]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/lynda.py
Merge tag 'upstream/2015.11.27.1'
[youtubedl] / youtube_dl / extractor / lynda.py
index cfd3b14f4bfd755a7600701e86900ece12b0c3ac..d4e1ae99d4d91c887cfc2d55c29d8211fc48a80f 100644 (file)
@@ -7,17 +7,17 @@ from .common import InfoExtractor
 from ..compat import (
     compat_str,
     compat_urllib_parse,
-    compat_urllib_request,
 )
 from ..utils import (
     ExtractorError,
+    clean_html,
     int_or_none,
+    sanitized_Request,
 )
 
 
 class LyndaBaseIE(InfoExtractor):
     _LOGIN_URL = 'https://www.lynda.com/login/login.aspx'
-    _SUCCESSFUL_LOGIN_REGEX = r'isLoggedIn: true'
     _ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
     _NETRC_MACHINE = 'lynda'
 
@@ -25,23 +25,23 @@ class LyndaBaseIE(InfoExtractor):
         self._login()
 
     def _login(self):
-        (username, password) = self._get_login_info()
+        username, password = self._get_login_info()
         if username is None:
             return
 
         login_form = {
-            'username': username,
-            'password': password,
+            'username': username.encode('utf-8'),
+            'password': password.encode('utf-8'),
             'remember': 'false',
             'stayPut': 'false'
         }
-        request = compat_urllib_request.Request(
-            self._LOGIN_URL, compat_urllib_parse.urlencode(login_form))
+        request = sanitized_Request(
+            self._LOGIN_URL, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
         login_page = self._download_webpage(
             request, None, 'Logging in as %s' % username)
 
         # Not (yet) logged in
-        m = re.search(r'loginResultJson = \'(?P<json>[^\']+)\';', login_page)
+        m = re.search(r'loginResultJson\s*=\s*\'(?P<json>[^\']+)\';', login_page)
         if m is not None:
             response = m.group('json')
             response_json = json.loads(response)
@@ -64,15 +64,33 @@ class LyndaBaseIE(InfoExtractor):
                     'remember': 'false',
                     'stayPut': 'false',
                 }
-                request = compat_urllib_request.Request(
-                    self._LOGIN_URL, compat_urllib_parse.urlencode(confirm_form))
+                request = sanitized_Request(
+                    self._LOGIN_URL, compat_urllib_parse.urlencode(confirm_form).encode('utf-8'))
                 login_page = self._download_webpage(
                     request, None,
                     'Confirming log in and log out from another device')
 
-        if re.search(self._SUCCESSFUL_LOGIN_REGEX, login_page) is None:
+        if all(not re.search(p, login_page) for p in ('isLoggedIn\s*:\s*true', r'logout\.aspx', r'>Log out<')):
+            if 'login error' in login_page:
+                mobj = re.search(
+                    r'(?s)<h1[^>]+class="topmost">(?P<title>[^<]+)</h1>\s*<div>(?P<description>.+?)</div>',
+                    login_page)
+                if mobj:
+                    raise ExtractorError(
+                        'lynda returned error: %s - %s'
+                        % (mobj.group('title'), clean_html(mobj.group('description'))),
+                        expected=True)
             raise ExtractorError('Unable to log in')
 
+    def _logout(self):
+        username, _ = self._get_login_info()
+        if username is None:
+            return
+
+        self._download_webpage(
+            'http://www.lynda.com/ajax/logout.aspx', None,
+            'Logging out', 'Unable to log out', fatal=False)
+
 
 class LyndaIE(LyndaBaseIE):
     IE_NAME = 'lynda'
@@ -99,52 +117,47 @@ class LyndaIE(LyndaBaseIE):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        page = self._download_webpage(
+        video = self._download_json(
             'http://www.lynda.com/ajax/player?videoId=%s&type=video' % video_id,
             video_id, 'Downloading video JSON')
-        video_json = json.loads(page)
 
-        if 'Status' in video_json:
+        if 'Status' in video:
             raise ExtractorError(
-                'lynda returned error: %s' % video_json['Message'], expected=True)
+                'lynda returned error: %s' % video['Message'], expected=True)
 
-        if video_json['HasAccess'] is False:
-            raise ExtractorError(
-                'Video %s is only available for members. '
-                % video_id + self._ACCOUNT_CREDENTIALS_HINT, expected=True)
+        if video.get('HasAccess') is False:
+            self.raise_login_required('Video %s is only available for members' % video_id)
 
-        video_id = compat_str(video_json['ID'])
-        duration = video_json['DurationInSeconds']
-        title = video_json['Title']
+        video_id = compat_str(video.get('ID') or video_id)
+        duration = int_or_none(video.get('DurationInSeconds'))
+        title = video['Title']
 
         formats = []
 
-        fmts = video_json.get('Formats')
+        fmts = video.get('Formats')
         if fmts:
-            formats.extend([
-                {
-                    'url': fmt['Url'],
-                    'ext': fmt['Extension'],
-                    'width': fmt['Width'],
-                    'height': fmt['Height'],
-                    'filesize': fmt['FileSize'],
-                    'format_id': str(fmt['Resolution'])
-                } for fmt in fmts])
-
-        prioritized_streams = video_json.get('PrioritizedStreams')
+            formats.extend([{
+                'url': f['Url'],
+                'ext': f.get('Extension'),
+                'width': int_or_none(f.get('Width')),
+                'height': int_or_none(f.get('Height')),
+                'filesize': int_or_none(f.get('FileSize')),
+                'format_id': compat_str(f.get('Resolution')) if f.get('Resolution') else None,
+            } for f in fmts if f.get('Url')])
+
+        prioritized_streams = video.get('PrioritizedStreams')
         if prioritized_streams:
-            formats.extend([
-                {
+            for prioritized_stream_id, prioritized_stream in prioritized_streams.items():
+                formats.extend([{
                     'url': video_url,
                     'width': int_or_none(format_id),
-                    'format_id': format_id,
-                } for format_id, video_url in prioritized_streams['0'].items()
-            ])
+                    'format_id': '%s-%s' % (prioritized_stream_id, format_id),
+                } for format_id, video_url in prioritized_stream.items()])
 
         self._check_formats(formats, video_id)
         self._sort_formats(formats)
 
-        subtitles = self.extract_subtitles(video_id, page)
+        subtitles = self.extract_subtitles(video_id)
 
         return {
             'id': video_id,
@@ -175,7 +188,7 @@ class LyndaIE(LyndaBaseIE):
         if srt:
             return srt
 
-    def _get_subtitles(self, video_id, webpage):
+    def _get_subtitles(self, video_id):
         url = 'http://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
         subs = self._download_json(url, None, False)
         if subs:
@@ -197,12 +210,13 @@ class LyndaCourseIE(LyndaBaseIE):
         course_path = mobj.group('coursepath')
         course_id = mobj.group('courseid')
 
-        page = self._download_webpage(
+        course = self._download_json(
             'http://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
             course_id, 'Downloading course JSON')
-        course_json = json.loads(page)
 
-        if 'Status' in course_json and course_json['Status'] == 'NotFound':
+        self._logout()
+
+        if course.get('Status') == 'NotFound':
             raise ExtractorError(
                 'Course %s does not exist' % course_id, expected=True)
 
@@ -212,12 +226,13 @@ class LyndaCourseIE(LyndaBaseIE):
         # Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
         # by single video API anymore
 
-        for chapter in course_json['Chapters']:
-            for video in chapter['Videos']:
-                if video['HasAccess'] is False:
+        for chapter in course['Chapters']:
+            for video in chapter.get('Videos', []):
+                if video.get('HasAccess') is False:
                     unaccessible_videos += 1
                     continue
-                videos.append(video['ID'])
+                if video.get('ID'):
+                    videos.append(video['ID'])
 
         if unaccessible_videos > 0:
             self._downloader.report_warning(
@@ -230,6 +245,6 @@ class LyndaCourseIE(LyndaBaseIE):
                 'Lynda')
             for video_id in videos]
 
-        course_title = course_json['Title']
+        course_title = course.get('Title')
 
         return self.playlist_result(entries, course_id, course_title)