[ Comparing and Sorting | The ECLiPSe Built-In Predicates | Reference Manual | Alphabetic Index ]

compare(-Ordering, ?Term1, ?Term2)

Succeeds if Ordering is a special atom which describes the ordering between Term1 and Term2.
Unifiable to a special atom describing the ordering between Term1 and Term2.
An arbitrary term.
An arbitrary term.


Succeeds if Ordering is one of the special atoms ('<', '>' or '=') describing the standard ordering between the terms Term1 and Term2:

Ordering is the atom '<' iff Term1 comes before Term2 in the standard ordering.

Ordering is the atom '>' iff Term1 comes after Term2 in the standard ordering.

Ordering is the atom '=' iff Term1 is identical to Term2.

The standard ordering of ECLiPSe terms is defined as the following increasing order:

(comparing two free variables yields an implementation-dependent and not necessarily reproducible result).
bounded reals
in ascending order (if bounds overlap, the order is by increasing lower bounds, then increasing upper bounds; if both bounds are the same, the two terms are considered equal).
in ascending order, with negative zeros (-0.0) being different and before positive zeros (0.0).
in ascending order.
in ascending order.
lexicographical (ASCII) order
lexicographical (ASCII) order
compound terms
first by arity, then by functor name, then by the arguments in left to right order.
Note in particular that numbers are first ordered by their type (integer, float, etc) and only then by their magnitude, i.e. when comparing numbers of different types, the result is not necessarily their numerical order.

Modes and Determinism


   compare(X, A, a), X = '<'.
   compare(X, a, A), X = '>'.
   compare('<', a(1,2), b(1,2)).
   compare(X, 1, 1), X = '='.
   compare(X, f(1), f(1)), X = '='.
   compare('<', 3.0, 2).              % not arithmetic order!
   compare('>', [a,b], [a|b]).
   compare('>', [a,b], [a|X]).
   compare('<', atomb, atoma).
   compare('=', 0, 1).

See Also

@> / 2, @< / 2, @=< / 2, @>= / 2