On Sun, 10 Nov 2002, Walt Armour wrote:

> I thought about using the Criteria for this (but only for a moment)
> before deciding it would be way too much of a pain.  Yeah, some approach
> would be needed for this.  Luckily for us at the moment we don't have a
> pressing need to see if a particular field has been set.

Part of the behavior in the patch that I sent in is to throw an
IllegalStateException in the getXXX() is it is unintialized.

With the system I am working on I need to have quite a few unit tests in
place to try out different usage scenarios and make sure that another
developer won't accidentally corrupt the database and though I don't
usually need to check if a value has been set or not, I do need to know
for certain that I am not writing anything accidental into the db.

The foreign key constraints should usually catch mussed keys, but there
are other situations where it won't complain and just put incorrect data
in the db.

> That would be nice.  BTW, you mentioned in the original patch post that
> it was for RC2.  Did you mean the CVS version?  I haven't seen any
> mention of RC2 coming out yet (and I'm praying it's not as painful as
> RC1).

Yeah; that's what I meant. rc2-dev. =)

Will


--
To unsubscribe, e-mail:   <mailto:turbine-torque-dev-unsubscribe@;jakarta.apache.org>
For additional commands, e-mail: <mailto:turbine-torque-dev-help@;jakarta.apache.org>

Reply via email to