I've got lazy loading collections working fine in other areas on the same code, but I can't figure out why a select on a Licence class which has a lazy=true one-to-many collection would result in a version update on Licence? I don't actually want the collection, and the collection is not brought back as can be seen in the log ([com.med.dom.Licence.points#1] (uninitialized)), but then AbstractFlushingEventListener says there is 1 update - within miliseconds of each other in the log??

Its been half a day now on this and I need some help!

Thanks,
adam

hql: from License where entitiyId = x;

2006-09-14 15:33:05,312 DEBUG [org.hibernate.engine.TwoPhaseLoad] - done materializing entity [com.med.dom.Licence#1]
2006-09-14 15:33:05,312 DEBUG [org.hibernate.engine.StatefulPersistenceContext] - initializing non-lazy collections
2006-09-14 15:33:05,312 DEBUG [org.hibernate.jdbc.ConnectionManager] - transaction completed on session with on_close connection release mode; be sure to close the session to release JDBC resources!
2006-09-14 15:33:05,312 DEBUG [org.hibernate.event.def.AbstractFlushingEventListener] - processing flush-time cascades
2006-09-14 15:33:05,312 DEBUG [org.hibernate.event.def.AbstractFlushingEventListener] - dirty checking collections
2006-09-14 15:33:05,312 DEBUG [org.hibernate.engine.Collections] - Collection found: [com.med.dom.Licence.points#1], was: [com.med.dom.Licence.points#1] (uninitialized)
2006-09-14 15:33:05,312 DEBUG [org.hibernate.event.def.AbstractFlushingEventListener] - Flushed: 0 insertions, 1 updates, 0 deletions to 1 objects
2006-09-14 15:33:05,312 DEBUG [org.hibernate.event.def.AbstractFlushingEventListener] - Flushed: 0 (re)creations, 0 updates, 0 removals to 1 collections
2006-09-14 15:33:05,312 DEBUG [org.hibernate.pretty.Printer] - listing entities:
2006-09-14 15:33:05,328 DEBUG [org.hibernate.pretty.Printer] - com.med.dom.Licence{roadSideExpiry=component[day,month,year]{month=null, day=null, year=null}, usersId=10, type=, notes=null, algorithm=null, points=<uninitialized>, id=1, roadSideCalloutNumber=null, version=6, internationalNumber=null, noClaimsStart=component[day,month,year]{month=null, day=null, year=null}, roadSideAccount=null, roadSideNumber=null, name=test54, roadSideInsurer=null, number=}
2006-09-14 15:33:05,328 DEBUG [org.hibernate.jdbc.AbstractBatcher] - about to open PreparedStatement (open PreparedStatements: 0, globally: 0)
2006-09-14 15:33:05,328 DEBUG [org.hibernate.SQL] - update licences set version=?, usersId=?, algorithm=?, name=?, type=?, number=?, internationalnumber=?, noclaimsstartday=?, noclaimsstartmonth=?, noclaimsstartyear=?, roadsideinsurer=?, roadsideaccount=?, roadsidenumber=?, roadsidecalloutnumber=?, roadsideexpiryday=?, roadsideexpirymonth=?, roadsideexpiryyear=?, notes=? where id=? and version=?
2006-09-14 15:33:05,359 DEBUG [org.hibernate.jdbc.AbstractBatcher] - about to close PreparedStatement (open PreparedStatements: 1, globally: 1)
2006-09-14 15:33:05,359 DEBUG [org.hibernate.jdbc.ConnectionManager] - releasing JDBC connection [ (open PreparedStatements: 0, globally: 0) (open ResultSets: 0, globally: 0)]
2006-09-14 15:33:05,359 DEBUG [org.hibernate.jdbc.ConnectionManager] - transaction completed on session with on_close connection release mode; be sure to close the session to release JDBC resources!


mapping file:

<class name="Licence" table="licences" lazy="true">

<id name="id" column="id" type="long" unsaved-value="null">
<generator class="identity"/>
</id>
<version name="version"/>

...

<set name="points" cascade="all-delete-orphan" lazy="true" inverse="true" order-by="id asc">
<key column="licencesId"/>
<one-to-many class="LicencePoint"/>
</set>

I believe because of the subtleness of the 'error' - in that it is not an error I just had noticed the version numbers changing all the time - I had not noticed the change when I upgraded from a previous version.

In the old version (2?) I did things like:

public void setBirthdate(MedDate birthdate) {
if (birthdate == null) this.birthdate = new MedDate();
else this.birthdate = birthdate;
}

because my view would complain when the component was null. So when hibernate assigns a null component, it actually creates an object with null default values.

In the new version (3.2) it appears that hibernate now detects this as a change, and hence the version changes when none of the data appears to (because the new value also contains nulls!)

So it was nothing to do with the collection after all.

Phew!

in case people are unsure what solutiuon I did....(heopfully others can comment of the approach)


in the mapping file:

<component ... access="field">

in the java file:

change

public void setBirthdate(MedDate birthdate) {
if (birthdate == null) this.birthdate = new MedDate();
else this.birthdate = birthdate;
}

to be along the lines of a normal bean property accessor

set...(...) {
this.prop = that;
}

but change the get to be (obviously ignore the actual method names):

public MedDate getNoClaimsStart() {
if (noClaimsStart == null) this.noClaimsStart = new MedDate();
return noClaimsStart;
}

that way hibernate is happy the views are happy (that there is no null object) and I am happy!