Skip to content

Queries and Unification

Queries

  • A query is a goal (or a comma‑separated list of goals) ending with a full stop.
  • The REPL prints bindings for variables that appear in the query.
?- parent(tom, bob).
true.

?- parent(tom, X).
X = bob.
  • Multiple goals use , (conjunction). Goals run left‑to‑right, backtracking as needed.
?- parent(tom, X), parent(X, Z).
X = bob, Z = ann.
  • Alternatives use ; (disjunction). Parenthesize when mixing with ,.
?- (X = a ; X = b), Y = t.
X = a, Y = t ;
X = b, Y = t.

Unification (=)

Unification tries to make two terms equal by binding variables.

?- X = a.
X = a.

?- point(X, 2) = point(1, Y).
X = 1, Y = 2.

?- a = b.
false.

Negated unification (\=) succeeds only if the terms cannot unify:

?- a \\= b.
true.
?- X \\= a.
false.

Structural vs arithmetic comparisons

  • Structural equality/inequality (no unification): ==/2, \==/2
?- X == Y.
false.  % different unbound variables are not identical
?- a == a.
true.
  • Arithmetic evaluation and comparison: use is/2 to evaluate, =:=/2 and =\=/2 to compare values.
?- X is 1+2.
X = 3.
?- 1+2 =:= 3.
true.
?- 1+2 = 3.
false.  % = is unification, not arithmetic

Multiple solutions and backtracking

Backtracking explores alternatives to find more solutions.

color(red). color(green). color(blue).

?- color(C).
C = red ;
C = green ;
C = blue ;
false.

Anonymous variables

  • _ is the anonymous variable; each occurrence is fresh and never shares a binding.
  • Use _ for “don’t care” positions to avoid cluttering output.
?- parent(tom, _).
true.

Occurs check

  • Like most Prologs, PyLog does not perform the occurs check by default. Queries such as X = f(X) will succeed and create a cyclic term. Prefer acyclic data where possible.

See also

  • Basics → Operators: for ,, ;, ->, and equality/compare operators
  • Basics → Arithmetic: is/2, =:=/2, =\=/2, <, =<, >, >=