Hello

I have a two objects Doctor and Contact to doctor.
Doctor can have more than one Contact.


i have select like this

select doctor from Doctor doctor left join fetch doctor.contactdoctorSet where doctor.specialization in (:specialization)

When doctor1 have three contacts it returns three rows with the same doctor1 like this, than other doctors with one contact only once:
[doctor1]
[doctor1]
[doctor2]
[doctor3]
...
[doctorN]


is any way how it can return collection of contacts as one record for doctor with already initialized collection

[doctor1]
[doctor2]
[doctor3]
...
[doctorN]


Maybe i can use uniqueresult but isn't it very slow ?

Can any1 help me ? plz

Thanks

one of the way would be using criteria
with

criteria.setResultTransformer(CriteriaSpecification.DISTINCT_ROOT_ENTITY);

The equivalent in HQL is to use select distinct:

thanks that is what i was asking for. Is any other way how to get unique result and do not use distinct. Distinct is very slow isn't it ?

what is faster :

using query.uniqueResult()

or

using your proposal with

criteria.setResultTransformer(CriteriaSpecification.DISTINCT_ROOT_ENTITY);

or any other way ... ?

I had the very same problem and have found a couple of ways to solve it. If you expect the result to be unique, use uniqueResult. I have personally observed distinct return dups when I joined more than two tables; it worked well when I joined two tables. The best solution seems to be code like the following:

List resultsWithDups = query.list();
Set resultSet = new HashSet(resultsWithDups);

The set doesn't allows dups so it effectively filters the list using the object's equals method.

Grant