
258 class LyndaCourseIE(LyndaBaseIE):
259     IE_NAME = 'lynda:course'
260     IE_DESC = 'lynda.com online courses'
261
262     # Course link equals to welcome/introduction video link of same course
263     # We will recognize it as course link
264     _VALID_URL = r'https?://(?:www|m)\.(?:lynda\.com|educourse\.ga)/(?P<coursepath>(?:[^/]+/){2,3}(?P<courseid>\d+))-2\.html'
265
266     _TESTS = [{
267         'url': 'https://www.lynda.com/Graphic-Design-tutorials/Grundlagen-guten-Gestaltung/393570-2.html',
268         'only_matching': True,
269     }, {
270         'url': 'https://www.lynda.com/de/Graphic-Design-tutorials/Grundlagen-guten-Gestaltung/393570-2.html',
271         'only_matching': True,
272     }]
273
274     def _real_extract(self, url):
275         mobj = re.match(self._VALID_URL, url)
276         course_path = mobj.group('coursepath')
277         course_id = mobj.group('courseid')
278
279         item_template = 'https://www.lynda.com/%s/%%s-4.html' % course_path
280
281         course = self._download_json(
282             'https://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
283             course_id, 'Downloading course JSON', fatal=False)
284
285         if not course:
286             webpage = self._download_webpage(url, course_id)
287             entries = [
288                 self.url_result(
289                     item_template % video_id, ie=LyndaIE.ie_key(),
290                     video_id=video_id)
291                 for video_id in re.findall(
292                     r'data-video-id=["\'](\d+)', webpage)]
293             return self.playlist_result(
294                 entries, course_id,
295                 self._og_search_title(webpage, fatal=False),
296                 self._og_search_description(webpage))
297
298         if course.get('Status') == 'NotFound':
299             raise ExtractorError(
300                 'Course %s does not exist' % course_id, expected=True)
301
302         unaccessible_videos = 0
303         entries = []
304
305         # Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
306         # by single video API anymore
307
308         for chapter in course['Chapters']:
309             for video in chapter.get('Videos', []):
310                 if video.get('HasAccess') is False:
311                     unaccessible_videos += 1
312                     continue
313                 video_id = video.get('ID')
314                 if video_id:
315                     entries.append({
316                         '_type': 'url_transparent',
317                         'url': item_template % video_id,
318                         'ie_key': LyndaIE.ie_key(),
319                         'chapter': chapter.get('Title'),
320                         'chapter_number': int_or_none(chapter.get('ChapterIndex')),
321                         'chapter_id': compat_str(chapter.get('ID')),
322                     })
323
324         if unaccessible_videos > 0:
325             self._downloader.report_warning(
326                 '%s videos are only available for members (or paid members) and will not be downloaded. '
327                 % unaccessible_videos + self._ACCOUNT_CREDENTIALS_HINT)
328
329         course_title = course.get('Title')
330         course_description = course.get('Description')
331
332         return self.playlist_result(entries, course_id, course_title, course_description)