MissingSchema fix

lethain · Jun 1, 2014 · 1281c49 · 1281c49
1 parent 6744c44
commit 1281c49
Showing 1 changed file with 2 additions and 0 deletions.
diff --git a/extraction/__init__.py b/extraction/__init__.py
@@ -183,6 +183,8 @@ def cleanup_url(self, value_url, source_url, mark):
             url = value_url
         else:
             url = urlparse.urljoin(source_url, value_url)
+        if url.startswith('//'):
+            url = 'http:' + url # MissingSchema fix
         return url + mark
 
     def cleanup(self, results, source_url=None, mark=""):