I'm having trouble with a native query which I formulated as:

"SELECT si.inst_number, c.name, feat.name, " +
"a.POSTALCODE, a.CITY, a.STREET, a.STREETNO " +
"FROM institution si " +
"inner join FEATUREITEM feat on si.STATUS=feat.ID " +
"left outer join CONTACTS c on si.INST_CONTACT_ID = c.ID " +
"left outer join address a on c.id = a.contact_id " +
"WHERE length(si.INST_NUMBER) > 5 ORDER by si.inst_number";

The problem is, that the returned object[] has the same values for two columns in object[1] and object[2].

If I put the SQL into a SQL Tool and let it process, this brings the right values.
It seems to me that Hibernate is confused by "c.name" and "feat.name" and it sees only "name". Is that true?

And: do you think it is better to use JPA-QL here? If, how can I formulate it. I have the following associations:
Institution -> Contact -> Address, where Address may appear more than once for one Contact, whereas Institution has exactly one Contact.

Any suggestions welcome!
Carlo

Hi,

I finally found that the cause confusing Hibernate. When executing the query "SELECT si.inst_number, c.name, feat.name...", that means Hibernates sees two attributes both named "name", because no aliases were given.

When I changed the code to this: "SELECT si.inst_number as InstNumber, c.name as contactName, feat.name as instStatus" I ran into an SQLException untitl I found out that the cause was the new MySQL driver (5.1.5). When I switched back to my former version(5.0.7), everything was as expected.

Carlo