On 6/7/10 4:19 PM, [email protected] wrote:
Author: felixk
Date: Mon Jun 7 14:19:51 2010
New Revision: 952239
URL: http://svn.apache.org/viewvc?rev=952239&view=rev
Log:
instanceof will always return true since all jdbm.helper.Serializer are
instances of java.io.Serializable
Modified:
directory/apacheds/trunk/jdbm/src/main/java/jdbm/btree/BTree.java
Modified: directory/apacheds/trunk/jdbm/src/main/java/jdbm/btree/BTree.java
URL:
http://svn.apache.org/viewvc/directory/apacheds/trunk/jdbm/src/main/java/jdbm/btree/BTree.java?rev=952239&r1=952238&r2=952239&view=diff
==============================================================================
--- directory/apacheds/trunk/jdbm/src/main/java/jdbm/btree/BTree.java (original)
+++ directory/apacheds/trunk/jdbm/src/main/java/jdbm/btree/BTree.java Mon Jun
7 14:19:51 2010
@@ -206,7 +206,7 @@ public class BTree<K, V> implements Exte
throw new IllegalArgumentException( I18n.err( I18n.ERR_519 ) );
}
- if ( keySerializer != null&& !( keySerializer instanceof Serializable
) )
+ if ( keySerializer != null )
This is wrong ! Actually, the previous test was covering (extra
cautiously, I agree) that we can't pass a serializer which is not an
instance of Serializer. If this is not the case, then we should remove
the test and the exception.
Atm, the simple test for a keySerializer not being null ends with an
IllegalArgumentException in almost all cases, which is bad...
I will revert the commit.
--
Regards,
Cordialement,
Emmanuel Lécharny
www.nextury.com
