[
https://issues.apache.org/jira/browse/JCR-811?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12489106
]
Julian Reschke commented on JCR-811:
------------------------------------
Any preferences about using existing name (such as "nodetype" or
"propertyname1") vs inventing new ones? Re-using the existing names would
affect existing TCK setups...
> SetPropertyAssumeTypeTest check for non-protected string array property
> -----------------------------------------------------------------------
>
> Key: JCR-811
> URL: https://issues.apache.org/jira/browse/JCR-811
> Project: Jackrabbit
> Issue Type: Bug
> Components: JCR TCK
> Reporter: Julian Reschke
> Assigned To: Julian Reschke
> Priority: Minor
>
> SetPropertyAssumeTypeTest.testValuesConstraintViolationExceptionBecauseOfInvalidTypeParameter
> tries to find a property definition for a writable, multivalued string
> property. It consults NodeTypeUtil.locatePropertyDef() for that purpose.
> In my setup, the property definition being returned is for
> jcr:valueConstraints, defined on nt:propertyDefinition. Nodes of that type in
> turn can not be created on the test node, thus the test fails already when
> trying to create the node.
> It seems the test suite tries to be too smart here. Can we change this so
> that the node type and the property name are configuration parameters?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
