Sure,
If we agree that this should be the solution, then I can write the CR. Are there objections to this solution?
Best regards
Stephan
>
Original Message
> From: ext Gyorgy Rethy (ETH) [This email address is being protected from spambots. You need JavaScript enabled to view it.]
> Sent: Donnerstag, 28. November 2002 14:58
> To: This email address is being protected from spambots. You need JavaScript enabled to view it.
> Subject: Re: Name clash after import statement
>
>
> Hi,
>
> Stephen is rigth, it will not compile. But it is allowed...
>
> Unfortunately I can not find any requirement saying that the
> module name shall not clash with any definition name declared
> inside it... only names WITHIN the scope unit shall be unique
> (and even more, module is not a scope unit... notice that
> there are different rules for module identifiers, for example
> need not be unique). Probably if I'm searching a little bit further...
>
> So, I think, what Stephen says is a possible solution (I
> would say the best) of the problem but not the status how it
> is defined in he standard.
>
> Stephen, would you wrigth a CR ? (if I'm rigth and have not
> skated over something).
>
> Best Regards, György
>
> > dr György RÉTHY
> > Manager
> > Test Services, Test House, Ericsson Hungary
> > Address: H-1037 Budapest, Laborc street 1.
> > Phone: + 36 (1) 437-7006
> > Mobile: + 36 (30) 297-7862
> > Fax: + 36 (1) 437-7767
> > Mail-to: This email address is being protected from spambots. You need JavaScript enabled to view it.
>
>
> >
Original Message
> >From: Stephen TOBIES [This email address is being protected from spambots. You need JavaScript enabled to view it.]
> >Sent: Thursday, November 28, 2002 1:25 PM
> >To: This email address is being protected from spambots. You need JavaScript enabled to view it.
> >Subject: Re: Name clash after import statement
> >
> >
> >Hi Mariusz,
> >
> >please see below for comments...
> >
> >Best regards
> >
> >Stephan Tobies
> >
> >>
Original Message
> >> From: ext Jochymczyk Mariusz-AMJ011
> >> [This email address is being protected from spambots. You need JavaScript enabled to view it.]
> >> Sent: Donnerstag, 28. November 2002 12:53
> >> To: This email address is being protected from spambots. You need JavaScript enabled to view it.
> >> Subject: Name clash after import statement
> >>
> >>
> >> Hi,
> >> Could anyone please help me answer the questions below:
> >> Is this code sample semantically correct?
> >> If it is - what is the value of 'dummy' after the assignment
> >- 0 or 1?
> >
> >Neither, module a won't compile because it contains a semantic
> >error. Then name b cannot be uniquely resolved: it refers both
> >to a local entity and an imported module. This is not allowed, IMO.
> >
> >If a module can never have the same name as another entity,
> >your problem vanishes.
> >
> >> Thanks,
> >> Mariusz Jochymczyk
> >>
> >> /////////////////////////////////
> >> module b
> >> {
> >> const integer c := 0;
> >> }
> >> /////////////////////////////////
> >> module a
> >> {
> >> type record MyType {integer c};
> >> const MyType b := {1};
> >
> >first entity named 'b'
> >
> >>
> >> import from b all;
> >
> >second entity named 'b'
> >
> >>
> >> control
> >> {
> >> var integer dummy;
> >> dummy := b.c; //0 or 1?
> >> }
> >> }
> >> /////////////////////////////////
> >>
> >
>
