built-in method

coinductive_success_hook/1-2

Description

coinductive_success_hook(Head, Hypothesis)
coinductive_success_hook(Head)

User-defined hook predicates that are automatically called in case of coinductive success when proving a query for a coinductive predicates. The hook predicates are called with the head of the coinductive predicate on coinductive success and, optionally, with the hypothesis that has used to reach coinductive success.

When both hook predicates are defined, the coinductive_success_hook/1 clauses are only used if no coinductive_success_hook/2 clause applies. The compiler ensures zero performance penalties when defining coinductive predicates without a corresponding definition for the coinductive success hook predicates.

The compiler assumes that these hook predicates are defined as static predicates in order to optimize their use.

Modes and number of proofs

coinductive_success_hook(+callable, +callable) - zero_or_one
coinductive_success_hook(+callable) - zero_or_one

Errors

(none)

Examples

% Are there "occurrences" of arg1 in arg2?
:- public(member/2).
:- coinductive(member/2).

member(X, [X| _]).
member(X, [_| T]) :-
    member(X, T).

% Are there infinitely many "occurrences" of arg1 in arg2?
:- public(comember/2).
:- coinductive(comember/2).
comember(X, [_| T]) :-
    comember(X, T).

coinductive_success_hook(member(_, _)) :-
    fail.
coinductive_success_hook(comember(X, L)) :-
    member(X, L).

See also

coinductive/1