Eclipse Community Forums: Epsilon » EVL: How to restrict a context evaluation

Help

Home

Home » Modeling » Epsilon » EVL: How to restrict a context evaluation

Show: Today's Messages :: Show Polls :: Message Navigator

EVL: How to restrict a context evaluation [message #1722006]

Tue, 02 February 2016 08:49

Alireza Rouhi

Messages: 148
Registered: December 2015

Senior Member

Hi,

I have an EVL module structure like the following:

pre {
	...
}

context Model {
	...
}

context Stereotype_1 {
	...
}
.
.
.
context Stereotype_N {
	...
}

On evaluating a given UML model, the constraint list is applied to the PrimitiveTypes too. My question is: How to restrict the constraint evaluation only to the Model context?

Best regards,
Alireza

Report message to a moderator

Re: EVL: How to restrict a context evaluation [message #1722188 is a reply to message #1722006]

Wed, 03 February 2016 09:57

Antonio Garcia-Dominguez

Messages: 594
Registered: January 2010
Location: Birmingham, UK

Senior Member

Have you tried using EVL guards on the contexts and/or the constraints?

Report message to a moderator

Re: EVL: How to restrict a context evaluation [message #1722191 is a reply to message #1722188]

Wed, 03 February 2016 10:19

Alireza Rouhi

Messages: 148
Registered: December 2015

Senior Member

Dear Antonio,

Hi,

Thanks a lot for your reply.
No, I have no guard condition settings in the model context.

Is there any trick (in the form of a condition) to put on the guard section of the model context to restrict the context evaluation to a single UML design model only and not for the library PrimitiveTypes?

Best regards,
Alireza

Report message to a moderator

Re: EVL: How to restrict a context evaluation [message #1722193 is a reply to message #1722191]

Wed, 03 February 2016 10:31

Antonio Garcia-Dominguez

Messages: 594
Registered: January 2010
Location: Birmingham, UK

Senior Member

For the Model context, I guess it would be enough with:

self.name <> 'PrimitiveType'

I'm not sure about the stereotypes, however. One simple option is to make the stereotype conditions lazy and then use satisfies to trigger them from the Model instances that meet the above guard. Another option is finding a path from the stereotype to the model and then checking its name.

Report message to a moderator