about summary refs log tree commit diff
path: root/youtube_dl/extractor/yahoo.py
diff options
context:
space:
mode:
authorYen Chi Hsuan <yan12125@gmail.com>2015-05-19 00:50:24 +0800
committerYen Chi Hsuan <yan12125@gmail.com>2015-05-19 00:50:24 +0800
commitecee5724110847b832a6074c66ca4a63758100f4 (patch)
tree18839df8f37dad55cecf17749a270e1a8e0d56fa /youtube_dl/extractor/yahoo.py
parent1b0427e6c433c0b6db5e210db6e3173e19e702ed (diff)
downloadyoutube-dl-ecee5724110847b832a6074c66ca4a63758100f4.tar.gz
youtube-dl-ecee5724110847b832a6074c66ca4a63758100f4.tar.xz
youtube-dl-ecee5724110847b832a6074c66ca4a63758100f4.zip
[yahoo] Add support for closed captions (closes #5714)
Diffstat (limited to 'youtube_dl/extractor/yahoo.py')
-rw-r--r--youtube_dl/extractor/yahoo.py18
1 files changed, 18 insertions, 0 deletions
diff --git a/youtube_dl/extractor/yahoo.py b/youtube_dl/extractor/yahoo.py
index bf4e659ac..f9afbdbab 100644
--- a/youtube_dl/extractor/yahoo.py
+++ b/youtube_dl/extractor/yahoo.py
@@ -15,6 +15,7 @@ from ..utils import (
     unescapeHTML,
     ExtractorError,
     int_or_none,
+    mimetype2ext,
 )
 
 from .nbc import NBCSportsVPlayerIE
@@ -236,6 +237,22 @@ class YahooIE(InfoExtractor):
 
         self._sort_formats(formats)
 
+        closed_captions = self._html_search_regex(
+            r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
+            default='[]')
+
+        cc_json = self._parse_json(closed_captions, video_id, fatal=False)
+        subtitles = {}
+        if cc_json:
+            for closed_caption in cc_json:
+                lang = closed_caption['lang']
+                if lang not in subtitles:
+                    subtitles[lang] = []
+                subtitles[lang].append({
+                    'url': closed_caption['url'],
+                    'ext': mimetype2ext(closed_caption['content_type']),
+                })
+
         return {
             'id': video_id,
             'display_id': display_id,
@@ -244,6 +261,7 @@ class YahooIE(InfoExtractor):
             'description': clean_html(meta['description']),
             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
             'duration': int_or_none(meta.get('duration')),
+            'subtitles': subtitles,
         }