🚀 go-pugleaf

The following will work for JPI which supports type
transfer functions (you may have to use VAL in other M2's):

PROCEDURE And(a,b:CARDINAL): CARDINAL;
BEGIN
  RETURN CARDINAL(BITSET(a) * BITSET(b))
END And;

PROCEDURE Or(a,b:CARDINAL): CARDINAL;
BEGIN
  RETURN CARDINAL(BITSET(a) + BITSET(b))
END Or;

PROCEDURE Not(a:CARDINAL): CARDINAL;
BEGIN
  RETURN CARDINAL(BITSET(0) - BITSET(a))
END Not;

PROCEDURE Xor(a,b:CARDINAL): CARDINAL;
BEGIN
  RETURN And(Or(a,b), Not(And(a,b)))
END Xor;


You can, of course, optimize Xor by 'inlining' by hand
the other ops.

This solution requires no recourse to assembly language.


-- Larry