|
|
|
|
Re: Opening UML meta-model using oclInEcore editor [message #1062534 is a reply to message #1062532] |
Sat, 08 June 2013 11:42 |
MinhTu TonThat Messages: 38 Registered: February 2012 |
Member |
|
|
Thank you for your answer.
Well, in some situations that I encountered, I have to evaluate some constraints on the concept that does not exist in the defined language. Let's say in the meta-model we have two groups of meta-classes: A1,A2,...,An and B1,B2,...,Bn. The notion of "group", which encompasses a set of meta-classes, obviously does not exist at the meta level but a possible constraint could be, let's say "GroupA must have certain relationships with GroupB". Do you have any idea on this?
If the dynamic addition of meta-class is impossible then I have another question: Imagine we have the meta-class Group at the meta-model. Could we somehow declare a new Group object at the model-level and evaluate some constraints on it? For instance, the syntax could be somehow like: let Group g = new Group(a1: A1, a2: A2,...) in evaluate(g).
Best regards.
|
|
|
Re: Opening UML meta-model using oclInEcore editor [message #1062734 is a reply to message #1062534] |
Mon, 10 June 2013 14:47 |
Ed Willink Messages: 7655 Registered: July 2009 |
Senior Member |
|
|
Hi
Perhaps profiles and stereotypes are what you should use.
Regards
Ed Willink
On 08/06/2013 12:42, Capi Tali wrote:
> Thank you for your answer.
> Well, in some situations that I encountered, I have to evaluate some
> constraints on the concept that does not exist in the defined
> language. Let's say in the meta-model we have two groups of
> meta-classes: A1,A2,...,An and B1,B2,...,Bn. The notion of "group",
> which encompasses a set of meta-classes, obviously does not exist at
> the meta level but a possible constraint could be, let's say "GroupA
> must have certain relationships with GroupB". Do you have any idea on
> this?
>
> If the dynamic addition of meta-class is impossible then I have
> another question: Imagine we have the meta-class Group at the
> meta-model. Could we somehow declare a new Group object at the
> model-level and evaluate some constraints on it? For instance, the
> syntax could be somehow like: let Group g = new Group(a1: A1, a2:
> A2,...) in evaluate(g).
>
> Best regards.
>
|
|
|
|
|
|
Powered by
FUDForum. Page generated in 0.03469 seconds