[news.eclipse.tools.emf] Re: EMF newbie question

[mc5686@xxxxxxxxx (Mauro Condarelli)]

Thanks.
I have evidently been unclear.
I already discovered the trick of leaving out the containment="true" 
attribute.
Problem is:
If I do that the generated editor will create a temporary IProduct 
instance and will refuse to save it (because it has to be contained 
somewhere).

How do I convince EMF to generate an editor that will not generate a new 
instance, but will allow me to select an existing IProduct contained in 
another class (instance of IFirm)?

Can this be done?
I suppose the answer is "yes", because it looks like a very common 
operation, but I found no references to it.

Thanks for the pointer. I'm printing it now :)

Regards
Mauro


Ed Merks wrote:

> Mauro,

> Comments below.

> Mauro Condarelli wrote:
> > Hi Everybody,
> > I am a complete newbie to EMF and thus, please, bear with me :)
> >
> > I am trying to setup my first (after tutorials) real EMF project.
> >
> > I am using annotated java to define my model (as per tutorials, again).
> >
> > My current problem is: how to fill-up references (contained="false") 
> > as references to a "real" container (something similar to enums, but 
> > with a List of real classes).
> >
> > I have a series of Interfaces:
> > IProduct: name, description, price, ...
> > IOrder: reference to IProduct, quantity, deliveryDate, ...
> > ICustomer: name, address, ..., List<IOrder>
> > IFirm: name, address, ..., List<IProduct>, List<ICustomer>
> >
> > I am simplifying, of course.
> > My problem is the "reference to IProduct" should not be owned by 
> > IOrder, but a reference to some IProduct already defined and owned by 
> > IFirm.
> > This would be a FOREIGN KEY in SQL.
> > I would expect an editor showing a (dynamically filled) list (or 
> > combo) but I have been unable to achieve this.
> Unless you add containment="true" you should end up with a 
> non-containment reference.  Note that you can always create an Ecore 
> model and generate the code to see what @model annotations come out for 
> that.  I'd also suggest you look at the nice EMF refcard. 
> http://refcardz.dzone.com/refcardz/essential-emf
> > Can someone point me in the right direction?
> > Thanks in advance.