Merge remote-tracking branch 'dstftw/generic-webpage-unescape'

[youtube-dl] / youtube_dl / extractor / ted.py

diff --git a/youtube_dl/extractor/ted.py b/youtube_dl/extractor/ted.py
index b9e65447f1f69c63fc61eab1469644070e5dc5da..ad1a46c3385713056b94d2f00e38558e8ea69b1c 100644 (file)
--- a/youtube_dl/extractor/ted.py
+++ b/youtube_dl/extractor/ted.py
@@ -7,12 +7,13 @@ from .subtitles import SubtitlesInfoExtractor
 
 from ..utils import (
     compat_str,
-    RegexNotFoundError,
 )
 
 
 class TEDIE(SubtitlesInfoExtractor):
-    _VALID_URL = r'''(?x)http://www\.ted\.com/
+    _VALID_URL = r'''(?x)
+        (?P<proto>https?://)
+        (?P<type>www|embed)(?P<urlmain>\.ted\.com/
         (
             (?P<type_playlist>playlists(?:/\d+)?) # We have a playlist
             |
@@ -20,12 +21,14 @@ class TEDIE(SubtitlesInfoExtractor):
         )
         (/lang/(.*?))? # The url may contain the language
         /(?P<name>\w+) # Here goes the name and then ".html"
+        .*)$
         '''
     _TEST = {
         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
-        'file': '102.mp4',
         'md5': '4ea1dada91e4174b53dac2bb8ace429d',
         'info_dict': {
+            'id': '102',
+            'ext': 'mp4',
             'title': 'The illusion of consciousness',
             'description': ('Philosopher Dan Dennett makes a compelling '
                 'argument that not only don\'t we understand our own '
@@ -48,6 +51,9 @@ class TEDIE(SubtitlesInfoExtractor):
 
     def _real_extract(self, url):
         m = re.match(self._VALID_URL, url, re.VERBOSE)
+        if m.group('type') == 'embed':
+            desktop_url = m.group('proto') + 'www' + m.group('urlmain')
+            return self.url_result(desktop_url, 'TED')
         name = m.group('name')
         if m.group('type_talk'):
             return self._talk_info(url, name)
@@ -86,18 +92,21 @@ class TEDIE(SubtitlesInfoExtractor):
         } for (format_id, format_url) in talk_info['nativeDownloads'].items()]
         self._sort_formats(formats)
 
-        video_id = talk_info['id']
+        video_id = compat_str(talk_info['id'])
         # subtitles
         video_subtitles = self.extract_subtitles(video_id, talk_info)
         if self._downloader.params.get('listsubtitles', False):
             self._list_available_subtitles(video_id, talk_info)
             return
 
+        thumbnail = talk_info['thumb']
+        if not thumbnail.startswith('http'):
+            thumbnail = 'http://' + thumbnail
         return {
             'id': video_id,
             'title': talk_info['title'],
             'uploader': talk_info['speaker'],
-            'thumbnail': talk_info['thumb'],
+            'thumbnail': thumbnail,
             'description': self._og_search_description(webpage),
             'subtitles': video_subtitles,
             'formats': formats,