]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/lynda.py
Merge tag 'upstream/2016.08.17'
[youtubedl] / youtube_dl / extractor / lynda.py
index 762cefa34ec35aa172102a5bbe9f78c129bdef92..a98c4c530ec4d62a18b437d5c17275daec5c3dfb 100644 (file)
 from __future__ import unicode_literals
 
 import re
 from __future__ import unicode_literals
 
 import re
-import json
 
 
-from .subtitles import SubtitlesInfoExtractor
 from .common import InfoExtractor
 from ..compat import (
 from .common import InfoExtractor
 from ..compat import (
+    compat_HTTPError,
     compat_str,
     compat_str,
-    compat_urllib_parse,
-    compat_urllib_request,
+    compat_urlparse,
 )
 from ..utils import (
     ExtractorError,
     int_or_none,
 )
 from ..utils import (
     ExtractorError,
     int_or_none,
+    urlencode_postdata,
 )
 
 
 )
 
 
-class LyndaIE(SubtitlesInfoExtractor):
+class LyndaBaseIE(InfoExtractor):
+    _SIGNIN_URL = 'https://www.lynda.com/signin'
+    _PASSWORD_URL = 'https://www.lynda.com/signin/password'
+    _USER_URL = 'https://www.lynda.com/signin/user'
+    _ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
+    _NETRC_MACHINE = 'lynda'
+
+    def _real_initialize(self):
+        self._login()
+
+    @staticmethod
+    def _check_error(json_string, key_or_keys):
+        keys = [key_or_keys] if isinstance(key_or_keys, compat_str) else key_or_keys
+        for key in keys:
+            error = json_string.get(key)
+            if error:
+                raise ExtractorError('Unable to login: %s' % error, expected=True)
+
+    def _login_step(self, form_html, fallback_action_url, extra_form_data, note, referrer_url):
+        action_url = self._search_regex(
+            r'<form[^>]+action=(["\'])(?P<url>.+?)\1', form_html,
+            'post url', default=fallback_action_url, group='url')
+
+        if not action_url.startswith('http'):
+            action_url = compat_urlparse.urljoin(self._SIGNIN_URL, action_url)
+
+        form_data = self._hidden_inputs(form_html)
+        form_data.update(extra_form_data)
+
+        try:
+            response = self._download_json(
+                action_url, None, note,
+                data=urlencode_postdata(form_data),
+                headers={
+                    'Referer': referrer_url,
+                    'X-Requested-With': 'XMLHttpRequest',
+                })
+        except ExtractorError as e:
+            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 500:
+                response = self._parse_json(e.cause.read().decode('utf-8'), None)
+                self._check_error(response, ('email', 'password'))
+            raise
+
+        self._check_error(response, 'ErrorMessage')
+
+        return response, action_url
+
+    def _login(self):
+        username, password = self._get_login_info()
+        if username is None:
+            return
+
+        # Step 1: download signin page
+        signin_page = self._download_webpage(
+            self._SIGNIN_URL, None, 'Downloading signin page')
+
+        # Already logged in
+        if any(re.search(p, signin_page) for p in (
+                'isLoggedIn\s*:\s*true', r'logout\.aspx', r'>Log out<')):
+            return
+
+        # Step 2: submit email
+        signin_form = self._search_regex(
+            r'(?s)(<form[^>]+data-form-name=["\']signin["\'][^>]*>.+?</form>)',
+            signin_page, 'signin form')
+        signin_page, signin_url = self._login_step(
+            signin_form, self._PASSWORD_URL, {'email': username},
+            'Submitting email', self._SIGNIN_URL)
+
+        # Step 3: submit password
+        password_form = signin_page['body']
+        self._login_step(
+            password_form, self._USER_URL, {'email': username, 'password': password},
+            'Submitting password', signin_url)
+
+
+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+/(\d+)-\d\.html'
-    _LOGIN_URL = 'https://www.lynda.com/login/login.aspx'
-    _NETRC_MACHINE = 'lynda'
+    _VALID_URL = r'https?://www\.lynda\.com/(?:[^/]+/[^/]+/\d+|player/embed)/(?P<id>\d+)'
 
 
-    _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.'
-
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
         'url': 'http://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
-        'md5': 'ecfc6862da89489161fb9cd5f5a6fac1',
+        # md5 is unstable
         'info_dict': {
             'id': '114408',
             'ext': 'mp4',
             'title': 'Using the exercise files',
             'duration': 68
         }
         'info_dict': {
             'id': '114408',
             'ext': 'mp4',
             'title': 'Using the exercise files',
             'duration': 68
         }
-    }
-
-    def _real_initialize(self):
-        self._login()
+    }, {
+        'url': 'https://www.lynda.com/player/embed/133770?tr=foo=1;bar=g;fizz=rt&fs=0',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group(1)
-
-        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_id = self._match_id(url)
 
 
-        if 'Status' in video_json:
-            raise ExtractorError('lynda returned error: %s' % video_json['Message'], expected=True)
+        video = self._download_json(
+            'http://www.lynda.com/ajax/player?videoId=%s&type=video' % video_id,
+            video_id, 'Downloading video JSON')
 
 
-        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)
 
 
-        video_id = compat_str(video_json['ID'])
-        duration = video_json['DurationInSeconds']
-        title = video_json['Title']
+        if video.get('HasAccess') is False:
+            self.raise_login_required('Video %s is only available for members' % video_id)
+
+        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)
 
-        if self._downloader.params.get('listsubtitles', False):
-            self._list_available_subtitles(video_id, page)
-            return
-
-        subtitles = self._fix_subtitles(self.extract_subtitles(video_id, page))
+        subtitles = self.extract_subtitles(video_id)
 
         return {
             'id': video_id,
 
         return {
             'id': video_id,
@@ -102,83 +165,37 @@ class LyndaIE(SubtitlesInfoExtractor):
             '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, subtitles):
-        if subtitles is None:
-            return subtitles  # subtitles not requested
-
-        fixed_subtitles = {}
-        for k, v in subtitles.items():
-            subs = json.loads(v)
-            if len(subs) == 0:
+    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'])
+            if m_current is None:
                 continue
                 continue
-            srt = ''
-            for pos in range(0, len(subs) - 1):
-                seq_current = subs[pos]
-                m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
-                if m_current is None:
-                    continue
-                seq_next = subs[pos + 1]
-                m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
-                if m_next is None:
-                    continue
-                appear_time = m_current.group('timecode')
-                disappear_time = m_next.group('timecode')
-                text = seq_current['Caption']
-                srt += '%s\r\n%s --> %s\r\n%s' % (str(pos), appear_time, disappear_time, text)
-            if srt:
-                fixed_subtitles[k] = srt
-        return fixed_subtitles
-
-    def _get_available_subtitles(self, video_id, webpage):
+            seq_next = subs[pos + 1]
+            m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
+            if m_next is None:
+                continue
+            appear_time = m_current.group('timecode')
+            disappear_time = m_next.group('timecode')
+            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
+
+    def _get_subtitles(self, video_id):
         url = 'http://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
         url = 'http://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
-        sub = self._download_webpage(url, None, False)
-        sub_json = json.loads(sub)
-        return {'en': url} if len(sub_json) > 0 else {}
+        subs = self._download_json(url, None, False)
+        if subs:
+            return {'en': [{'ext': 'srt', 'data': self._fix_subtitles(subs)}]}
+        else:
+            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'
 
@@ -191,37 +208,42 @@ 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)
+        if course.get('Status') == 'NotFound':
+            raise ExtractorError(
+                'Course %s does not exist' % course_id, expected=True)
 
         unaccessible_videos = 0
 
         unaccessible_videos = 0
-        videos = []
-        (username, _) = self._get_login_info()
+        entries = []
 
         # 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'])
+                video_id = video.get('ID')
+                if video_id:
+                    entries.append({
+                        '_type': 'url_transparent',
+                        'url': 'http://www.lynda.com/%s/%s-4.html' % (course_path, video_id),
+                        'ie_key': LyndaIE.ie_key(),
+                        'chapter': chapter.get('Title'),
+                        'chapter_number': int_or_none(chapter.get('ChapterIndex')),
+                        'chapter_id': compat_str(chapter.get('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)
-
-        entries = [
-            self.url_result('http://www.lynda.com/%s/%s-4.html' %
-                            (course_path, video_id),
-                            'Lynda')
-            for video_id in videos]
+            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)
 
 
-        course_title = course_json['Title']
+        course_title = course.get('Title')
+        course_description = course.get('Description')
 
 
-        return self.playlist_result(entries, course_id, course_title)
+        return self.playlist_result(entries, course_id, course_title, course_description)