-----Original Message-----
From: Brian Ingerson [mailto:[EMAIL PROTECTED]]
Sent: Friday, December 07, 2001 5:13 PM
To: Inline Mailing List
Subject: Re: Inline::yacc/bison ?
On 07/12/01 11:09 +0100, Nadim Khemir wrote:
> What I would like to see in the future is Inline::C using Inline::Yacc (or
> equivalent) to speed up things. But Inline::Yacc should use Inline::C this
> is an egg-chicken paradox. Anyone has an idea of how to make it work
> (bootstrap).
I have no problem with replacing the P::RD code with something better.
However it must fit all definitions of "better". That said, I like the
idea of Inline::C using Inline::Yacc for parsing C as an option.
Better is something we want you to define Brian, I think better is 'as
right' but faster. Other might have other definitions
The bootstrapping problem is not really a problem at all. Inline::C will
merely autodetect Inline::Yacc, and use it. Otherwise it will default
back to Parse::RecDescent. Problem solved?
Problem not resolved. I thought Inline::Yacc would use Inline::C that would
use Inline::Yacc .... To compile the yacc lexer used by Inline::C one needs
inline::C to compile the parser generated by Inline::Yacc. You get the
point.
I'd also like to remove the dependency for P::RD from Inline altogether.
This is easily done by precompiling the Parse::RecDescent grammar into
pure Perl and just distributing that. Are there any Parse::RecDescent
gurus in the house who would like to submit a patch?
I vote for writing the grammar in Byacc.
Cheers, Brian
