Merge remote-tracking branch 'AGSPhoenix/teamcoco-fix'

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

diff --git a/youtube_dl/extractor/ted.py b/youtube_dl/extractor/ted.py
index cf10be2d02cb5bc538e2561eee2c3fb5e47643c6..a8d8e8b29668f5db59f27565c21ee9064190498b 100644 (file)
--- a/youtube_dl/extractor/ted.py
+++ b/youtube_dl/extractor/ted.py
@@ -11,16 +11,21 @@ from ..utils import (
 
 
 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
             |
             ((?P<type_talk>talks)) # We have a simple talk
+            |
+            (?P<type_watch>watch)/[^/]+/[^/]+
         )
         (/lang/(.*?))? # The url may contain the language
-        /(?P<name>\w+) # Here goes the name and then ".html"
+        /(?P<name>[\w-]+) # Here goes the name and then ".html"
+        .*)$
         '''
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
         'md5': '4ea1dada91e4174b53dac2bb8ace429d',
         'info_dict': {
@@ -33,7 +38,17 @@ class TEDIE(SubtitlesInfoExtractor):
                 'actively fooling us.'),
             'uploader': 'Dan Dennett',
         }
-    }
+    }, {
+        'url': 'http://www.ted.com/watch/ted-institute/ted-bcg/vishal-sikka-the-beauty-and-power-of-algorithms',
+        'md5': '226f4fb9c62380d11b7995efa4c87994',
+        'info_dict': {
+            'id': 'vishal-sikka-the-beauty-and-power-of-algorithms',
+            'ext': 'mp4',
+            'title': 'Vishal Sikka: The beauty and power of algorithms',
+            'thumbnail': 're:^https?://.+\.jpg',
+            'description': 'Adaptive, intelligent, and consistent, algorithms are emerging as the ultimate app for everything from matching consumers to products to assessing medical diagnoses. Vishal Sikka shares his appreciation for the algorithm, charting both its inherent beauty and its growing power.',
+        }
+    }]
 
     _FORMATS_PREFERENCE = {
         'low': 1,
@@ -48,9 +63,14 @@ 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)
+        elif m.group('type_watch'):
+            return self._watch_info(url, name)
         else:
             return self._playlist_videos_info(url, name)
 
@@ -93,11 +113,14 @@ class TEDIE(SubtitlesInfoExtractor):
             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,
@@ -114,3 +137,26 @@ class TEDIE(SubtitlesInfoExtractor):
         else:
             self._downloader.report_warning(u'video doesn\'t have subtitles')
             return {}
+
+    def _watch_info(self, url, name):
+        webpage = self._download_webpage(url, name)
+
+        config_json = self._html_search_regex(
+            r"data-config='([^']+)", webpage, 'config')
+        config = json.loads(config_json)
+        video_url = config['video']['url']
+        thumbnail = config.get('image', {}).get('url')
+
+        title = self._html_search_regex(
+            r"(?s)<h1(?:\s+class='[^']+')?>(.+?)</h1>", webpage, 'title')
+        description = self._html_search_regex(
+            r'(?s)<h4 class="[^"]+" id="h3--about-this-talk">.*?</h4>(.*?)</div>',
+            webpage, 'description', fatal=False)
+
+        return {
+            'id': name,
+            'url': video_url,
+            'title': title,
+            'thumbnail': thumbnail,
+            'description': description,
+        }