jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Marcel Reutegger <marcel.reuteg...@gmx.net>
Subject Re: Bug in NodeReadMethodsTest.java?
Date Thu, 15 Sep 2005 08:23:01 GMT
Hi David,

thanks for reporting this issue, this is clearly a bug in the test code.

I've fixed it as you suggested (svn revision: 289176)

 > I assume this is the place to report this kind of problem.

yes it is, though if you feel a test case should be exclude from the 
official TCK (1.0) because it poses a problem for your implementation, 
you should follow the appeals process as defined by the JCP. We just 
host the test code here in jackrabbit.

regards
  marcel

David Caruana wrote:
> The testGetIndex() test searches for a node with same name siblings
> somewhere in the repository. It seems it will only find that node, if:
> 
> 1) The node is the one identified by the test config value
> javax.jcr.tck.NodeReadMethodsTest.testroot
> 2) Its first child is one of the same name siblings
> 
> It's possible that people may be mislead in that they believe the test
> has passed, but in fact, their same name siblings were never found.
> 
> The method locateNodeWithSameNameSiblings(Node node) within
> NodeReadMethodsTest.java looks like the culprit:
> 
> 1011          } else {
> 1012              Node returnedNode = locateNodeWithSameNameSiblings(n);
> 1013              if (n != null) {
> 1014                 return returnedNode;
> 1015              }
> 1016           }
> 
> line 1013 should be if (returnedNode != null) {
> 
> Changing to the above has allowed me to execute the test without messing
> around with my test data.
> 
> I assume this is the place to report this kind of problem.
> 
> Regards,
> David Caruana
> www.alfresco.org
> 
> 

Mime
View raw message