Re: [abc] Scanner

On Fri, 24 Sep 2004, Pavel Avgustinov wrote:

> >
> Errrm... do we? I don't think we do. It's just that we rely on the
> lexer's greedy match to parse "get*" as a pattern rather than a 'get' +
> '*'. Hmm... We should see if we deal with pointcuts on functions named
> get() correctly.

No, it is in the grammar where we add them as alternatives. See,
aspectj_reserved_identifier ....

So, in fact an extension needs to add an alternative into that rule when
it adds a new token to the pointcut language.

>
> The issue Oege was addressing was that (for example in the IBM slides on
> the scanner/parser, page 6), we claim "Reserved words in [the POINTCUT]
> state are only: adviceexecution, args, [..]". This does not correspond
> to the implementation, and is also clearly incorrect - as we *need* at
> least the primitive types and method modifiers to be POINTCUT keywords
> as well.

Yes, it should say "extra reserved words are only" ... I think I said that
when I gave the talk. It should be clarified in the slides. It quite
clearly says in the scanner/parser document in section 2.3.2 what the
keywords are.

Laurie

>
> As you say, the way it has always worked is that Java keywords are
> keywords everywhere. We just seemingly create the wrong impression in
> the docs.
>
> I'll try a call(* get(..)) pointcut now.
>
> - P
>
> >
> >Does that clarify anything?
> >
> >Laurie
> >
> >
> >+-------------------------------------------------------------+
> >| Laurie Hendren, Professor, School of Computer Science |
> >| McGill University |
> >| 318 McConnell Engineering Building tel: (514) 398-7391 |
> >| 3480 University Street fax: (514) 398-3883 |
> >| Montreal, Quebec H3A 2A7 hendren@cs.mcgill.ca |
> >| CANADA http://www.sable.mcgill.ca/~hendren |
> >+-------------------------------------------------------------+
> >
> >
> >
> >
> >
> >
>
>
Received on Fri Sep 24 15:22:35 2004