Xref: utzoo comp.lang.misc:7533 comp.object:3256
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!hsdndev!cmcl2!sbcs!eeserv1.ic.sunysb.edu!jallen
From: jallen@eeserv1.ic.sunysb.edu (Joseph Allen)
Newsgroups: comp.lang.misc,comp.object
Subject: Re:  Type Systems and Dynamic Binding
Message-ID: <1991Apr19.085233.2352@sbcs.sunysb.edu>
Date: 19 Apr 91 08:52:33 GMT
References: <3843@ssc-bee.ssc-vax.UUCP> <27313:Apr1623:30:0391@kramden.acf.nyu.edu> <17APR91.21054415@uc780.umd.edu>
Sender: usenet@sbcs.sunysb.edu (Usenet poster)
Organization: State University of New York at Stony Brook
Lines: 36

In article <17APR91.21054415@uc780.umd.edu> cs450a03@uc780.umd.edu writes:
>Let's say I was trying to implement a dynamically typed language in C.
>(Of course, I wouldn't _really_ be that crazy  :-)  Let's say I wanted
>to detect when, for instance, the addition of 2 integers produces an
>invalid result.  Let's say that I can not accept a severe performance
>hit for this checking (most of the time, it's not going to happen.
>When it does happen, I'm going to switch to float -- performance must
>be better than floating point).

>How do I do it?

#define add(a,b) (a<0 && b<0 && a+b>=0 || a>=0 && b>=0 && a+b<0 ? error():a+b)

:-)

I believe C is not really the problem.  What you want is for the processor
to give an integer overflow signal so that the calculation can be repeated
in floating point.  Unfortunately most (all?) don't.

Even in an interpretive language you'd have to:

	add
	bov error
	sub
	bov error
	mul
	bov error

which is quite gross.
--
#define h 23 /* Height */         /* jallen@ic.sunysb.edu (129.49.12.74) */
#define w 79 /* Width */                       /* Amazing */
int i,r,b[]={-w,w,1,-1},d,a[w*h];m(p){a[p]=2;while(d=(p>2*w?!a[p-w-w]?1:0:0)|(
p<w*(h-2)?!a[p+w+w]?2:0:0)|(p%w!=w-2?!a[p+2]?4:0:0)|(p%w!=1?!a[p-2]?8:0:0)){do
i=3&(r=(r*57+1))/d;while(!(d&(1<<i)));a[p+b[i]]=2;m(p+2*b[i]);}}main(){r=time(
0L);m(w+1);for(i=0;i%w?0:printf("\n"),i!=w*h;i++)printf("#\0 "+a[i]);}