debian/patches: Removed, they came from upstream.

[youtubedl] / youtube_dl / extractor / lynda.py

diff --git a/youtube_dl/extractor/lynda.py b/youtube_dl/extractor/lynda.py
index 5dc22da22a6a5324887379fdff339f09f13d0309..d4e1ae99d4d91c887cfc2d55c29d8211fc48a80f 100644 (file)
--- a/youtube_dl/extractor/lynda.py
+++ b/youtube_dl/extractor/lynda.py
@@ -7,26 +7,99 @@ from .common import InfoExtractor
 from ..compat import (
     compat_str,
     compat_urllib_parse,
 from ..compat import (
     compat_str,
     compat_urllib_parse,

-    compat_urllib_request,

 )
 from ..utils import (
     ExtractorError,
 )
 from ..utils import (
     ExtractorError,

+    clean_html,

     int_or_none,
     int_or_none,

+    sanitized_Request,

 )
 
 
 )
 
 

-class LyndaIE(InfoExtractor):
+class LyndaBaseIE(InfoExtractor):
+    _LOGIN_URL = 'https://www.lynda.com/login/login.aspx'
+    _ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
+    _NETRC_MACHINE = 'lynda'
+
+    def _real_initialize(self):
+        self._login()
+
+    def _login(self):
+        username, password = self._get_login_info()
+        if username is None:
+            return
+
+        login_form = {
+            'username': username.encode('utf-8'),
+            'password': password.encode('utf-8'),
+            'remember': 'false',
+            'stayPut': 'false'
+        }
+        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\s*=\s*\'(?P<json>[^\']+)\';', login_page)
+        if m is not None:
+            response = m.group('json')
+            response_json = json.loads(response)
+            state = response_json['state']
+
+            if state == 'notlogged':
+                raise ExtractorError(
+                    'Unable to login, incorrect username and/or password',
+                    expected=True)
+
+            # This is when we get popup:
+            # > You're already logged in to lynda.com on two devices.
+            # > If you log in here, we'll log you out of another device.
+            # So, we need to confirm this.
+            if state == 'conflicted':
+                confirm_form = {
+                    'username': '',
+                    'password': '',
+                    'resolve': 'true',
+                    'remember': 'false',
+                    'stayPut': 'false',
+                }
+                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 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'
     IE_DESC = 'lynda.com videos'
     IE_NAME = 'lynda'
     IE_DESC = 'lynda.com videos'

-    _VALID_URL = r'https?://www\.lynda\.com/(?:[^/]+/[^/]+/\d+|player/embed)/(\d+)'
-    _LOGIN_URL = 'https://www.lynda.com/login/login.aspx'
+    _VALID_URL = r'https?://www\.lynda\.com/(?:[^/]+/[^/]+/\d+|player/embed)/(?P<id>\d+)'

     _NETRC_MACHINE = 'lynda'
 
     _NETRC_MACHINE = 'lynda'
 

-    _SUCCESSFUL_LOGIN_REGEX = r'isLoggedIn: true'

     _TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
 
     _TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
 

-    ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
-

     _TESTS = [{
         'url': 'http://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
         'md5': 'ecfc6862da89489161fb9cd5f5a6fac1',
     _TESTS = [{
         'url': 'http://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
         'md5': 'ecfc6862da89489161fb9cd5f5a6fac1',


@@ -41,56 +114,50 @@ class LyndaIE(InfoExtractor):
         'only_matching': True,
     }]
 
         'only_matching': True,
     }]
 

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

     def _real_extract(self, url):
     def _real_extract(self, url):

-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group(1)
+        video_id = self._match_id(url)

 
 

-        page = self._download_webpage('http://www.lynda.com/ajax/player?videoId=%s&type=video' % video_id, video_id,
-                                      'Downloading video JSON')
-        video_json = json.loads(page)
+        video = self._download_json(
+            'http://www.lynda.com/ajax/player?videoId=%s&type=video' % video_id,
+            video_id, 'Downloading video JSON')

 
 

-        if 'Status' in video_json:
-            raise ExtractorError('lynda returned error: %s' % video_json['Message'], expected=True)
-
-        if video_json['HasAccess'] is False:
+        if 'Status' in video:

             raise ExtractorError(
             raise ExtractorError(

-                'Video %s is only available for members. ' % video_id + self.ACCOUNT_CREDENTIALS_HINT, expected=True)
+                'lynda returned error: %s' % video['Message'], 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 = []
 
 
         formats = []
 

-        fmts = video_json.get('Formats')
+        fmts = video.get('Formats')

         if fmts:
         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:
         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),
                     '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)
 
 
         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,
 
         return {
             'id': video_id,


@@ -100,50 +167,9 @@ class LyndaIE(InfoExtractor):
             'formats': formats
         }
 
             'formats': formats
         }
 

-    def _login(self):
-        (username, password) = self._get_login_info()
-        if username is None:
-            return
-
-        login_form = {
-            'username': username,
-            'password': password,
-            'remember': 'false',
-            'stayPut': 'false'
-        }
-        request = compat_urllib_request.Request(self._LOGIN_URL, compat_urllib_parse.urlencode(login_form))
-        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)
-        if m is not None:
-            response = m.group('json')
-            response_json = json.loads(response)
-            state = response_json['state']
-
-            if state == 'notlogged':
-                raise ExtractorError('Unable to login, incorrect username and/or password', expected=True)
-
-            # This is when we get popup:
-            # > You're already logged in to lynda.com on two devices.
-            # > If you log in here, we'll log you out of another device.
-            # So, we need to confirm this.
-            if state == 'conflicted':
-                confirm_form = {
-                    'username': '',
-                    'password': '',
-                    'resolve': 'true',
-                    'remember': 'false',
-                    'stayPut': 'false',
-                }
-                request = compat_urllib_request.Request(self._LOGIN_URL, compat_urllib_parse.urlencode(confirm_form))
-                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:
-            raise ExtractorError('Unable to log in')
-

     def _fix_subtitles(self, subs):
         srt = ''
     def _fix_subtitles(self, subs):
         srt = ''

+        seq_counter = 0

         for pos in range(0, len(subs) - 1):
             seq_current = subs[pos]
             m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
         for pos in range(0, len(subs) - 1):
             seq_current = subs[pos]
             m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])


@@ -155,12 +181,14 @@ class LyndaIE(InfoExtractor):
                 continue
             appear_time = m_current.group('timecode')
             disappear_time = m_next.group('timecode')
                 continue
             appear_time = m_current.group('timecode')
             disappear_time = m_next.group('timecode')

-            text = seq_current['Caption'].lstrip()
-            srt += '%s\r\n%s --> %s\r\n%s' % (str(pos), appear_time, disappear_time, text)
+            text = seq_current['Caption'].strip()
+            if text:
+                seq_counter += 1
+                srt += '%s\r\n%s --> %s\r\n%s\r\n\r\n' % (seq_counter, appear_time, disappear_time, text)

         if srt:
             return srt
 
         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:
         url = 'http://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
         subs = self._download_json(url, None, False)
         if subs:


@@ -169,7 +197,7 @@ class LyndaIE(InfoExtractor):
             return {}
 
 
             return {}
 
 

-class LyndaCourseIE(InfoExtractor):
+class LyndaCourseIE(LyndaBaseIE):

     IE_NAME = 'lynda:course'
     IE_DESC = 'lynda.com online courses'
 
     IE_NAME = 'lynda:course'
     IE_DESC = 'lynda.com online courses'
 


@@ -182,37 +210,41 @@ class LyndaCourseIE(InfoExtractor):
         course_path = mobj.group('coursepath')
         course_id = mobj.group('courseid')
 
         course_path = mobj.group('coursepath')
         course_id = mobj.group('courseid')
 

-        page = self._download_webpage('http://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
-                                      course_id, 'Downloading course JSON')
-        course_json = json.loads(page)
+        course = self._download_json(
+            'http://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
+            course_id, 'Downloading course JSON')

 
 

-        if 'Status' in course_json and course_json['Status'] == 'NotFound':
-            raise ExtractorError('Course %s does not exist' % course_id, expected=True)
+        self._logout()
+
+        if course.get('Status') == 'NotFound':
+            raise ExtractorError(
+                'Course %s does not exist' % course_id, expected=True)

 
         unaccessible_videos = 0
         videos = []
 
         unaccessible_videos = 0
         videos = []

-        (username, _) = self._get_login_info()

 
         # Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
         # by single video API anymore
 
 
         # 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 username is None and 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
                     unaccessible_videos += 1
                     continue

-                videos.append(video['ID'])
+                if video.get('ID'):
+                    videos.append(video['ID'])

 
         if unaccessible_videos > 0:
 
         if unaccessible_videos > 0:

-            self._downloader.report_warning('%s videos are only available for members and will not be downloaded. '
-                                            % unaccessible_videos + LyndaIE.ACCOUNT_CREDENTIALS_HINT)
+            self._downloader.report_warning(
+                '%s videos are only available for members (or paid members) and will not be downloaded. '
+                % unaccessible_videos + self._ACCOUNT_CREDENTIALS_HINT)

 
         entries = [
 
         entries = [

-            self.url_result('http://www.lynda.com/%s/%s-4.html' %
-                            (course_path, video_id),
-                            'Lynda')
+            self.url_result(
+                'http://www.lynda.com/%s/%s-4.html' % (course_path, video_id),
+                'Lynda')

             for video_id in videos]
 
             for video_id in videos]
 

-        course_title = course_json['Title']
+        course_title = course.get('Title')

 
         return self.playlist_result(entries, course_id, course_title)
 
         return self.playlist_result(entries, course_id, course_title)