debian/NEWS: Add link to my webpage to give more details.

[youtubedl] / youtube_dl / extractor / ted.py

diff --git a/youtube_dl/extractor/ted.py b/youtube_dl/extractor/ted.py
index ccad1c7c2087b533712b62aedaed43abcfa0bf90..8b31caa92c1e44473aa42953427b3cc2d71762f7 100644 (file)
--- a/youtube_dl/extractor/ted.py
+++ b/youtube_dl/extractor/ted.py
@@ -1,10 +1,16 @@
+from __future__ import unicode_literals
+

 import json
 import re
 
 import json
 import re
 

-from .common import InfoExtractor
+from .subtitles import SubtitlesInfoExtractor
+
+from ..utils import (
+    RegexNotFoundError,
+)

 
 
 
 

-class TEDIE(InfoExtractor):
+class TEDIE(SubtitlesInfoExtractor):

     _VALID_URL=r'''http://www\.ted\.com/
                    (
                         ((?P<type_playlist>playlists)/(?P<playlist_id>\d+)) # We have a playlist
     _VALID_URL=r'''http://www\.ted\.com/
                    (
                         ((?P<type_playlist>playlists)/(?P<playlist_id>\d+)) # We have a playlist


@@ -14,6 +20,15 @@ class TEDIE(InfoExtractor):
                    (/lang/(.*?))? # The url may contain the language
                    /(?P<name>\w+) # Here goes the name and then ".html"
                    '''
                    (/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': {
+            "description": "md5:c6fa72e6eedbd938c9caf6b2702f5922",
+            "title": "Dan Dennett: The illusion of consciousness"
+        }
+    }

 
     @classmethod
     def suitable(cls, url):
 
     @classmethod
     def suitable(cls, url):


@@ -23,33 +38,32 @@ class TEDIE(InfoExtractor):
     def _real_extract(self, url):
         m=re.match(self._VALID_URL, url, re.VERBOSE)
         if m.group('type_talk'):
     def _real_extract(self, url):
         m=re.match(self._VALID_URL, url, re.VERBOSE)
         if m.group('type_talk'):

-            return [self._talk_info(url)]
+            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)]
 
         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)]
 

-    def _playlist_videos_info(self,url,name,playlist_id=0):
+
+    def _playlist_videos_info(self, url, name, playlist_id):

         '''Returns the videos of the playlist'''
         '''Returns the videos of the playlist'''

-        video_RE=r'''
-                     <li\ id="talk_(\d+)"([.\s]*?)data-id="(?P<video_id>\d+)"
-                     ([.\s]*?)data-playlist_item_id="(\d+)"
-                     ([.\s]*?)data-mediaslug="(?P<mediaSlug>.+?)"
-                     '''
-        video_name_RE=r'<p\ class="talk-title"><a href="(?P<talk_url>/talks/(.+).html)">(?P<fullname>.+?)</a></p>'
-        webpage=self._download_webpage(url, playlist_id, 'Downloading playlist webpage')
-        m_videos=re.finditer(video_RE,webpage,re.VERBOSE)
-        m_names=re.finditer(video_name_RE,webpage)
+
+        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')
 
 
         playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
                                                  webpage, 'playlist title')
 

-        playlist_entries = []
-        for m_video, m_name in zip(m_videos,m_names):
-            talk_url='http://www.ted.com%s' % m_name.group('talk_url')
-            playlist_entries.append(self.url_result(talk_url, 'TED'))
-        return self.playlist_result(playlist_entries, playlist_id = playlist_id, playlist_title = playlist_title)
+        playlist_entries = [
+            self.url_result(u'http://www.ted.com' + m.group('talk_url'), 'TED')
+            for m in matches
+        ]
+        return self.playlist_result(
+            playlist_entries, playlist_id=playlist_id, playlist_title=playlist_title)

 
     def _talk_info(self, url, video_id=0):
         """Return the video for the talk in the url"""
 
     def _talk_info(self, url, video_id=0):
         """Return the video for the talk in the url"""


@@ -58,7 +72,7 @@ class TEDIE(InfoExtractor):
         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
         self.report_extraction(video_name)
         # If the url includes the language we get the title translated
         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
         self.report_extraction(video_name)
         # If the url includes the language we get the title translated

-        title = self._html_search_regex(r'<span id="altHeadline" >(?P<title>.*)</span>',
+        title = self._html_search_regex(r'<span .*?id="altHeadline".+?>(?P<title>.*)</span>',

                                         webpage, 'title')
         json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
                                     webpage, 'json data')
                                         webpage, 'title')
         json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
                                     webpage, 'json data')


@@ -68,12 +82,39 @@ class TEDIE(InfoExtractor):
         
         thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
                                        webpage, 'thumbnail')
         
         thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
                                        webpage, 'thumbnail')

-        info = {
-                'id': info['id'],
-                'url': info['htmlStreams'][-1]['file'],
-                'ext': 'mp4',
-                'title': title,
-                'thumbnail': thumbnail,
-                'description': desc,
-                }
-        return info
+        formats = [{
+            'ext': 'mp4',
+            'url': stream['file'],
+            'format': stream['id']
+        } for stream in info['htmlStreams']]
+
+        video_id = info['id']
+
+        # subtitles
+        video_subtitles = self.extract_subtitles(video_id, webpage)
+        if self._downloader.params.get('listsubtitles', False):
+            self._list_available_subtitles(video_id, webpage)
+            return
+
+        return {
+            'id': video_id,
+            'title': title,
+            'thumbnail': thumbnail,
+            'description': desc,
+            'subtitles': video_subtitles,
+            'formats': formats,
+        }
+
+    def _get_available_subtitles(self, video_id, webpage):
+        try:
+            options = self._search_regex(r'(?:<select name="subtitles_language_select" id="subtitles_language_select">)(.*?)(?:</select>)', webpage, 'subtitles_language_select', flags=re.DOTALL)
+            languages = re.findall(r'(?:<option value=")(\S+)"', options)
+            if languages:
+                sub_lang_list = {}
+                for l in languages:
+                    url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
+                    sub_lang_list[l] = url
+                return sub_lang_list
+        except RegexNotFoundError:
+            self._downloader.report_warning(u'video doesn\'t have subtitles')
+        return {}