Merge remote-tracking branch 'origin/wheels'

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

diff --git a/youtube_dl/extractor/ted.py b/youtube_dl/extractor/ted.py
index 10cb1e4be5b905171493efbfc76c6f4c8455fc63..3968b718e9ad7a62664c3224452e6d57dce042e0 100644 (file)
--- a/youtube_dl/extractor/ted.py
+++ b/youtube_dl/extractor/ted.py
@@ -6,27 +6,31 @@ import re
 from .subtitles import SubtitlesInfoExtractor
 
 from ..utils import (
-    RegexNotFoundError,
+    compat_str,
 )
 
 
 class TEDIE(SubtitlesInfoExtractor):
-    _VALID_URL=r'''(?x)http://www\.ted\.com/
-                   (
-                        ((?P<type_playlist>playlists)/(?P<playlist_id>\d+)) # We have a playlist
-                        |
-                        ((?P<type_talk>talks)) # We have a simple talk
-                   )
-                   (/lang/(.*?))? # The url may contain the language
-                   /(?P<name>\w+) # Here goes the name and then ".html"
-                   '''
+    _VALID_URL = r'''(?x)http://www\.ted\.com/
+        (
+            (?P<type_playlist>playlists(?:/\d+)?) # We have a playlist
+            |
+            ((?P<type_talk>talks)) # We have a simple talk
+        )
+        (/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 consciousness, but that half the time our brains are actively fooling us.',
+            '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',
         }
     }
@@ -37,46 +41,41 @@ class TEDIE(SubtitlesInfoExtractor):
         'high': 3,
     }
 
+    def _extract_info(self, webpage):
+        info_json = self._search_regex(r'q\("\w+.init",({.+})\)</script>',
+            webpage, 'info json')
+        return json.loads(info_json)
+
     def _real_extract(self, url):
-        m=re.match(self._VALID_URL, url, re.VERBOSE)
+        m = re.match(self._VALID_URL, url, re.VERBOSE)
+        name = m.group('name')
         if m.group('type_talk'):
-            return self._talk_info(url)
-        else :
-            playlist_id=m.group('playlist_id')
-            name=m.group('name')
-            self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
-            return [self._playlist_videos_info(url,name,playlist_id)]
-
+            return self._talk_info(url, name)
+        else:
+            return self._playlist_videos_info(url, name)
 
-    def _playlist_videos_info(self, url, name, playlist_id):
+    def _playlist_videos_info(self, url, name):
         '''Returns the videos of the playlist'''
 
-        webpage = self._download_webpage(
-            url, playlist_id, 'Downloading playlist webpage')
-        matches = re.finditer(
-            r'<p\s+class="talk-title[^"]*"><a\s+href="(?P<talk_url>/talks/[^"]+\.html)">[^<]*</a></p>',
-            webpage)
-
-        playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
-                                                 webpage, 'playlist title')
+        webpage = self._download_webpage(url, name,
+            'Downloading playlist webpage')
+        info = self._extract_info(webpage)
+        playlist_info = info['playlist']
 
         playlist_entries = [
-            self.url_result(u'http://www.ted.com' + m.group('talk_url'), 'TED')
-            for m in matches
+            self.url_result(u'http://www.ted.com/talks/' + talk['slug'], self.ie_key())
+            for talk in info['talks']
         ]
         return self.playlist_result(
-            playlist_entries, playlist_id=playlist_id, playlist_title=playlist_title)
+            playlist_entries,
+            playlist_id=compat_str(playlist_info['id']),
+            playlist_title=playlist_info['title'])
 
-    def _talk_info(self, url, video_id=0):
-        """Return the video for the talk in the url"""
-        m = re.match(self._VALID_URL, url)
-        video_name = m.group('name')
-        webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
+    def _talk_info(self, url, video_name):
+        webpage = self._download_webpage(url, video_name)
         self.report_extraction(video_name)
 
-        info_json = self._search_regex(r'"talkPage.init",({.+})\)</script>', webpage, 'info json')
-        info = json.loads(info_json)
-        talk_info = info['talks'][0]
+        talk_info = self._extract_info(webpage)['talks'][0]
 
         formats = [{
             'ext': 'mp4',
@@ -87,18 +86,21 @@ 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,