[antlr-interest] Re: how useful would a generic grammar "action" language be?

[micheal_jor]

Kunle had this to say (see below):

FWIW, I kinda agree with him that a statement/expression level generic
action language probably won't provide vastly improved cross-langauge
reuse due to the differences in the libraries available on each
platform. Not even for Java/C# despite their similarities.

Cheers,

Micheal
ANTLR/C#

> -----Original Message-----
> From: Kunle Odutola
> Sent: 28 October 2003 15:12
> To: Micheal J
> Subject: Re: how useful would a generic grammar "action" language be?
> 
> 
> --- In antlr-interest at yahoogroups.com, Terence Parr <parrt at c...> wrote:
> > Howdy,
> >
> > Just added a blog entry about a generic action grammar language to
> > support reuse of grammars even when they have (simple) actions:
> >
> > http://www.antlr.org/blog/antlr3/codegen.tml
> >
> > Is the idea a useful line of research?  I'll have the intermediate
form
> > underneath the covers anyway...
> 
> Micheal,
> 
> This touches on my previous comments a little doesn't it?. I still think
> that this is an ad hoc approach to an eventual "attribute 
> grammar" type
> implementation. Whilst it would be nice to have such a generic action
> grammar language, the differences in the libraries (standard and custom)
> available in the different languages suggests that a higher level action
> language is also needed to make real cross-langauge grammar reuse
possible.
> 
> What I mean by higher level (generic) action language is this: the
generic
> action language should support the sort of operations that action
code in
> grammars are typically used for - attribute 
> evaluation/access/modification,
> symbol table build/update/access, typechecking, AST
> manipulation/transformation, etc. Issues with this approach? -
resources.
> Someone would have to write basic supporting libraries for symbol
tables,
> etc....
> 
> Of course we might simply provide a framework (specified as 
> interfaces) for
> such a library that ANTLR users can fill in if they want to use 
> the higher-level language
> agnostic features. This framework might include interfaces such as
ISymbolTable,
> IContext, IScope, IDeclaration etc.. Not sure how well this gels 
> with non-OO
> and non-imperative languages though.
> 
> Kunle
> 
> > Ter
> > --
> > Professor Comp. Sci., University of San Francisco
> > Creator, ANTLR Parser Generator, http://www.antlr.org
> > Co-founder, http://www.jguru.com
> > Co-founder, http://www.knowspam.net enjoy email again!
> > Co-founder, http://www.peerscope.com pure link sharing


 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/