Merge branch 'atomicparsley' (closes #2436)

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

diff --git a/youtube_dl/extractor/ted.py b/youtube_dl/extractor/ted.py
index b9e65447f1f69c63fc61eab1469644070e5dc5da..be7a0f866cb0ce37d977e6ce504ff3d612af5bbb 100644 (file)
--- a/youtube_dl/extractor/ted.py
+++ b/youtube_dl/extractor/ted.py
@@ -7,38 +7,54 @@ 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
             |
             ((?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',
-        '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 '
                 'consciousness, but that half the time our brains are '
                 'actively fooling us.'),
             'uploader': 'Dan Dennett',
+            'width': 854,
         }
-    }
+    }, {
+        '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,
-        'medium': 2,
-        'high': 3,
+    _NATIVE_FORMATS = {
+        'low': {'preference': 1, 'width': 320, 'height': 180},
+        'medium': {'preference': 2, 'width': 512, 'height': 288},
+        'high': {'preference': 3, 'width': 854, 'height': 480},
     }
 
     def _extract_info(self, webpage):
@@ -48,9 +64,14 @@ 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)
 
@@ -78,26 +99,31 @@ class TEDIE(SubtitlesInfoExtractor):
         talk_info = self._extract_info(webpage)['talks'][0]
 
         formats = [{
-            'ext': 'mp4',
             'url': format_url,
             'format_id': format_id,
             'format': format_id,
-            'preference': self._FORMATS_PREFERENCE.get(format_id, -1),
         } for (format_id, format_url) in talk_info['nativeDownloads'].items()]
+        for f in formats:
+            finfo = self._NATIVE_FORMATS.get(f['format_id'])
+            if finfo:
+                f.update(finfo)
         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,
@@ -114,3 +140,26 @@ 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,
+        }