Updated version 2017.09.24 from 'upstream/2017.09.24'

[youtubedl] / youtube_dl / extractor / pluralsight.py

diff --git a/youtube_dl/extractor/pluralsight.py b/youtube_dl/extractor/pluralsight.py
index fd32836ccaf61b171f9dcafbbfce24c7764b18c1..f6a9131b19bf693511a84cfa3118d89030053c6d 100644 (file)
--- a/youtube_dl/extractor/pluralsight.py
+++ b/youtube_dl/extractor/pluralsight.py
@@ -1,39 +1,94 @@
 from __future__ import unicode_literals
 
 from __future__ import unicode_literals
 

-import re
+import collections

 import json
 import json

+import os
+import random

 
 from .common import InfoExtractor
 from ..compat import (
     compat_str,
 
 from .common import InfoExtractor
 from ..compat import (
     compat_str,

-    compat_urllib_parse,
-    compat_urllib_request,

     compat_urlparse,
 )
 from ..utils import (
     compat_urlparse,
 )
 from ..utils import (

+    dict_get,

     ExtractorError,
     ExtractorError,

+    float_or_none,

     int_or_none,
     parse_duration,
     int_or_none,
     parse_duration,

+    qualities,
+    srt_subtitles_timecode,
+    try_get,
+    update_url_query,
+    urlencode_postdata,

 )
 
 
 )
 
 

-class PluralsightIE(InfoExtractor):
+class PluralsightBaseIE(InfoExtractor):
+    _API_BASE = 'https://app.pluralsight.com'
+
+    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(
+            '%s/player/functions/rpc' % self._API_BASE, display_id,
+            'Downloading course JSON',
+            data=json.dumps({
+                'fn': 'bootstrapPlayer',
+                'payload': {
+                    'courseId': course_id,
+                },
+            }).encode('utf-8'),
+            headers={
+                'Content-Type': 'application/json;charset=utf-8',
+                'Referer': url,
+            })
+
+        course = try_get(response, lambda x: x['payload']['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'
     IE_NAME = 'pluralsight'

-    _VALID_URL = r'https?://(?:www\.)?pluralsight\.com/training/player\?author=(?P<author>[^&]+)&name=(?P<name>[^&]+)(?:&mode=live)?&clip=(?P<clip>\d+)&course=(?P<course>[^&]+)'
-    _LOGIN_URL = 'https://www.pluralsight.com/id/'
+    _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:training/)?player\?'
+    _LOGIN_URL = 'https://app.pluralsight.com/id/'
+

     _NETRC_MACHINE = 'pluralsight'
 
     _NETRC_MACHINE = 'pluralsight'
 

-    _TEST = {
+    _TESTS = [{

         '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',
         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
         'info_dict': {
             'id': 'hosting-sql-server-windows-azure-iaas-m7-mgmt-04',
             'ext': 'mp4',
         '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',
         'md5': '4d458cf5cf4c593788672419a8dd4cf8',
         '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',
             'duration': 338,
         },
         'skip': 'Requires pluralsight account credentials',

-    }
+    }, {
+        'url': 'https://app.pluralsight.com/training/player?course=angularjs-get-started&author=scott-allen&name=angularjs-get-started-m1-introduction&clip=0&mode=live',
+        'only_matching': True,
+    }, {
+        # 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,
+    }]

 
     def _real_initialize(self):
         self._login()
 
     def _real_initialize(self):
         self._login()


@@ -41,7 +96,7 @@ class PluralsightIE(InfoExtractor):
     def _login(self):
         (username, password) = self._get_login_info()
         if username is None:
     def _login(self):
         (username, password) = self._get_login_info()
         if username is None:

-            self.raise_login_required('Pluralsight account is required')
+            return

 
         login_page = self._download_webpage(
             self._LOGIN_URL, None, 'Downloading login page')
 
         login_page = self._download_webpage(
             self._LOGIN_URL, None, 'Downloading login page')


@@ -49,8 +104,8 @@ class PluralsightIE(InfoExtractor):
         login_form = self._hidden_inputs(login_page)
 
         login_form.update({
         login_form = self._hidden_inputs(login_page)
 
         login_form.update({

-            'Username': username.encode('utf-8'),
-            'Password': password.encode('utf-8'),
+            'Username': username,
+            'Password': password,

         })
 
         post_url = self._search_regex(
         })
 
         post_url = self._search_regex(


@@ -60,12 +115,10 @@ class PluralsightIE(InfoExtractor):
         if not post_url.startswith('http'):
             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
 
         if not post_url.startswith('http'):
             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
 

-        request = compat_urllib_request.Request(
-            post_url, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
-        request.add_header('Content-Type', 'application/x-www-form-urlencoded')
-

         response = self._download_webpage(
         response = self._download_webpage(

-            request, None, 'Logging in as %s' % username)
+            post_url, None, 'Logging in as %s' % username,
+            data=urlencode_postdata(login_form),
+            headers={'Content-Type': 'application/x-www-form-urlencoded'})

 
         error = self._search_regex(
             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',
 
         error = self._search_regex(
             r'<span[^>]+class="field-validation-error"[^>]*>([^<]+)</span>',


@@ -73,30 +126,88 @@ class PluralsightIE(InfoExtractor):
         if error:
             raise ExtractorError('Unable to login: %s' % error, expected=True)
 
         if error:
             raise ExtractorError('Unable to login: %s' % error, expected=True)
 

+        if all(p not in response for p in ('__INITIAL_STATE__', '"currentUser"')):
+            BLOCKED = 'Your account has been blocked due to suspicious activity'
+            if BLOCKED in response:
+                raise ExtractorError(
+                    'Unable to login: %s' % BLOCKED, expected=True)
+            raise ExtractorError('Unable to log in')
+
+    def _get_subtitles(self, author, clip_id, lang, name, duration, video_id):
+        captions_post = {
+            'a': author,
+            'cn': clip_id,
+            '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)),
+                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))
+            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):
     def _real_extract(self, url):

-        mobj = re.match(self._VALID_URL, url)
-        author = mobj.group('author')
-        name = mobj.group('name')
-        clip_id = mobj.group('clip')
-        course = mobj.group('course')
+        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_name = qs.get('course', [None])[0]
+
+        if any(not f for f in (author, name, clip_id, course_name,)):
+            raise ExtractorError('Invalid URL', expected=True)

 
         display_id = '%s-%s' % (name, clip_id)
 
 
         display_id = '%s-%s' % (name, clip_id)
 

-        webpage = self._download_webpage(url, display_id)
+        course = self._download_course(course_name, url, display_id)

 
 

-        collection = self._parse_json(
-            self._search_regex(
-                r'moduleCollection\s*:\s*new\s+ModuleCollection\((\[.+?\])\s*,\s*\$rootScope\)',
-                webpage, 'modules'),
-            display_id)
+        collection = course['modules']

 
 

-        module, clip = None, None
+        clip = None

 
         for module_ in collection:
 
         for module_ in collection:

-            if module_.get('moduleName') == name:
-                module = module_
+            if name in (module_.get('moduleName'), module_.get('name')):

                 for clip_ in module_.get('clips', []):
                     clip_index = clip_.get('clipIndex')
                 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 clip_index is None:
                         continue
                     if compat_str(clip_index) == clip_id:


@@ -106,67 +217,128 @@ class PluralsightIE(InfoExtractor):
         if not clip:
             raise ExtractorError('Unable to resolve clip')
 
         if not clip:
             raise ExtractorError('Unable to resolve clip')
 

+        title = clip['title']
+

         QUALITIES = {
             'low': {'width': 640, 'height': 480},
             'medium': {'width': 848, 'height': 640},
             'high': {'width': 1024, 'height': 768},
         QUALITIES = {
             'low': {'width': 640, 'height': 480},
             'medium': {'width': 848, 'height': 640},
             'high': {'width': 1024, 'height': 768},

+            'high-widescreen': {'width': 1280, 'height': 720},

         }
 
         }
 

+        QUALITIES_PREFERENCE = ('low', 'medium', 'high', 'high-widescreen',)
+        quality_key = qualities(QUALITIES_PREFERENCE)
+
+        AllowedQuality = collections.namedtuple('AllowedQuality', ['ext', 'qualities'])
+

         ALLOWED_QUALITIES = (
         ALLOWED_QUALITIES = (

-            ('webm', ('high',)),
-            ('mp4', ('low', 'medium', 'high',)),
+            AllowedQuality('webm', ['high', ]),
+            AllowedQuality('mp4', ['low', 'medium', 'high', ]),

         )
 
         )
 

+        # Some courses also offer widescreen resolution for high quality (see
+        # https://github.com/rg3/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:
+                allowed_quality.qualities.append(best_quality)
+
+        # In order to minimize the number of calls to ViewClip API and reduce
+        # the probability of being throttled or banned by Pluralsight we will request
+        # only single format until formats listing was explicitly requested.
+        if self._downloader.params.get('listformats', False):
+            allowed_qualities = ALLOWED_QUALITIES
+        else:
+            def guess_allowed_qualities():
+                req_format = self._downloader.params.get('format') or 'best'
+                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, )), )
+                req_ext = 'webm' if self._downloader.params.get('prefer_free_formats') else 'mp4'
+                return (AllowedQuality(req_ext, (best_quality, )), )
+            allowed_qualities = guess_allowed_qualities()
+

         formats = []
         formats = []

-        for ext, qualities in ALLOWED_QUALITIES:
-            for quality in qualities:
+        for ext, qualities_ in allowed_qualities:
+            for quality in qualities_:

                 f = QUALITIES[quality].copy()
                 clip_post = {
                 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_id),
+                    'courseName': course_name,
+                    'locale': 'en',
+                    'moduleName': name,
+                    'mediaType': ext,
+                    'quality': '%dx%d' % (f['width'], f['height']),

                 }
                 }

-                request = compat_urllib_request.Request(
-                    'http://www.pluralsight.com/training/Player/ViewClip',
-                    json.dumps(clip_post).encode('utf-8'))
-                request.add_header('Content-Type', 'application/json;charset=utf-8')

                 format_id = '%s-%s' % (ext, quality)
                 format_id = '%s-%s' % (ext, quality)

-                clip_url = self._download_webpage(
-                    request, display_id, 'Downloading %s URL' % format_id, fatal=False)
-                if not clip_url:
+                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).
+                # 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 viewclip:

                     continue
                     continue

-                f.update({
-                    'url': clip_url,
-                    'ext': ext,
-                    'format_id': format_id,
-                })
-                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)
 
         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_id, 'en', name, duration, display_id)

 
         return {
 
         return {

-            'id': clip['clipName'],
-            'title': '%s - %s' % (module['title'], clip['title']),
-            'duration': int_or_none(clip.get('duration')) or parse_duration(clip.get('formattedDuration')),
+            'id': clip.get('clipName') or clip['name'],
+            'title': title,
+            'duration': duration,

             'creator': author,
             'creator': author,

-            'formats': formats
+            'formats': formats,
+            'subtitles': subtitles,

         }
 
 
         }
 
 

-class PluralsightCourseIE(InfoExtractor):
+class PluralsightCourseIE(PluralsightBaseIE):

     IE_NAME = 'pluralsight:course'
     IE_NAME = 'pluralsight:course'

-    _VALID_URL = r'https?://(?:www\.)?pluralsight\.com/courses/(?P<id>[^/]+)'
-    _TEST = {
+    _VALID_URL = r'https?://(?:(?:www|app)\.)?pluralsight\.com/(?:library/)?courses/(?P<id>[^/]+)'
+    _TESTS = [{

         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',
         # Free course from Pluralsight Starter Subscription for Microsoft TechNet
         # https://offers.pluralsight.com/technet?loc=zTS3z&prod=zOTprodz&tech=zOttechz&prog=zOTprogz&type=zSOz&media=zOTmediaz&country=zUSz
         'url': 'http://www.pluralsight.com/courses/hosting-sql-server-windows-azure-iaas',


@@ -176,32 +348,52 @@ class PluralsightCourseIE(InfoExtractor):
             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
         },
         'playlist_count': 31,
             'description': 'md5:61b37e60f21c4b2f91dc621a977d0986',
         },
         'playlist_count': 31,

-    }
+    }, {
+        # available without pluralsight account
+        'url': 'https://www.pluralsight.com/courses/angularjs-get-started',
+        'only_matching': True,
+    }, {
+        'url': 'https://app.pluralsight.com/library/courses/understanding-microsoft-azure-amazon-aws/table-of-contents',
+        'only_matching': True,
+    }]

 
     def _real_extract(self, url):
         course_id = self._match_id(url)
 
         # TODO: PSM cookie
 
 
     def _real_extract(self, url):
         course_id = self._match_id(url)
 
         # TODO: PSM cookie
 

-        course = self._download_json(
-            'http://www.pluralsight.com/data/course/%s' % course_id,
-            course_id, 'Downloading course JSON')
+        course = self._download_course(course_id, url, course_id)

 
         title = course['title']
 
         title = course['title']

+        course_name = course['name']
+        course_data = course['modules']

         description = course.get('description') or course.get('shortDescription')
 
         description = course.get('description') or course.get('shortDescription')
 

-        course_data = self._download_json(
-            'http://www.pluralsight.com/data/course/content/%s' % course_id,
-            course_id, 'Downloading course data JSON')
-

         entries = []
         entries = []

-        for module in course_data:
+        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', []):
             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
                     continue

-                entries.append(self.url_result(
-                    'http://www.pluralsight.com/training/player?%s' % player_parameters,
-                    'Pluralsight'))
+                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': clip_url,
+                    'ie_key': PluralsightIE.ie_key(),
+                    'chapter': module.get('title'),
+                    'chapter_number': num,
+                    'chapter_id': module.get('moduleRef'),
+                })

 
         return self.playlist_result(entries, course_id, title, description)
 
         return self.playlist_result(entries, course_id, title, description)