release 2015.05.20

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

diff --git a/youtube_dl/extractor/viki.py b/youtube_dl/extractor/viki.py
index 957e3c01ea548609f05c274eb9552b9491e68e10..cf6af1e5cdb6315d325d2bd355d384cc283a3e0c 100644 (file)
--- a/youtube_dl/extractor/viki.py
+++ b/youtube_dl/extractor/viki.py
@@ -11,6 +11,8 @@ from ..utils import (
     unescapeHTML,
     unified_strdate,
     US_RATINGS,
+    determine_ext,
+    mimetype2ext,
 )
 from .common import InfoExtractor
 
@@ -44,6 +46,19 @@ class VikiIE(InfoExtractor):
             'upload_date': '20150430',
             'title': '\'The Avengers: Age of Ultron\' Press Conference',
         }
+    }, {
+        'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
+        'info_dict': {
+            'id': '1048879v',
+            'ext': 'mp4',
+            'upload_date': '20140820',
+            'description': 'md5:54ff56d51bdfc7a30441ec967394e91c',
+            'title': 'Ankhon Dekhi',
+        },
+        'params': {
+            # requires ffmpeg
+            'skip_download': True,
+        }
     }]
 
     def _real_extract(self, url):
@@ -71,12 +86,29 @@ class VikiIE(InfoExtractor):
         req.add_header('User-Agent', self._USER_AGENT)
         info_webpage = self._download_webpage(
             req, video_id, note='Downloading info page')
-        if re.match(r'\s*<div\s+class="video-error', info_webpage):
-            raise ExtractorError(
-                'Video %s is blocked from your location.' % video_id,
-                expected=True)
-        video_url = self._html_search_regex(
-            r'<source[^>]+src="([^"]+)"', info_webpage, 'video URL')
+        err_msg = self._html_search_regex(r'<div[^>]+class="video-error[^>]+>(.+)</div>', info_webpage, 'error message', default=None)
+        if err_msg:
+            if 'not available in your region' in err_msg:
+                raise ExtractorError(
+                    'Video %s is blocked from your location.' % video_id,
+                    expected=True)
+            else:
+                raise ExtractorError('Viki said: ' + err_msg)
+        mobj = re.search(
+            r'<source[^>]+type="(?P<mime_type>[^"]+)"[^>]+src="(?P<url>[^"]+)"', info_webpage)
+        if not mobj:
+            raise ExtractorError('Unable to find video URL')
+        video_url = unescapeHTML(mobj.group('url'))
+        video_ext = mimetype2ext(mobj.group('mime_type'))
+
+        if determine_ext(video_url) == 'm3u8':
+            formats = self._extract_m3u8_formats(
+                video_url, video_id, ext=video_ext)
+        else:
+            formats = [{
+                'url': video_url,
+                'ext': video_ext,
+            }]
 
         upload_date_str = self._html_search_regex(
             r'"created_at":"([^"]+)"', info_webpage, 'upload date')
@@ -92,7 +124,7 @@ class VikiIE(InfoExtractor):
         return {
             'id': video_id,
             'title': title,
-            'url': video_url,
+            'formats': formats,
             'description': description,
             'thumbnail': thumbnail,
             'age_limit': age_limit,