Using SHOP variables in Lisp

[ko56]

Is it possible to access the value of a SHOP variable, say that of an atom, in a Lisp function?

[rpgoldman]

You can do this in the context of the theorem-prover. In the theorem-prover, what actually happens is the value of a SHOP variable can be substituted into an expression that is interpreted as lisp code.

For example from a board game playing program:

(:- (distance ?location1 ?location2 ?distance)
    ((hex-location ?location1 ?hex1)
     (hex-location ?location2 ?hex2)
     (assign ?distance (hex-distance ?hex1 ?hex2)))

Here we have used collected the coordinate values ?hex1 and ?hex2 and then they are passed into a lisp function, hex-distance.

Note an obscure issue here -- if these coordinates were, for some reason, lisp symbols, like 'A12 and 'B14(instead of, say 0112 and 0214) then the above would cause an error, because the values would be plugged in, and then hex-distance would be invoked, and A12 and B14 would look like unbound variables to Lisp.

In cases like that, you would have to modify the above to use:

     (assign ?distance (hex-distance (quote ?hex1) (quote ?hex2))))

[ko56]

Thanks, I'm already doing things like the example you gave. My question was not properly asked.
What I meant to ask is if there's a way to access the atoms without passing them as arguments to the Lisp function.
E.g. I have a predicate "greater", and I have defined/set
(greater 2 1)
(greater b a)
...
These atoms are part of the current state, so maybe that way, somehow?

[rpgoldman]

I'm still not sure I understand what you mean by a SHOP variable. SHOP variables are internal to the theorem-prover, and so are only accessible through mechanisms like the one I described earlier.

If you want to access atoms from Lisp code, you can use the special *current-state* variable. See section 3.4 in the manual.

[ko56]

I was using the term "variable" very loosely. I should have said "atom".
I know about current-state, but I never paid attention to its structure.
I will do so now.