Yeah, that's a good point of course. I liked using a map as it very naturally expressed that there may be only one rule per "from" type, but making the API evolvable is more important I guess.

How about using ConvertGroup directly, i.e. return a Set<ConvertGroup>? In that context the name doesn't read that well, but OTOH GroupConvertionDescriptor would more or less resemble ConvertGroup. WDYT?


2013/1/31 Emmanuel Bernard <emmanuel@hibernate.org>
I'd rather have a GroupConversionDescriptor representing the from / to.
My reason is that if we add a new attribute to @ConvertGroup later, we
will be able to add it on the descriptor as well.

I think we tend to use a Set for the collection of elements in the spec.

So that would become a Set<GroupConvertionDescriptor>.

Emmanuel

On Wed 2013-01-30 22:39, Gunnar Morling wrote:
> Hi all,
>
> I think one natural way for representing group conversions in the metadata
> API would be to add a method
>
>     Map<Class<?>, Class<?>> getGroupConversions()
>
> to ParameterDescriptor, ReturnValueDescriptor and PropertyDescriptor
> (that's also how conversions are represented internally in the RI).
>
> Any thoughts on that, or other suggestions? If no one objects, I'd move
> forward and add the method to the spec and API.
>
> --Gunnar
>
>
>
> 2013/1/18 Gunnar Morling <gunnar@hibernate.org>
>
> > > We should.
> >
> > Ok, I've created https://hibernate.onjira.com/browse/BVAL-361 for this.
> >
> > --Gunnar
> >
> >
> > 2013/1/18 Emmanuel Bernard <emmanuel@hibernate.org>
> >
> >> We should.
> >>
> >> The reason the metadata API is complete is so that non Java system can
> >> propagate the constraints and logic of validation. Client side presentation
> >> framework was the canonical example.
> >>
> >> On 17 janv. 2013, at 21:44, Gunnar Morling <gunnar@hibernate.org> wrote:
> >>
> >> > Hi all,
> >> >
> >> > As the title says: should we provide a way to access configured group
> >> conversions via the meta-data API?
> >> >
> >> > I'm undecided, I don't really see where this would be useful, OTOH
> >> we've exposed all BV-related configurations via the meta-data API so far,
> >> so I guess we probably should for the sake of completeness.
> >> >
> >> > Thoughts?
> >> >
> >> > --Gunnar
> >> >
> >> > _______________________________________________
> >> > beanvalidation-dev mailing list
> >> > beanvalidation-dev@lists.jboss.org
> >> > https://lists.jboss.org/mailman/listinfo/beanvalidation-dev
> >>
> >> _______________________________________________
> >> beanvalidation-dev mailing list
> >> beanvalidation-dev@lists.jboss.org
> >> https://lists.jboss.org/mailman/listinfo/beanvalidation-dev
> >>
> >
> >

> _______________________________________________
> beanvalidation-dev mailing list
> beanvalidation-dev@lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/beanvalidation-dev

_______________________________________________
beanvalidation-dev mailing list
beanvalidation-dev@lists.jboss.org
https://lists.jboss.org/mailman/listinfo/beanvalidation-dev