RE: [Axiom-developer] Re: newrule from metalex.lisp

[Weiss, Juergen]

Hi,

There should be only one parse - for sure. The old one does
not support rules for the interpreter. The bootstrap of
the old parser generator is another difficulty. With at least
the possibility to use Aldor as the new compiler there
are quite a few arguments in favor of the new parser. It is
similar in style to the boot translator.

A few years ago I made an argument in favor of the
old parser on this mailing list -- as the one doing the
boot translation, the interpreter parsing and the
parsing for the SPAD compiler. But the code seems to be
quite fragile. In the mean time I think the new
parse could be the better solution.

Regards

Juergen Weiss

Juergen Weiss     | Universitaet Mainz, Zentrum fuer Datenverarbeitung,
address@hidden| 55099 Mainz, Tel: +49(6131)39-26361, FAX:
+49(6131)39-26407
 

> -----Original Message-----
> From: Gabriel Dos Reis [mailto:address@hidden 
> Sent: Thursday, August 09, 2007 9:59 PM
> To: Weiss, Juergen
> Cc: Stephen Wilson; address@hidden
> Subject: RE: [Axiom-developer] Re: newrule from metalex.lisp
> 
> Hi Juergen,  thanks for the background.
> 
> You're quite right the current situation is confusing.  The current
> interpreter uses the new parser (cparse.boot), while the compiler
> uses the old parser whose grammar is generated as you describe above.
> Except that the grammar is not actually generated anymore: 
> instead the 
> corresponding Lisp is cached.  I believe it would make sense that
> both the interpreter and the compiler uses the same parser, to reduce
> confusion in syntax, etc.  From that perspective, I would 
> think that the long
> term would be to remove the remanants.
> 
> Would you object to moving to the new parser (cparse.boot)?
> 
> -- Gaby
>