Concept for adversarial-search domains that can estimate a child score without modifying state (incremental evaluator for Alpha_Beta Estimated_Score move ordering). More...

#include <state_search_common.H>

Concept definition

template<typename Domain>
concept Aleph::IncrementalEvaluator = 
  requires(Domain &domain,
           const typename Domain::State &state,
           const typename Domain::Move  &move) {
    { domain.evaluate_after(state, move) }
      -> std::convertible_to<typename Domain::Score>;
  }

Detailed Description

Concept for adversarial-search domains that can estimate a child score without modifying state (incremental evaluator for Alpha_Beta Estimated_Score move ordering).

When a domain satisfies this concept, Aleph::Alpha_Beta and any other engine that supports Estimated_Score move ordering can compute shallow evaluation priorities for move ordering without calling apply()/undo() on each candidate move.

The expression domain.evaluate_after(state, move) must return a value convertible to Domain::Score and must represent a cheap heuristic estimate of the score of the child state obtained by applying move, without altering state.

Template Parameters

Domain Adversarial game domain type.

Definition at line 216 of file state_search_common.H.