Xref: news2.mv.net comp.os.msdos.djgpp:1069
From: Steve Higgins <steve AT alphdyn DOT demon DOT co DOT uk>
Newsgroups: comp.os.msdos.djgpp
Subject: More Complex preprocessors
Date: Tue, 13 Feb 1996 11:16:01 +0000
Lines: 60
Message-ID: <312072F1.41C6@alphdyn.demon.co.uk>
NNTP-Posting-Host: alphdyn.demon.co.uk
MIME-Version: 1.0
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp

Hi All,

Every so often I keep wishing that C/C++ had a decent macro preprocessor
of the order of most macro-assemblers. I've had a play around with the
different tools available and cannot find anything that really suits me.

Can anyone suggest anything?

Here is what I have looked at so far.

1) m4   This does give me much of what I want, however the m4 commands 
        use words which could easily be used in the c code (shift, eval, 
        index, len, etc...) causing all kinds of possible bugs.

        I would prefer each preprocessor line to be obvious ie starting
        with # or some other flag.

2) perl, awk
        These are all useful however there are a couple of problems I
	have with them. The most important of which is that I would 
	need to create a command file which would recognise my own
	preprocessor directives in the c code and execute them. This
	command file is not such a trivial thing to create. (c source
	includes headers with our macros in it, there would also be a
	requirement for an expression parser for preprocesor arithmetic
	constructs like,

        #define NUM 1
        #rept 20
        int var_NUM;
        #define NUM NUM+1
        #endr
	
	or
	#define CELLS ROWS*COLS

   	#rept CELLS
	/* some initialiser */
	#endr
	
3) sh   UNIX shells have a similar problem to (2) with the restriction
	of unix platforms (has anyone found a decent MSDOS shell which
	doesn't require me to install LINUX?).

4) emacs lisp
        It would be quite nice to be able to use this as I could then
	expand the C source inside my editor and see the resultant 
	code, however I would also need a command line equivelent to 
	run in makefiles and of course all of the issues in (2) would
	apply here as well.

What does anyone else think.



Steve.

-- 
Steve Higgins