🚀 go-pugleaf

In article <1990Aug27.001721.9563@comp.vuw.ac.nz> L.Parkes@comp.vuw.ac.nz (Lloyd Parkes) writes:
>	2) Why doesn't xlobget/setvalue search the ivars of the object
>	   being sent the message, when looking for local bindings of
>	   symbols?

It does...

Through the miracle of gnuemacs tags (M-.) , I traced out the following
sequence of calls in xlisp 2.1:

(eval...)-->xleval()-->xlgetvalue()-->xlxgetvalue()-->xlobgetvalue():

[from xlsym.c]

|/* xlxgetvalue - get the value of a symbol */
|LVAL xlxgetvalue(sym)
|  LVAL sym;
|{
|    register LVAL fp,ep;
|    LVAL val;
|
|    /* check the environment list */
|    for (fp = xlenv; fp; fp = cdr(fp))
|
|	/* check for an instance variable */
+	if ((ep = car(fp)) && objectp(car(ep))) {
+	    if (xlobgetvalue(ep,sym,&val))
+		return (val);
|	}
|
|	/* check an environment stack frame */
|	else {
|	    for (; ep; ep = cdr(ep))
|		if (sym == car(car(ep)))
|		    return (cdr(car(ep)));
|	}
|
|    /* return the global value */
|    return (getvalue(sym));
|}

[from xlobj.c]

|/* xlobgetvalue - get the value of an instance variable */
|int xlobgetvalue(pair,sym,pval)
|  LVAL pair,sym,*pval;
|{
|    LVAL cls,names;
|    int ivtotal,n;
|
|    /* find the instance or class variable */
|    for (cls = cdr(pair); objectp(cls); cls = getivar(cls,SUPERCLASS)) {
|
|	/* check the instance variables */
|	names = getivar(cls,IVARS);
|	ivtotal = getivcnt(cls,IVARTOTAL);
|	for (n = ivtotal - getivcnt(cls,IVARCNT); n < ivtotal; ++n) {
|	    if (car(names) == sym) {
|		*pval = getivar(car(pair),n);
|		return (TRUE);
|	    }
|	    names = cdr(names);
|	}
|
|	/* check the class variables */
|	names = getivar(cls,CVARS);
|	for (n = 0; consp(names); ++n) {
|	    if (car(names) == sym) {
|		*pval = getelement(getivar(cls,CVALS),n);
|		return (TRUE);
|	    }
|	    names = cdr(names);
|	}
|    }
|
|    /* variable not found */
|    return (FALSE);
|}

Note the lines above marked with '+'.

>	   Why should it?

Because otherwise, what's the point of having instance variables, or OOP,
for that matter.

As to your other questions, hopefully somebody else will answer, as I am
very busy.

-------------------------------------------------------------------------------
	    Niels Mayer -- hplabs!mayer -- mayer@hplabs.hp.com
		  Human-Computer Interaction Department
		       Hewlett-Packard Laboratories
			      Palo Alto, CA.
				   *