[Xsb-development] incremental tabling (fwd)

[David W. <wa...@cs...>]

Terrance Swift writes:
 > ---------- Forwarded message ----------
 > Date: Fri, 8 Jun 2007 11:52:51 +0200
 > From: Sven Kloppenburg <klo...@st...>
 > To: ts...@us...
 > Subject: incremental tabling
 > 
 > Hi,
 > 
 > 
 > I am working with the incremental tabling feature of XSB.
 > It seems to be broken. When I try the example from the manual:
 > 
 > p(X,Y) :- q(X,Y),Y =< 5.
 > :- dynamic q/2 as incremental.
 > :- use_incremental_dynamic q/2.
 > q(a,1).
 > q(b,3).
 > q(c,5).
 > q(d,7).
 > 
 > :- p(X,Y),writeln([X,Y]),fail.
 > :- incr_assert(q(d,4)).
 > :- writeln('where is the 4?').
 > :- p(X,Y),writeln([X,Y]),fail.
 > :- writeln('strange.').
 > :- p(X,4),writeln([X,Y]),fail.
 > 
 > the following is written to stdout:
 > 
 > [c,5]
 > [b,3]
 > [a,1]
 > where is the 4?
 > [c,5]
 > [b,3]
 > [a,1]
 > strange.
 > [d,_h3923]
 > 
 > 
 > so, XSB knows, what char corresponds to p(A,4), but does not know,
 > that p(_,4)
 > exists. After a "abolish_all_tables" the 4 is returned, but then the
 > incremental tabling is not incremental.
 > 
 > 
 > What am I doing wrong?
 > 
 > BTW:
 > my XSB is a recent CVS checkout (from tonight) and I am on MacosX:
 > 
 > XSB Version 3.0.1+ (CVS (Sagres)) of August 7, 2006
 > [i686-apple-darwin8.9.1; mode: optimal; engine: slg-wam; gc:
 > indirection; scheduling: local]

When I run it on my system (current CVS plus a few minor changes
completely unrelated to incremental tabling), under Windows with
cygwin, I get:

bash-3.1$ xsb
[xsb_configuration loaded]
[sysinitrc loaded]

XSB Version 3.0.1+ (CVS (Sagres)) of August 7, 2006
[i686-pc-cygwin; mode: optimal; engine: slg-wam; gc: indirection; scheduling: local]

| ?- [incr_tabling_bug_sven].
[Compiling .\incr_tabling_bug_sven]
[incr_tabling_bug_sven compiled, cpu time used: 0.0470 seconds]
[incr_tabling_bug_sven loaded]
[a,1]
[b,3]
[c,5]
where is the 4?
[a,1]
[b,3]
[c,5]
[d,4]
strange.
[d,_h3874]

yes
| ?- 

So I do get the 4.  (I did have to import incr_assert/1 from
increval.)  And I also get the same result when running the MSVC-compiled
version. 

So I can't reproduce your problem.  I'm wondering if there could be a
difference on the Mac?  Terry, can you see what result you get?

%%%% program
p(X,Y) :- q(X,Y),Y =< 5.
:- dynamic q/2 as incremental.
:- use_incremental_dynamic q/2.
q(a,1).
q(b,3).
q(c,5).
q(d,7).

:- import incr_assert/1 from increval.

:- p(X,Y),writeln([X,Y]),fail.
:- incr_assert(q(d,4)).
:- writeln('where is the 4?').
:- p(X,Y),writeln([X,Y]),fail.
:- writeln('strange.').
:- p(X,4),writeln([X,Y]),fail.
%%%% program

Thanks,
-David