Simon Cozens <[EMAIL PROTECTED]> wrote:
> On Wed, Nov 29, 2000 at 02:02:31PM -0500, Dan Sugalski wrote:
> > I'm really thinking that the lexer, parser, and tokenizer can't be
> anywhere
> > near as separate as we'd like. I think we're going to end up with a
> rather
> > odd mutant beast. Hopefully one that's understandable by reasonably
sane
> > people...
>
> This would *honestly* be my preference; I think it would be far easier
to
> write and understand than anything else. So long as it's nicely
re-entrant
> we
> should be fine. My only worry is, how do we reconcile this with the
idea of
> Perl having an easily modifiable grammar and being a good environment
for
> little-language stuff?
That's basically where I've been talking about a "creole processor", which
would in these terms be a pre-preprocessor, I imagine. This was also my
original source of confusion, since I thought that this was the primary
goal of the "pre-processor". What I've worked on is something that takes
in "creoles" or modes of perl or "little-language stuff" and turns it into
pure perl before actual interpretation. I think it's simpler and may make
more sense to turn it into perl than to have each creole spit out syntax
trees. It's the difference between a bunch of little anthill add-ons
versus a bunch of big everest add-ons, whether compiled in or linked,
whether perl or api.
David Grove
[EMAIL PROTECTED]