Hmm, I agree it's definitely more maintainable. Maybe my question would
be better put as "why not focus on writing the visitor for the
functionality you need, instead of replacing a parser that exists".
Since you can alway replace the parser later, if they're produce the
same ASTs.

If you're doing it anyway, I'd like to hear about how the alternative
parser does when combined with the RB. Might be a good way for the RB to
become lighter as well.

Daniel

Anthony Hannan <ajh18 at cornell.edu> wrote:
> Daniel Vainsencher <danielv at netvision.net.il> wrote:
> > I think that's a cool idea [use SmaCC and RB AST for Closure Compiler].
> > Just a question, why do you use SmaCC if you're using the RB AST,
> > why not use the RB Parser?
> 
> Because RB Parser is hand-written imperative code just like our current
> parser.  SmaCC uses a simple declarative syntax (context free grammar)
> which is easier to understand and modify.
> 
> Cheers,
> Anthony