Path: utzoo!utgpu!watmath!att!pacbell!ames!zodiac!joyce!unix!husc6!ginosko!xanth!mcnc!spl
From: spl@mcnc.org (Steve Lamont)
Newsgroups: comp.lang.c
Subject: Re: What I'd really like to see in an if-statement...
Message-ID: <5040@alvin.mcnc.org>
Date: 8 Aug 89 11:55:46 GMT
References: <5024@alvin.mcnc.org> <1300@atanasoff.cs.iastate.edu> <456@helios.prosys.se> <14521@bfmny0.UUCP> <1864@salgado.Solbourne.COM> <14522@bfmny0.UUCP>
Reply-To: spl@mcnc.org.UUCP (Steve Lamont)
Organization: Microelectronics Center of NC; RTP, NC
Lines: 56

In article <14522@bfmny0.UUCP> tneff@bfmny0.UUCP (Tom Neff) writes:
>I suppose I might as well point out that if speed is not crucial,
>you could always code a C function to do the triadic compare:
>
>	if (triadc(min, '<', thisval, '<=', max)) ...
>

Though the following isn't exactly what I intended (generating even more
keystrokes than the simpler and, to the mathematical among us, more
straightforward

		( a < b < c )

you don't need to code triadc() as a function.  You could indeed code it as a
macro, to wit

	#define triadc(a,b,c,d,e) ( ( a b c ) && ( c d e ) )

	...

	if ( triadc( 0.0, <, x, <, 1.0 ) )
	...

which will expand to

	if ( ( ( 0.0 < x ) && ( x < 1.0 ) ) )
	...

or at least it does on my IRIS.  Haven't tested it on other versions of CPP,
so I can't say whether it is a general solution or one unique to the
preprocessor on the IRIS 4D series of machines.

Note that while this effectively does more or less what I want, it is a kludge
at best and I don't know how one would go about implementing 

		( a < b < c )

in C without either overloading the inequality operators, breaking currently
working programs [does anyone really code something like the above to mean

		( a < b ) < c
		
in a useful, working C program???  I'd be curious to see an example where such
a construction would be meaningful.]

Perhaps in ANSI C II (K&R Strike Back?? :-) ) we could add operators .<., .>.,
.<=., and .>=. (big :-) here ).  Wouldn't you just love to code

		( a .<. b .<. c ) ?????


-- 
							spl
Steve Lamont, sciViGuy			EMail:	spl@ncsc.org
North Carolina Supercomputing Center	Phone: (919) 248-1120
Box 12732/RTP, NC 27709