Hi Ralf,
I finally did that:
http://jira.codehaus.org/browse/CASTOR-2615
Thanks a lot,
J
--- On Tue, 12/16/08, Ralf Joachim wrote:
>
> such an issue has not been reported before. Can you please
> create a new issue in jira about it.
>
> The behaviour you recognized is not intended but it may be
> caused by
> Castor's internal design. It is expected that the
> resultset of a query
> returns multiple records of the entity you queried (A).
> This will happen
> if the entity (A) has a 1:n relation to another entity (B).
> The records
> returned will not only contain columns of the entity
> queried (A) but
> also columns of the related entities (B). While columns of
> the queried
> entity (A) do not differ, the columns belonging to the
> related entity (B) will.
>
> To be on the save side I think we should add a default
> order to the
> original query internally as without this order the
> sequence of the
> records will depend on the underlaying database engine.
> Another option
> would be to change the loading strategy and load related
> entities with separate queries.
>
> Having said that I've started to refactor Castor's
> database layer step
> by step a while ago but as it looks at the moment this will
> take some more time.
---------------------------------------------------------------------
To unsubscribe from this list, please visit:
http://xircles.codehaus.org/manage_email
