Merge pull request #14225 from Tithen-Firion/openload-phantomjs-method
[youtube-dl] / youtube_dl / extractor / pluralsight.py
1 from __future__ import unicode_literals
2
3 import collections
4 import json
5 import os
6 import random
7
8 from .common import InfoExtractor
9 from ..compat import (
10     compat_str,
11     compat_urlparse,
12 )
13 from ..utils import (
14     dict_get,
15     ExtractorError,
16     float_or_none,
17     int_or_none,
18     parse_duration,
19     qualities,
20     srt_subtitles_timecode,
21     try_get,
22     update_url_query,
23     urlencode_postdata,
24 )
25
26
27 class PluralsightBaseIE(InfoExtractor):
28     _API_BASE = 'https://app.pluralsight.com'
29
30     def _download_course(self, course_id, url, display_id):
31         try:
32             return self._download_course_rpc(course_id, url, display_id)
33         except ExtractorError:
34             # Old API fallback
35             return self._download_json(
36                 'https://app.pluralsight.com/player/user/api/v1/player/payload',
37                 display_id, data=urlencode_postdata({'courseId': course_id}),
38                 headers={'Referer': url})
39
40     def _download_course_rpc(self, course_id, url, display_id):
41         response = self._download_json(
42             '%s/player/functions/rpc' % self._API_BASE, display_id,
43             'Downloading course JSON',
44             data=json.dumps({
45                 'fn': 'bootstrapPlayer',
46                 'payload': {
47                     'courseId': course_id,
48                 },
49             }).encode('utf-8'),
50             headers={
51                 'Content-Type': 'application/json;charset=utf-8',
52                 'Referer': url,
53             })
54
55         course = try_get(response, lambda x: x['payload']['course'], dict)
56         if course:
57             return course
58
59         raise ExtractorError(
60             '%s said: %s' % (self.IE_NAME, response['error']['message']),
61             expected=True)
62
63
64 class PluralsightIE(PluralsightBaseIE):
65     IE_NAME = 'pluralsight'
66     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:training/)?player\?'
67     _LOGIN_URL = 'https://app.pluralsight.com/id/'
68
69     _NETRC_MACHINE = 'pluralsight'
70
71     _TESTS = [{
72         'url': 'http://www.pluralsight.com/training/player?author=mike-mckeown&name=hosting-sql-server-windows-azure-iaas-m7-mgmt&mode=live&clip=3&course=hosting-sql-server-windows-azure-iaas',
73         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
74         'info_dict': {
75             'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04',
76             'ext': 'mp4',
77             'title': 'Demo Monitoring',
78             'duration': 338,
79         },
80         'skip': 'Requires pluralsight account credentials',
81     }, {
82         'url': 'https://app.pluralsight.com/training/player?course=angularjs-get-started&author=scott-allen&name=angularjs-get-started-m1-introduction&clip=0&mode=live',
83         'only_matching': True,
84     }, {
85         # available without pluralsight account
86         'url': 'http://app.pluralsight.com/training/player?author=scott-allen&name=angularjs-get-started-m1-introduction&mode=live&clip=0&course=angularjs-get-started',
87         'only_matching': True,
88     }, {
89         'url': 'https://app.pluralsight.com/player?course=ccna-intro-networking&author=ross-bagurdes&name=ccna-intro-networking-m06&clip=0',
90         'only_matching': True,
91     }]
92
93     def _real_initialize(self):
94         self._login()
95
96     def _login(self):
97         (username, password) = self._get_login_info()
98         if username is None:
99             return
100
101         login_page = self._download_webpage(
102             self._LOGIN_URL, None, 'Downloading login page')
103
104         login_form = self._hidden_inputs(login_page)
105
106         login_form.update({
107             'Username': username,
108             'Password': password,
109         })
110
111         post_url = self._search_regex(
112             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
113             'post url', default=self._LOGIN_URL, group='url')
114
115         if not post_url.startswith('http'):
116             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
117
118         response = self._download_webpage(
119             post_url, None, 'Logging in as %s' % username,
120             data=urlencode_postdata(login_form),
121             headers={'Content-Type': 'application/x-www-form-urlencoded'})
122
123         error = self._search_regex(
124             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',
125             response, 'error message', default=None)
126         if error:
127             raise ExtractorError('Unable to login: %s' % error, expected=True)
128
129         if all(p not in response for p in ('__INITIAL_STATE__', '"currentUser"')):
130             BLOCKED = 'Your account has been blocked due to suspicious activity'
131             if BLOCKED in response:
132                 raise ExtractorError(
133                     'Unable to login: %s' % BLOCKED, expected=True)
134             raise ExtractorError('Unable to log in')
135
136     def _get_subtitles(self, author, clip_id, lang, name, duration, video_id):
137         captions_post = {
138             'a': author,
139             'cn': clip_id,
140             'lc': lang,
141             'm': name,
142         }
143         captions = self._download_json(
144             '%s/player/retrieve-captions' % self._API_BASE, video_id,
145             'Downloading captions JSON', 'Unable to download captions JSON',
146             fatal=False, data=json.dumps(captions_post).encode('utf-8'),
147             headers={'Content-Type': 'application/json;charset=utf-8'})
148         if captions:
149             return {
150                 lang: [{
151                     'ext': 'json',
152                     'data': json.dumps(captions),
153                 }, {
154                     'ext': 'srt',
155                     'data': self._convert_subtitles(duration, captions),
156                 }]
157             }
158
159     @staticmethod
160     def _convert_subtitles(duration, subs):
161         srt = ''
162         TIME_OFFSET_KEYS = ('displayTimeOffset', 'DisplayTimeOffset')
163         TEXT_KEYS = ('text', 'Text')
164         for num, current in enumerate(subs):
165             current = subs[num]
166             start, text = (
167                 float_or_none(dict_get(current, TIME_OFFSET_KEYS)),
168                 dict_get(current, TEXT_KEYS))
169             if start is None or text is None:
170                 continue
171             end = duration if num == len(subs) - 1 else float_or_none(
172                 dict_get(subs[num + 1], TIME_OFFSET_KEYS))
173             if end is None:
174                 continue
175             srt += os.linesep.join(
176                 (
177                     '%d' % num,
178                     '%s --> %s' % (
179                         srt_subtitles_timecode(start),
180                         srt_subtitles_timecode(end)),
181                     text,
182                     os.linesep,
183                 ))
184         return srt
185
186     def _real_extract(self, url):
187         qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
188
189         author = qs.get('author', [None])[0]
190         name = qs.get('name', [None])[0]
191         clip_id = qs.get('clip', [None])[0]
192         course_name = qs.get('course', [None])[0]
193
194         if any(not f for f in (author, name, clip_id, course_name,)):
195             raise ExtractorError('Invalid URL', expected=True)
196
197         display_id = '%s-%s' % (name, clip_id)
198
199         course = self._download_course(course_name, url, display_id)
200
201         collection = course['modules']
202
203         clip = None
204
205         for module_ in collection:
206             if name in (module_.get('moduleName'), module_.get('name')):
207                 for clip_ in module_.get('clips', []):
208                     clip_index = clip_.get('clipIndex')
209                     if clip_index is None:
210                         clip_index = clip_.get('index')
211                     if clip_index is None:
212                         continue
213                     if compat_str(clip_index) == clip_id:
214                         clip = clip_
215                         break
216
217         if not clip:
218             raise ExtractorError('Unable to resolve clip')
219
220         title = clip['title']
221
222         QUALITIES = {
223             'low': {'width': 640, 'height': 480},
224             'medium': {'width': 848, 'height': 640},
225             'high': {'width': 1024, 'height': 768},
226             'high-widescreen': {'width': 1280, 'height': 720},
227         }
228
229         QUALITIES_PREFERENCE = ('low', 'medium', 'high', 'high-widescreen',)
230         quality_key = qualities(QUALITIES_PREFERENCE)
231
232         AllowedQuality = collections.namedtuple('AllowedQuality', ['ext', 'qualities'])
233
234         ALLOWED_QUALITIES = (
235             AllowedQuality('webm', ['high', ]),
236             AllowedQuality('mp4', ['low', 'medium', 'high', ]),
237         )
238
239         # Some courses also offer widescreen resolution for high quality (see
240         # https://github.com/rg3/youtube-dl/issues/7766)
241         widescreen = course.get('supportsWideScreenVideoFormats') is True
242         best_quality = 'high-widescreen' if widescreen else 'high'
243         if widescreen:
244             for allowed_quality in ALLOWED_QUALITIES:
245                 allowed_quality.qualities.append(best_quality)
246
247         # In order to minimize the number of calls to ViewClip API and reduce
248         # the probability of being throttled or banned by Pluralsight we will request
249         # only single format until formats listing was explicitly requested.
250         if self._downloader.params.get('listformats', False):
251             allowed_qualities = ALLOWED_QUALITIES
252         else:
253             def guess_allowed_qualities():
254                 req_format = self._downloader.params.get('format') or 'best'
255                 req_format_split = req_format.split('-', 1)
256                 if len(req_format_split) > 1:
257                     req_ext, req_quality = req_format_split
258                     req_quality = '-'.join(req_quality.split('-')[:2])
259                     for allowed_quality in ALLOWED_QUALITIES:
260                         if req_ext == allowed_quality.ext and req_quality in allowed_quality.qualities:
261                             return (AllowedQuality(req_ext, (req_quality, )), )
262                 req_ext = 'webm' if self._downloader.params.get('prefer_free_formats') else 'mp4'
263                 return (AllowedQuality(req_ext, (best_quality, )), )
264             allowed_qualities = guess_allowed_qualities()
265
266         formats = []
267         for ext, qualities_ in allowed_qualities:
268             for quality in qualities_:
269                 f = QUALITIES[quality].copy()
270                 clip_post = {
271                     'author': author,
272                     'includeCaptions': False,
273                     'clipIndex': int(clip_id),
274                     'courseName': course_name,
275                     'locale': 'en',
276                     'moduleName': name,
277                     'mediaType': ext,
278                     'quality': '%dx%d' % (f['width'], f['height']),
279                 }
280                 format_id = '%s-%s' % (ext, quality)
281                 viewclip = self._download_json(
282                     '%s/video/clips/viewclip' % self._API_BASE, display_id,
283                     'Downloading %s viewclip JSON' % format_id, fatal=False,
284                     data=json.dumps(clip_post).encode('utf-8'),
285                     headers={'Content-Type': 'application/json;charset=utf-8'})
286
287                 # Pluralsight tracks multiple sequential calls to ViewClip API and start
288                 # to return 429 HTTP errors after some time (see
289                 # https://github.com/rg3/youtube-dl/pull/6989). Moreover it may even lead
290                 # to account ban (see https://github.com/rg3/youtube-dl/issues/6842).
291                 # To somewhat reduce the probability of these consequences
292                 # we will sleep random amount of time before each call to ViewClip.
293                 self._sleep(
294                     random.randint(2, 5), display_id,
295                     '%(video_id)s: Waiting for %(timeout)s seconds to avoid throttling')
296
297                 if not viewclip:
298                     continue
299
300                 clip_urls = viewclip.get('urls')
301                 if not isinstance(clip_urls, list):
302                     continue
303
304                 for clip_url_data in clip_urls:
305                     clip_url = clip_url_data.get('url')
306                     if not clip_url:
307                         continue
308                     cdn = clip_url_data.get('cdn')
309                     clip_f = f.copy()
310                     clip_f.update({
311                         'url': clip_url,
312                         'ext': ext,
313                         'format_id': '%s-%s' % (format_id, cdn) if cdn else format_id,
314                         'quality': quality_key(quality),
315                         'source_preference': int_or_none(clip_url_data.get('rank')),
316                     })
317                     formats.append(clip_f)
318
319         self._sort_formats(formats)
320
321         duration = int_or_none(
322             clip.get('duration')) or parse_duration(clip.get('formattedDuration'))
323
324         # TODO: other languages?
325         subtitles = self.extract_subtitles(
326             author, clip_id, 'en', name, duration, display_id)
327
328         return {
329             'id': clip.get('clipName') or clip['name'],
330             'title': title,
331             'duration': duration,
332             'creator': author,
333             'formats': formats,
334             'subtitles': subtitles,
335         }
336
337
338 class PluralsightCourseIE(PluralsightBaseIE):
339     IE_NAME = 'pluralsight:course'
340     _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:library/)?courses/(?P<id>[^/]+)'
341     _TESTS = [{
342         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
343         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
344         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',
345         'info_dict': {
346             'id': 'hosting-sql-server-windows-azure-iaas',
347             'title': 'Hosting SQL Server in Microsoft Azure IaaS Fundamentals',
348             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
349         },
350         'playlist_count': 31,
351     }, {
352         # available without pluralsight account
353         'url': 'https://www.pluralsight.com/courses/angularjs-get-started',
354         'only_matching': True,
355     }, {
356         'url': 'https://app.pluralsight.com/library/courses/understanding-microsoft-azure-amazon-aws/table-of-contents',
357         'only_matching': True,
358     }]
359
360     def _real_extract(self, url):
361         course_id = self._match_id(url)
362
363         # TODO: PSM cookie
364
365         course = self._download_course(course_id, url, course_id)
366
367         title = course['title']
368         course_name = course['name']
369         course_data = course['modules']
370         description = course.get('description') or course.get('shortDescription')
371
372         entries = []
373         for num, module in enumerate(course_data, 1):
374             author = module.get('author')
375             module_name = module.get('name')
376             if not author or not module_name:
377                 continue
378             for clip in module.get('clips', []):
379                 clip_index = int_or_none(clip.get('index'))
380                 if clip_index is None:
381                     continue
382                 clip_url = update_url_query(
383                     '%s/player' % self._API_BASE, query={
384                         'mode': 'live',
385                         'course': course_name,
386                         'author': author,
387                         'name': module_name,
388                         'clip': clip_index,
389                     })
390                 entries.append({
391                     '_type': 'url_transparent',
392                     'url': clip_url,
393                     'ie_key': PluralsightIE.ie_key(),
394                     'chapter': module.get('title'),
395                     'chapter_number': num,
396                     'chapter_id': module.get('moduleRef'),
397                 })
398
399         return self.playlist_result(entries, course_id, title, description)