制約処理ルールによる無限再帰の回避

Question

Constraint Handling Rulesを使用して、SWI-Prologに簡単な制約条件を記述しました。私はmeans([3,is,equal,to,4],[3,equals,4])はtrueことが期待が、それは代わりに、無限再帰を引き起こしているようだ

%If A means B, then B means A. 
means(A,B) ==> means(B,A).  
%If A means B and A means C, then B means C. 
means(A,B),means(A,C) ==> means(B,C). 

：それは推論の2つの比較的簡単なルールを使用しています

:- use_module(library(chr)). 
:- chr_constraint means/2. 
:- initialization(main). 


means([A,equals,B],[A,'=',B]). 
means([A,is,equal,to,B],[A,'=',B]). 
means([A,equals,B],[A,and,B,are,equal]). 


%These are the rules of inference for this program. 
    %If A means B, then B means A. 
    means(A,B) ==> means(B,A).  
    %If A means B and A means C, then B means C. 
    means(A,B),means(A,C) ==> means(B,C). 

main :- 
    %This part works as expected. X = [3,'=',4]. 
    means([3,is,equal,to,4],X),writeln(X), 

    %This statement should be true, so why does it produce an infinite recursion? 
    means([3,is,equal,to,4],[3,and,4,are,equal]). 

私はこのプログラムにsimpagationルールを追加しましたが、それまだOut of local stackエラーにつながる：推論規則を再書き込みする

:- use_module(library(chr)). 
:- chr_constraint means/2. 
:- initialization(main). 


%These are the rules of inference for this program. 
    %If A means B, then B means A. 
    means(A,B) ==> means(B,A).  
    %If A means B and A means C, then B means C. 
    means(A,B),means(A,C) ==> means(B,C). 
    means(A,B) \ means(A,B) <=> true. 
    means(A,A) <=> true. 

means([A,equals,B],[A,'=',B]). 
means([A,is,equal,to,B],[A,'=',B]). 
means([A,equals,B],[A,and,B,are,equal]). 

main :- 
    %This part works as expected. X = [3,'=',4]. 
    means([3,is,equal,to,4],X),writeln(X), 

    %This statement should be true, so why does it produce an infinite recursion? 
    means([3,is,equal,to,4],[3,and,4,are,equal]). 

ことが可能なように、彼ら無限再帰を生成しませんか？

Answer 1

  CHRのこのような側面の詳細については、入手可能なCHR文献をお読みください。期待どおりCHRのsimpagationルールを追加する場合はこのように、あなたの例では、作品

9. Set Semantics

The CHR system allows the presence of identical constraints, i.e. multiple constraints with the same functor, arity and arguments. For most constraint solvers, this is not desirable: it affects efficiency and possibly termination. Hence appropriate simpagation rules should be added of the form: constraint \ constraint <=> true

：

例えば、Tips for CHR programmingはプログラミングのヒントに含まれてい

means(A,B) \ means(A,B) <=> true.

サンプルクエリ結果：

 
?- means([3,is,equal,to,4],[3,and,4,are,equal]). 
means([3, and, 4, are, equal], [3, is, equal, to, 4]), 
means([3, is, equal, to, 4], [3, and, 4, are, equal]).