Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!bbn!bbn.com!mesard
From: mesard@bbn.com (Wayne Mesard)
Newsgroups: comp.sys.mac.programmer
Subject: Re: ?Bug? in LsC Preprocessor.
Summary: No bug.
Keywords: LsC, C, Preprocessor
Message-ID: <37855@bbn.COM>
Date: 28 Mar 89 15:38:14 GMT
References: <11796@umn-cs.CS.UMN.EDU>
Sender: news@bbn.COM
Reply-To: mesard@BBN.COM (Wayne Mesard)
Distribution: usa
Organization: Bolt Beranek and Newman Inc., Cambridge MA
Lines: 34

In article <11796@umn-cs.CS.UMN.EDU> goers@umn-cs.cs.umn.edu (Kenny Goers) writes:
>Ho Netland,
>
>I'm having a problem with a macro in LsC.  The following macro is preprocessed
>"correctly" with the Sun and Sequent C preprocessors but in LsC I get
>just a syntax error.  
>
>#define check_size(name) \
>	if (e_/**/name >= l_/**/name) { \
>	    register nsize = l_/**/name-s_/**/name+400; \
>	    name/**/buf = (char *) realloc(name/**/buf, nsize); \
>	    e_/**/name = name/**/buf + (e_/**/name-s_/**/name) + 1; \
>	    l_/**/name = name/**/buf + nsize - 5; \
>	    s_/**/name = name/**/buf + 1; \
>	}
>

As dozens of others have, no doubt, said by now, using "/**/" to
concatenate to macro arguments is an artifact of BSD's pcc.  To the best
of my knowledge it was never intended, and has certainly never been
portable!

The good news is that the ever-emergent ANSI standard does provide a
portable way of doing this (although it's nearly as ugly): the double
pound sign.  E.g.:

	#define CONCAT(A,B)  A##B

This will work in LSC.

-- 
unsigned *Wayne_Mesard();    "Newborns are slimy!  I like nice clean
MESARD@BBN.COM                babies in front-packs."
BBN, Cambridge, MA                                         -DB.