Hi Jukka, > -----Original Message----- > From: Jukka Zitting [mailto:[EMAIL PROTECTED] > Sent: Thursday, February 23, 2006 3:47 AM > To: [email protected] > Subject: Re: Automatic Mixins
> Yes. Your approach of adding the mixin types as supertypes of > a primary node type is the correct one. > > Section 6.7.5 of the JCR specification states (with the > emphasis) that: > > Note that the mixin node types listed in the jcr:mixinTypes > property are those that have been *explicitly assigned* (using > Node.addMixin) to a node. It does not include mixin types that > may be among the supertypes of a node's primary type. > > Thus Jackrabbit is correct in not exposing such mixin types > in the jcr:mixinTypes property. Ah, right! Now I see. No matter how many times I've been through the spec I always seem to miss some details. :) This makes sense though as the mixin declared as a supertype cannot really be 'removed' as it forms part of the primary node type. The mixin would still be discoverable via the NodeType API though. > However I believe that there > is an error with //element(*, mix:type). AFAIK the query > should match the correct nodes even if the given type is a > (mixin) supertype. You may want to file a bug report in Jira > about this. Will do. Just need to factor it down to a small test case to ensure that the problem isn't on my end. Thanks again! Best Regards, -- Mike -------------------------------------------------------- NOTICE: If received in error, please destroy and notify sender. Sender does not waive confidentiality or privilege, and use is prohibited.
