Reputation: 189
I haven't found how I can get the first/second element of a tuple in Prolog, I tried arg, but it only works on predicates. My list looks like this:
Pairs = [
(pos(1, 2), pos(1, 3)),
(pos(1, 2), pos(2, 1)),
(pos(1, 2), pos(2, 3)),
(pos(1, 2), pos(3, 1)),
(pos(1, 2), pos(3, 2)),
(pos(1, 3), pos(2, 1)),
(pos(1, 3), pos(2, 3)),
(pos(..., ...), pos(..., ...)),
(..., ...)|...].
Upvotes: 3
Views: 3110
Reputation: 18950
You don't need any special predicate to do that, just unification, e.g.:
?- (A, B) = (pos(1, 2), pos(4, 3)).
A = pos(1, 2),
B = pos(4, 3).
?- (pos(X1, Y1), pos(X2, Y2)) = (pos(1, 2), pos(4, 3)).
X1 = 1,
Y1 = 2,
X2 = 4,
Y2 = 3.
Unification can happen in the body of a rule, or directly in the head of the rule. Consider the following example "visiting" all pairs:
visit_pairs([]). % same as: % visit_pairs([]).
visit_pairs([(A, B) | Xs]) :- % --> % visit_pairs([X|Xs]) :-
writeln(first_pair(A)), % % (A, B) = X,
writeln(second_pair(B)), % % writeln(first_pair(A)),
visit_pairs(Xs). % % writeln(second_pair(B)),
% % visit_pairs(Xs).
visit_pairs
will work only with lists of tuples, and each element of the list must unify with (A, B)
.
?- visit_pairs([(pos(1, 2), pos(1, 3)), (pos(1, 2), pos(2, 1)), (pos(1, 2), pos(2, 3)), (pos(1, 2), pos(3, 1)), (pos(1, 2), pos(3, 2)), (pos(1, 3), pos(2, 1)), (pos(1, 3), pos(2, 3))]).
first_pair(pos(1,2))
second_pair(pos(1,3))
first_pair(pos(1,2))
second_pair(pos(2,1))
first_pair(pos(1,2))
second_pair(pos(2,3))
first_pair(pos(1,2))
second_pair(pos(3,1))
first_pair(pos(1,2))
second_pair(pos(3,2))
first_pair(pos(1,3))
second_pair(pos(2,1))
first_pair(pos(1,3))
second_pair(pos(2,3))
true.
Another way could be to use nth0/3
to access one specific element by index, and then again using unification to unpack the tuple:
?- Pairs = [(pos(1, 2), pos(1, 3)), ...], nth0(0, Pairs, Elem), (A, B) = Elem.
Elem = (pos(1, 2), pos(1, 3)),
A = pos(1, 2),
B = pos(1, 3).
but this looks rather atypical.
Upvotes: 4