[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
Re: [epsilon-dev] NPE in InMemoryEmfModel
|
Hello,
AbstractModel already has a field to store a/the registry, so by using its
getPackageRegistry() method everything works ok.
Without a ResourceSet the Resource creation factory does not make sense (as it
requires a ResourceSet to create/store multiple Resources), so this could be
easily skipped.
Therefore some null-checks were enough to support 'standalone resources' in
InMemoryEmfModel.
I have committed these and everything should be working as it was (besides the
new support for the 'standalone resources').
Cheers,
Maarten
On Wednesday 05 February 2014 09:24:47 Dimitris Kolovos wrote:
> Hi Maarten,
>
> When there's no resource set associated with the resource, we should
> indeed be keeping the provided ePackages in a different package
> registry. One option would be to introduce a local one and then
> override AbstractModel's getPackageRegistry().
>
> Cheers,
> Dimitris
>
> On 4 February 2014 12:21, Maarten Bezemer <maarten.bezemer@xxxxxxxxx> wrote:
> > Hello,
> >
> > I stumbled upon an NPE when InMemoryEmfModel is used for resources that
> > are
> > not part of a ResourceSet.
> >
> > The init() method, assumes there is a ResourceSet and uses it to register
> > EPackage instances. It is fairly easy to check if there is a ResourceSet
> > before registering the packages.
> >
> > But, this would result in the packages not being registered when there is
> > no ResourceSet. I do not know whether this becomes a problem later or
> > not...
> >
> > So before changing the behaviour of the init() method (and disabling
> > registering the packages when there is no ResourceSet), I was wondering
> > about your thoughts on this matter.
> >
> > Cheers,
> >
> > Maarten
> >
> > _______________________________________________
> > epsilon-dev mailing list
> > epsilon-dev@xxxxxxxxxxx
> > https://dev.eclipse.org/mailman/listinfo/epsilon-dev