Re: [mdr-users] Operation as a context of ActivityGraph
Jaroslav Drazan <[email protected]> Thu, 23 Jun 2005 18:00:33 +0200 (CEST)
| Newsgroups | gmane.comp.java.netbeans.modules.mdr.user |
|---|---|
| Message-ID | <[email protected]> |
Hi all, UML 1.5 (and less) defines context as aggregation but not composition i.e. it should be possible to set both - Namespace and Context and MDR shouldn't throw CompositionViolationException (at least in case that it is same in metamodel you really use). Otherway I applogize I didn't see UML metamodel and I thougt that context is composition (because of the Exception). Regards Jarda > Hello Martin, > > It is not possible because the context is a BehaviorFeature and > BehaviorFeature does not extends Namespace. > > > > Thanks, > > Leo > > 2005/6/23, Martin Matula <[email protected]>: > > Set the namespace to the same value as the context. > > Regards, > > Martin > > > > Leonardo Gresta Paulino Murta wrote: > > > > >Hi Jarda, > > > > > > Thanks for the hint! However, when I do not set the namespace of the > > >ActivityGraph, the ActivityGraph is not written in XMI file. In other words: > > >I do not receive the CompositionViolationException and the generated XMI > > >file correctly has an element relating the xmiid of the Operation with the > > >xmiid of the ActivityGraph, but the ActivityGraph itself is not there... > > > > > > The attached file (test.xml) is the generated XMI file. It is using > > >SPEM metamodel, but SPEM is derived from UML metamodel. The element > > >"SPEM:A_behavior_contex" relates a WorkDefinition, which extends Operation, > > >with an ActivityGraph (xmi.idref = "a7"). However, the ActivityGraph itself > > >(xmiid = "a7") is not there! Do you have any other hints? > > > > > > Thanks a lot!!! > > > > > > Leo > > > > > > > > > > > >>-----Original Message----- > > >>From: Jaroslav Drazan [mailto:[email protected]] > > >>Sent: Thursday, June 23, 2005 6:33 AM > > >>To: [email protected] > > >>Subject: Re: [mdr-users] Operation as a context of ActivityGraph > > >> > > >>Hello Leo, > > >>You can't set the composition more than once. That's the reason why all > > >>compositions defined by UML (ModelElement in Namespace, Feature in > > >>Classifier...) are optional (multiplicity is 0..1). So if you want to > > >>specify ActivityGraph for some BehaviorFeature you mustn't specify its > > >>namespace (if it has been already set you must first set it to null). > > >> > > >>Regards > > >>Jarda > > >> > > >> > > >> > > >>>Hello, > > >>> > > >>> I am using UML metamodel and trying to set the context of an > > >>>ActivityGraph to an Operation. The specification says the context of an > > >>>ActivityGraph may be a BehavioralFeature. Moreover, Operation inherits > > >>> > > >>> > > >>from > > >> > > >> > > >>>BehavioralFeature. However, a CompositionViolationException is thrown. > > >>> > > >>> > > >>It > > >> > > >> > > >>>complains that I am trying to set more than one context to the > > >>>ActivityGraph. I suppose it considers the namespace where the > > >>> > > >>> > > >>ActivityGraph > > >> > > >> > > >>>is inserted as an existing context. Am I doing something wrong? What is > > >>> > > >>> > > >>the > > >> > > >> > > >>>correct way to set the context of an ActivityGraph to an Operation? > > >>> > > >>> > > >>Where > > >> > > >> > > >>>should I insert the ActivityGraph to avoid the > > >>>CompositionViolationException? > > >>> > > >>> Thanks, > > >>> > > >>> Leo > > >>> > > >>> > > >>> > > >>____________________________________________________________ > > >>http://www.listicka.cz > > >> > > >> > > > > ____________________________________________________________ http://www.listicka.cz