X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/9dc487f48b50767cf540fa36c3de2c386fd74c04..96981814751d51f37fe7347060de3a51c982f5ec:/youtube_dl/extractor/pluralsight.py?ds=sidebyside diff --git a/youtube_dl/extractor/pluralsight.py b/youtube_dl/extractor/pluralsight.py index 9aab776..abd08bc 100644 --- a/youtube_dl/extractor/pluralsight.py +++ b/youtube_dl/extractor/pluralsight.py @@ -1,9 +1,10 @@ from __future__ import unicode_literals -import re +import collections import json +import os import random -import collections +import re from .common import InfoExtractor from ..compat import ( @@ -11,22 +12,107 @@ from ..compat import ( compat_urlparse, ) from ..utils import ( + dict_get, ExtractorError, + float_or_none, int_or_none, parse_duration, qualities, - sanitized_Request, + srt_subtitles_timecode, + try_get, + update_url_query, urlencode_postdata, ) class PluralsightBaseIE(InfoExtractor): - _API_BASE = 'http://app.pluralsight.com' + _API_BASE = 'https://app.pluralsight.com' + + _GRAPHQL_EP = '%s/player/api/graphql' % _API_BASE + _GRAPHQL_HEADERS = { + 'Content-Type': 'application/json;charset=UTF-8', + } + _GRAPHQL_COURSE_TMPL = ''' +query BootstrapPlayer { + rpc { + bootstrapPlayer { + profile { + firstName + lastName + email + username + userHandle + authed + isAuthed + plan + } + course(courseId: "%s") { + name + title + courseHasCaptions + translationLanguages { + code + name + } + supportsWideScreenVideoFormats + timestamp + modules { + name + title + duration + formattedDuration + author + authorized + clips { + authorized + clipId + duration + formattedDuration + id + index + moduleIndex + moduleTitle + name + title + watched + } + } + } + } + } +}''' + + def _download_course(self, course_id, url, display_id): + try: + return self._download_course_rpc(course_id, url, display_id) + except ExtractorError: + # Old API fallback + return self._download_json( + 'https://app.pluralsight.com/player/user/api/v1/player/payload', + display_id, data=urlencode_postdata({'courseId': course_id}), + headers={'Referer': url}) + + def _download_course_rpc(self, course_id, url, display_id): + response = self._download_json( + self._GRAPHQL_EP, display_id, data=json.dumps({ + 'query': self._GRAPHQL_COURSE_TMPL % course_id, + 'variables': {} + }).encode('utf-8'), headers=self._GRAPHQL_HEADERS) + + course = try_get( + response, lambda x: x['data']['rpc']['bootstrapPlayer']['course'], + dict) + if course: + return course + + raise ExtractorError( + '%s said: %s' % (self.IE_NAME, response['error']['message']), + expected=True) class PluralsightIE(PluralsightBaseIE): IE_NAME = 'pluralsight' - _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/training/player\?' + _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:training/)?player\?' _LOGIN_URL = 'https://app.pluralsight.com/id/' _NETRC_MACHINE = 'pluralsight' @@ -37,7 +123,7 @@ class PluralsightIE(PluralsightBaseIE): 'info_dict': { 'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04', 'ext': 'mp4', - 'title': 'Management of SQL Server - Demo Monitoring', + 'title': 'Demo Monitoring', 'duration': 338, }, 'skip': 'Requires pluralsight account credentials', @@ -48,13 +134,38 @@ class PluralsightIE(PluralsightBaseIE): # available without pluralsight account 'url': 'http://app.pluralsight.com/training/player?author=scott-allen&name=angularjs-get-started-m1-introduction&mode=live&clip=0&course=angularjs-get-started', 'only_matching': True, + }, { + 'url': 'https://app.pluralsight.com/player?course=ccna-intro-networking&author=ross-bagurdes&name=ccna-intro-networking-m06&clip=0', + 'only_matching': True, }] + GRAPHQL_VIEWCLIP_TMPL = ''' +query viewClip { + viewClip(input: { + author: "%(author)s", + clipIndex: %(clipIndex)d, + courseName: "%(courseName)s", + includeCaptions: %(includeCaptions)s, + locale: "%(locale)s", + mediaType: "%(mediaType)s", + moduleName: "%(moduleName)s", + quality: "%(quality)s" + }) { + urls { + url + cdn + rank + source + }, + status + } +}''' + def _real_initialize(self): self._login() def _login(self): - (username, password) = self._get_login_info() + username, password = self._get_login_info() if username is None: return @@ -75,12 +186,10 @@ class PluralsightIE(PluralsightBaseIE): if not post_url.startswith('http'): post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url) - request = sanitized_Request( - post_url, urlencode_postdata(login_form)) - request.add_header('Content-Type', 'application/x-www-form-urlencoded') - response = self._download_webpage( - request, None, 'Logging in as %s' % username) + post_url, None, 'Logging in', + data=urlencode_postdata(login_form), + headers={'Content-Type': 'application/x-www-form-urlencoded'}) error = self._search_regex( r']+class="field-validation-error"[^>]*>([^<]+)', @@ -88,56 +197,118 @@ class PluralsightIE(PluralsightBaseIE): if error: raise ExtractorError('Unable to login: %s' % error, expected=True) - if all(p not in response for p in ('__INITIAL_STATE__', '"currentUser"')): + if all(not re.search(p, response) for p in ( + r'__INITIAL_STATE__', r'["\']currentUser["\']', + # new layout? + r'>\s*Sign out\s*<')): + BLOCKED = 'Your account has been blocked due to suspicious activity' + if BLOCKED in response: + raise ExtractorError( + 'Unable to login: %s' % BLOCKED, expected=True) + MUST_AGREE = 'To continue using Pluralsight, you must agree to' + if any(p in response for p in (MUST_AGREE, '>Disagree<', '>Agree<')): + raise ExtractorError( + 'Unable to login: %s some documents. Go to pluralsight.com, ' + 'log in and agree with what Pluralsight requires.' + % MUST_AGREE, expected=True) + raise ExtractorError('Unable to log in') + def _get_subtitles(self, author, clip_idx, clip_id, lang, name, duration, video_id): + captions = None + if clip_id: + captions = self._download_json( + '%s/transcript/api/v1/caption/json/%s/%s' + % (self._API_BASE, clip_id, lang), video_id, + 'Downloading captions JSON', 'Unable to download captions JSON', + fatal=False) + if not captions: + captions_post = { + 'a': author, + 'cn': int(clip_idx), + 'lc': lang, + 'm': name, + } + captions = self._download_json( + '%s/player/retrieve-captions' % self._API_BASE, video_id, + 'Downloading captions JSON', 'Unable to download captions JSON', + fatal=False, data=json.dumps(captions_post).encode('utf-8'), + headers={'Content-Type': 'application/json;charset=utf-8'}) + if captions: + return { + lang: [{ + 'ext': 'json', + 'data': json.dumps(captions), + }, { + 'ext': 'srt', + 'data': self._convert_subtitles(duration, captions), + }] + } + + @staticmethod + def _convert_subtitles(duration, subs): + srt = '' + TIME_OFFSET_KEYS = ('displayTimeOffset', 'DisplayTimeOffset') + TEXT_KEYS = ('text', 'Text') + for num, current in enumerate(subs): + current = subs[num] + start, text = ( + float_or_none(dict_get(current, TIME_OFFSET_KEYS, skip_false_values=False)), + dict_get(current, TEXT_KEYS)) + if start is None or text is None: + continue + end = duration if num == len(subs) - 1 else float_or_none( + dict_get(subs[num + 1], TIME_OFFSET_KEYS, skip_false_values=False)) + if end is None: + continue + srt += os.linesep.join( + ( + '%d' % num, + '%s --> %s' % ( + srt_subtitles_timecode(start), + srt_subtitles_timecode(end)), + text, + os.linesep, + )) + return srt + def _real_extract(self, url): qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query) author = qs.get('author', [None])[0] name = qs.get('name', [None])[0] - clip_id = qs.get('clip', [None])[0] - course = qs.get('course', [None])[0] + clip_idx = qs.get('clip', [None])[0] + course_name = qs.get('course', [None])[0] - if any(not f for f in (author, name, clip_id, course,)): + if any(not f for f in (author, name, clip_idx, course_name,)): raise ExtractorError('Invalid URL', expected=True) - display_id = '%s-%s' % (name, clip_id) - - webpage = self._download_webpage(url, display_id) + display_id = '%s-%s' % (name, clip_idx) - modules = self._search_regex( - r'moduleCollection\s*:\s*new\s+ModuleCollection\((\[.+?\])\s*,\s*\$rootScope\)', - webpage, 'modules', default=None) + course = self._download_course(course_name, url, display_id) - if modules: - collection = self._parse_json(modules, display_id) - else: - # Webpage may be served in different layout (see - # https://github.com/rg3/youtube-dl/issues/7607) - collection = self._parse_json( - self._search_regex( - r'var\s+initialState\s*=\s*({.+?});\n', webpage, 'initial state'), - display_id)['course']['modules'] + collection = course['modules'] - module, clip = None, None + clip = None for module_ in collection: if name in (module_.get('moduleName'), module_.get('name')): - module = module_ for clip_ in module_.get('clips', []): clip_index = clip_.get('clipIndex') if clip_index is None: clip_index = clip_.get('index') if clip_index is None: continue - if compat_str(clip_index) == clip_id: + if compat_str(clip_index) == clip_idx: clip = clip_ break if not clip: raise ExtractorError('Unable to resolve clip') + title = clip['title'] + clip_id = clip.get('clipName') or clip.get('name') or clip['clipId'] + QUALITIES = { 'low': {'width': 640, 'height': 480}, 'medium': {'width': 848, 'height': 640}, @@ -156,9 +327,8 @@ class PluralsightIE(PluralsightBaseIE): ) # Some courses also offer widescreen resolution for high quality (see - # https://github.com/rg3/youtube-dl/issues/7766) - widescreen = True if re.search( - r'courseSupportsWidescreenVideoFormats\s*:\s*true', webpage) else False + # https://github.com/ytdl-org/youtube-dl/issues/7766) + widescreen = course.get('supportsWideScreenVideoFormats') is True best_quality = 'high-widescreen' if widescreen else 'high' if widescreen: for allowed_quality in ALLOWED_QUALITIES: @@ -175,6 +345,7 @@ class PluralsightIE(PluralsightBaseIE): req_format_split = req_format.split('-', 1) if len(req_format_split) > 1: req_ext, req_quality = req_format_split + req_quality = '-'.join(req_quality.split('-')[:2]) for allowed_quality in ALLOWED_QUALITIES: if req_ext == allowed_quality.ext and req_quality in allowed_quality.qualities: return (AllowedQuality(req_ext, (req_quality, )), ) @@ -187,56 +358,82 @@ class PluralsightIE(PluralsightBaseIE): for quality in qualities_: f = QUALITIES[quality].copy() clip_post = { - 'a': author, - 'cap': 'false', - 'cn': clip_id, - 'course': course, - 'lc': 'en', - 'm': name, - 'mt': ext, - 'q': '%dx%d' % (f['width'], f['height']), + 'author': author, + 'includeCaptions': 'false', + 'clipIndex': int(clip_idx), + 'courseName': course_name, + 'locale': 'en', + 'moduleName': name, + 'mediaType': ext, + 'quality': '%dx%d' % (f['width'], f['height']), } - request = sanitized_Request( - '%s/training/Player/ViewClip' % self._API_BASE, - json.dumps(clip_post).encode('utf-8')) - request.add_header('Content-Type', 'application/json;charset=utf-8') format_id = '%s-%s' % (ext, quality) - clip_url = self._download_webpage( - request, display_id, 'Downloading %s URL' % format_id, fatal=False) + + try: + viewclip = self._download_json( + self._GRAPHQL_EP, display_id, + 'Downloading %s viewclip graphql' % format_id, + data=json.dumps({ + 'query': self.GRAPHQL_VIEWCLIP_TMPL % clip_post, + 'variables': {} + }).encode('utf-8'), + headers=self._GRAPHQL_HEADERS)['data']['viewClip'] + except ExtractorError: + # Still works but most likely will go soon + viewclip = self._download_json( + '%s/video/clips/viewclip' % self._API_BASE, display_id, + 'Downloading %s viewclip JSON' % format_id, fatal=False, + data=json.dumps(clip_post).encode('utf-8'), + headers={'Content-Type': 'application/json;charset=utf-8'}) # Pluralsight tracks multiple sequential calls to ViewClip API and start # to return 429 HTTP errors after some time (see - # https://github.com/rg3/youtube-dl/pull/6989). Moreover it may even lead - # to account ban (see https://github.com/rg3/youtube-dl/issues/6842). + # https://github.com/ytdl-org/youtube-dl/pull/6989). Moreover it may even lead + # to account ban (see https://github.com/ytdl-org/youtube-dl/issues/6842). # To somewhat reduce the probability of these consequences # we will sleep random amount of time before each call to ViewClip. self._sleep( random.randint(2, 5), display_id, '%(video_id)s: Waiting for %(timeout)s seconds to avoid throttling') - if not clip_url: + if not viewclip: continue - f.update({ - 'url': clip_url, - 'ext': ext, - 'format_id': format_id, - 'quality': quality_key(quality), - }) - formats.append(f) + + clip_urls = viewclip.get('urls') + if not isinstance(clip_urls, list): + continue + + for clip_url_data in clip_urls: + clip_url = clip_url_data.get('url') + if not clip_url: + continue + cdn = clip_url_data.get('cdn') + clip_f = f.copy() + clip_f.update({ + 'url': clip_url, + 'ext': ext, + 'format_id': '%s-%s' % (format_id, cdn) if cdn else format_id, + 'quality': quality_key(quality), + 'source_preference': int_or_none(clip_url_data.get('rank')), + }) + formats.append(clip_f) + self._sort_formats(formats) - # TODO: captions - # http://www.pluralsight.com/training/Player/ViewClip + cap = true - # or - # http://www.pluralsight.com/training/Player/Captions - # { a = author, cn = clip_id, lc = end, m = name } + duration = int_or_none( + clip.get('duration')) or parse_duration(clip.get('formattedDuration')) + + # TODO: other languages? + subtitles = self.extract_subtitles( + author, clip_idx, clip.get('clipId'), 'en', name, duration, display_id) return { - 'id': clip.get('clipName') or clip['name'], - 'title': '%s - %s' % (module['title'], clip['title']), - 'duration': int_or_none(clip.get('duration')) or parse_duration(clip.get('formattedDuration')), + 'id': clip_id, + 'title': title, + 'duration': duration, 'creator': author, - 'formats': formats + 'formats': formats, + 'subtitles': subtitles, } @@ -267,26 +464,34 @@ class PluralsightCourseIE(PluralsightBaseIE): # TODO: PSM cookie - course = self._download_json( - '%s/data/course/%s' % (self._API_BASE, course_id), - course_id, 'Downloading course JSON') + course = self._download_course(course_id, url, course_id) title = course['title'] + course_name = course['name'] + course_data = course['modules'] description = course.get('description') or course.get('shortDescription') - course_data = self._download_json( - '%s/data/course/content/%s' % (self._API_BASE, course_id), - course_id, 'Downloading course data JSON') - entries = [] for num, module in enumerate(course_data, 1): + author = module.get('author') + module_name = module.get('name') + if not author or not module_name: + continue for clip in module.get('clips', []): - player_parameters = clip.get('playerParameters') - if not player_parameters: + clip_index = int_or_none(clip.get('index')) + if clip_index is None: continue + clip_url = update_url_query( + '%s/player' % self._API_BASE, query={ + 'mode': 'live', + 'course': course_name, + 'author': author, + 'name': module_name, + 'clip': clip_index, + }) entries.append({ '_type': 'url_transparent', - 'url': '%s/training/player?%s' % (self._API_BASE, player_parameters), + 'url': clip_url, 'ie_key': PluralsightIE.ie_key(), 'chapter': module.get('title'), 'chapter_number': num,