base_* semantics in stereotype generalizations [message #661590] |
Fri, 25 March 2011 13:55 |
Eclipse User |
|
|
|
Originally posted by: ss.thinkersfoot.net
hey,
working with helios, i wonder what the exact semantics of base_*
references are along a generalization hierarchy of stereotypes.
given:
<<stereotype>> superStereo <|- <<stereotype>> subStereo
and
<<metaclass>> aMetaClass <|-(extends)- <<stereotype>> subStereo
and
<<subStereo>> anInstance:AMetaClass
the base_AMetaClass will be set as a tagged value in the context of the
subStereo instance.
however, if I specify OCL constraints in the context of superStereo,
then base_Classifier (or any base_* reference depending on the extension
rel) will not be set:
self.base_Classifier = null
does the OCL spec elaborate on that? is the a way in MDT/OCL to write
constraints in the context of generalising stereotypes (comforting when
it comes to constraint reuse!) while their specialising stereotypes are
actually applied.
thx for a heads up!
//stefan
|
|
|
Re: base_* semantics in stereotype generalizations [message #661767 is a reply to message #661590] |
Sun, 27 March 2011 16:32 |
Ed Willink Messages: 7655 Registered: July 2009 |
Senior Member |
|
|
Hi Stefan
I think you are now approaching the point where your UML-perspective of
Eclipse OCL exceeds that of the current committers.
I came across base_xxx when the OCL UML models started generating
validation errors (https://bugs.eclipse.org/bugs/show_bug.cgi?id=321461)
I do not really understand the problem or the fix, but when Nicolas
Rouquette joins in to question the semantics, I know there are deep
issues. At present base_xxx just seems really wierd to me, so I really
cannot help you.
Your problems may be cured or caused by the fix, depending on which
version you use.
Once I get UML profiles sort of working in the new UML-aligned code with
a modeled library, I should be in a position to start understanding the
subtleties. May be you would like to help.
Regards
Ed Willink
On 25/03/2011 06:55, Stefan Sobernig wrote:
> hey,
>
> working with helios, i wonder what the exact semantics of base_*
> references are along a generalization hierarchy of stereotypes.
>
> given:
>
> <<stereotype>> superStereo <|- <<stereotype>> subStereo
>
> and
>
> <<metaclass>> aMetaClass <|-(extends)- <<stereotype>> subStereo
>
> and
>
> <<subStereo>> anInstance:AMetaClass
>
> the base_AMetaClass will be set as a tagged value in the context of
> the subStereo instance.
>
> however, if I specify OCL constraints in the context of superStereo,
> then base_Classifier (or any base_* reference depending on the
> extension rel) will not be set:
>
> self.base_Classifier = null
>
> does the OCL spec elaborate on that? is the a way in MDT/OCL to write
> constraints in the context of generalising stereotypes (comforting
> when it comes to constraint reuse!) while their specialising
> stereotypes are actually applied.
>
> thx for a heads up!
>
> //stefan
|
|
|
Powered by
FUDForum. Page generated in 0.04041 seconds