Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!mcsun!hp4nl!and!jos
From: jos@and.nl (J. Horsmeier)
Newsgroups: comp.lang.c
Subject: Re: LL(1) C grammar
Message-ID: <958@baby.and.nl>
Date: 19 Jun 91 10:28:11 GMT
References: <91-06-021@comp.compilers> <18959@prometheus.megatest.UUCP>
Organization: AND Software BV Rotterdam
Lines: 39

In article <18959@prometheus.megatest.UUCP> djones@megatest.UUCP (Dave Jones) writes:
>From article <91-06-021@comp.compilers>, by jos@and.nl (J. Horsmeier):
>> ... brew LL(1) parser generator ...
>> ... For several reasons (interpretation
>> versus compilation, incremental parser building etc.), we do not want to use
>> lex/yacc.
>
>
>Huh? There's nothing that can be done with an LL parser
>that can't be done with an LR parser. What do you mean by "interpretation
>versus compilation, incremental parser building etc"?

Hi there, 

I know. I do have an ANSI C LALR(1) yac description of the grammar, but
I have to use an interactive parser generator tool (home brew), which allows
a user to fiddle/diddle with a grammar on the fly. Parser and scanner generation
are both wrapped up in one. They want to use the thing for 'automated' 
reference documentation generation. It has some real nice features in it,
but it's LL(1) :-(

The LL(1) grammar is not supposed to be a full fletched 'C' grammar, e.g.
operator precedence may be left out. Recognizing user defined types is done
by the scanner etc. I am certainly *not* very optimistic about the whole
project, but we'll give it a try.

Hope this clarifies things a bit,

Jos

ps

I got a mail this morning: someone is willing to give me a LL(1) C grammar
description, I'll let you know about the results.


|O   J.A. Horsmeier AND Software B.V.        phone : +31 10 4367100   O|
|O                  Westersingel 106/108     fax   : +31 10 4367110   O|
|O                  3015 LD  Rotterdam NL    e-mail: jos@and.nl       O|