If I would have the following mapping:

public class Wallet {
private int id;
private List transfers;
// getters and setters
....
}

public class Transfer {
private int id;
private int walletId;
private long timestamp;
// getters and setters
....
}

<hibernate-mapping>
<class name="Wallet" table="wallet">
<id name="id" type="int" column="id">
<generator class="assigned"/>
</id>
<bag name="transfers" table="transfer" inverse="true" cascade="save-update" >
<key column="wallet_id"/>
<one-to-many class="Transfer"/>
</bag>
<class>

<class name="Transfer" table="transfer">
<id name="id" type="int" column="id">
<generator class="assigned"/>
</id>
<property name="walletId" column="wallet_id" type="int"/>
<property name="timestamp" column="timestamp" type="long"/>
</class>
</hibernate-mapping>


Now I want to retrieve the wallet but only with the transfer collection to consist of all the tranfers within a certain time period (select on timestamp). Additionally if no transfers within that period can be found I just want the wallet with an empty collection.

What query should I use ?

Thanks,

I don't think this is possible to specify in the mapping unless your date range is always the same. (Or if your date range is contained within each Wallet instance it would also be possible.) Otherwise, I think the only way to do this is with a runtime HQL query.

You always can make a separate method (additional to "standard" getTransfers getter) like getTransfers(Date, Date) which retrieves only part of collection using collection filter (session.filter(...))

Thanks, this worked partially. I now get transfers within a give time period but if I look at sql hibernate executes it first retrieves the wallet, then selects all the transfers and only at the end are transfers selected based upon the given time period. Is there a way to avoid selecting all the transfers ?

of course - make your transfers set (or bag) lazy.
add lazy="true" attribute and Hibernate will not load the set right after loading parent object