|Re: [MDT-OCL] Defining OCL constraints [message #647517]
||Thu, 06 January 2011 13:16
| Ed Willink
Registered: July 2009
[Please use the OCL newsgroup for OCL queries.]
1) MDT/OCL supports Complete OCL (Clause 12 of the OMG specification)
which allows a complementary document for any meta-model. MDT/OCL 1.3.0
allows such documents to be parsed and evaluated programmatically
against Ecore or UML meta-models. MDT/OCL 3.0 Examples provides an
Xtext-based Complete OCL editor for Ecore meta-models that is useful but
far from perfect. Expect much better in MDT/OCL 3.1.0.
2) Work in progress is hiding the Ecore/UML distinctions through use of
a neutral Pivot Model. Other meta-modeling dialects can be supported by
alternate to/from Piovot model converters.
3) The pivot model support should be 100% OMG compliant so use of your
MOF constraints should be no problem.
[The pivot model will appear in the Indigo MDT/OCL 3.1.0 release as
Examples plugins that will be promoted to become the preferred API in
MDT/OCL 4.0 in Indigo+1.]
On 06/01/2011 15:42, Ashwani Kr Sharma wrote:
> Currently i am aware of process of defining the OCL constraints in EMF
> Model and using them for validation.
> My questions:
> 1) Is there any way OCL constrains can be defined in some place other
> then EMF Model ? (Motivation is the developers does not want people to
> touch their emf model. quite personal to them :) )
> 2) MDT OCL, i think support only EMF and UML. What has to be done to
> implement it for some other MOF based metamodel ? Some guidance will
> be helpful
> 3) We have a situation where we convert this MOF based metamodel into
> EMF model dynamically. So, if i am able to define OCL constraints on
> MOF based metamodel, can i somehow re-use them EMF model ?
> These questions might some very abstract. Sorry for that, but i am
> just starting with OCL and this new business scenario. :)
> Ashwani Kr Sharma
Powered by FUDForum
. Page generated in 0.02436 seconds