[news.eclipse.modeling.m2m] Re: [ATL] "Warning, could not find mathing n

[Guillaume Gauffre <gauffre@xxxxxxx>]

I found bug 147313 which talk about iterative patterns and "Lazy rules", 
which I don't know what is it about.
Are Iterative target patterns in CalledRules implemented in ATL ?
The manual says yes !

Any idea ?

Guillaume


Guillaume Gauffre a écrit :
> Hi,
>
> I'm working on a transformation which use a CalledRule to factorize some 
> operations.
> When I compile it, I've got this error :
> "Warning, could not find matching node for ForEachOutPatternElement in 
> mode calledRuleCreate" and "... in mode calledRuleInit"
>
> It reports "ERROR: no slot reserved for variable: view used at 
> XX:XX-XX:XX" for my first Target pattern and the same for the second 
> pattern.
>
> Here is the code of my CalledRule :
> ------------------
> rule Entities(c : asur!ComputerSystem) {
>     using {
>         sources : Sequence(asur!DataExchangeRelation) =      
> c.relationShipSources->select(r|r.oclIsKindOf(asur!DataExchangeRelation) 
> and r.target.oclIsKindOf(asur!AOut));
>         targets : Sequence(asur!DataExchangeRelation) = 
> c.relationShipTargets->select(r|r.oclIsKindOf(asur!DataExchangeRelation) 
> and r.source.oclIsKindOf(asur!AIn));
>     }
>     to
>         view : distinct asurIL!E_View foreach(e in sources) (     
> Name<-'V_'.concat(c.Name.concat('To'.concat(e.target.Name))),
>             Module_ID <- 'null'
>         ),
>         control : distinct asurIL!E_Control foreach(e in targets) (     
> Name<-'C_'.concat(e.source.Name.concat('To'.concat(c.Name))),
>             Module_ID <- 'null'
>         )
>     do {
>         thisModule.model.components <- view;
>         thisModule.model.components <- control;
>     }
> }
> -------------------
> "asur" is my source metamodel and "asurIL" my target metamodel.
>
> When I use this code in my MatchedRule, it works fine, but when I define 
> this CalledRule, I've got this error.
>
> I call it in the "do" section of 3 Matched rules.
>
> If you see something bad, please tell me where I'm wrong.
>
> Thank you
>
> Guillaume