Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!rutgers!umnd-cs!jwabik
From: jwabik@cs.D.UMN.EDU (Jeff Wabik)
Newsgroups: comp.unix.questions
Subject: Followup YACC & LEX Question
Message-ID: <719@umnd-cs.D.UMN.EDU>
Date: Mon, 10-Aug-87 17:05:25 EDT
Article-I.D.: umnd-cs.719
Posted: Mon Aug 10 17:05:25 1987
Date-Received: Tue, 11-Aug-87 05:40:06 EDT
Organization: U. of Minnesota-Duluth, Computer Science
Lines: 41
Keywords: Selective Parsing

Thanks to all who responded to my recent LEX question..  I managed to take
two of the 4 responses I got and merge them for a solution..  Another question
comes up, now:


I'm still having fun with preprocessors.  What I now need to do is to 
implement a (YACC) grammar to preprocess hunks of code, disregarding
code that is the actual program body wherever necesary..

For example, Here is a possibility:

	? (VAR debug_mode:boolean := TRUE,
	       unhappy_mode:boolean := TRUE; );?
	       .
	<lots of program body here #1>
	       .
	?IF (debug_mode)
	    PROCEDURE [XREF] Moo (x:integer);
	       .
	       .
	<lots of program body here #2>
	       .
	    PROCEND Moo;
	?ELSE
	    PROCEDURE [XREF] Moo2 (x:real);
	       .
	       .
	?IFEND



For clarification of my ambiguous comment above, what I need to do is to 
take a YACC grammar that will accept preprocessor directives but that has
no clue about the programming language itself.  Thus, I will need to 
disregard code that is in "<lots of program body here #1>".  I can deal
with program bodies #2 and #3 and their appropriate manipulation, but 
can't comprehend how to disregard all that middle code.  Anyone have an
idea?  Thanks in advance!


	-Jeff