On 10.08.2011, at 11:02, Jukka Zitting wrote:

> Hi,
> 
> On Wed, Aug 10, 2011 at 10:41 AM, Lukas Kahwe Smith <[email protected]> 
> wrote:
>> i do not understand why Jackrabbit hasnt implemented access to the path in 
>> SQL2.
> 
> The more specific ISSAMENODE, ISCHILDNODE and ISDESCENDANTNODE
> constraints in SQL2 were designed to replace the rather vague
> "jcr:path LIKE ..." construct in SQL1.

But these do not replace all uses cases for wanting to filter on a path.

> Regarding your specific use case: If you need a simple way to query
> for all the day nodes, you could use a specific node type or a marker
> mixin for all those nodes. Then you could simply use a query like
> SELECT * FROM [my:day].


yeah, but then i would still have no way to do the sort i need, nor ensure that 
only day's within a certain subpath are selected:

in the end i am now just denormalizing the path and the depth:
SELECT * FROM [nt:unstructured] WHERE depth = 4 AND name LIKE '/issue/%' ORDER 
BY name DESC

if PATH and DEPTH would exist as listed on 
http://docs.jboss.org/modeshape/latest/manuals/reference/html/jcr-query-and-search.html
 i wouldnt need to denormalize at all.

regards,
Lukas Kahwe Smith
[email protected]



Reply via email to