varlist : correctly implemented ?

[victor]

Hello,

I am wondering if varlist has not a problem for predicates defined in list and using unification :

varlist::select(A, [A, C, D], R).
R = [C, D] ;
A = C,
R = [C, D] ;
A = D,
R = [D, C] ;
false.

I expected this behavior :

varlist::select(A, [A, C, D], R).
R = [C, D] ;
false.

Maybe a solution would be to rewrite completely varlist implementing listp ...

I can do it if you think it is a good idea.

[victor]

I forgot to say :

maybe I misunderstand varlist, but shouldn't it be named strictlist (like in "strict equality") instead of varlist ?

I personally use memberchk from it because it does not unify ...

[victor]

Or maybe keep the extension of list, but reimplementing more predicates that needs it ... since there is a lot of them that does not need a rewrite.

[pmoura]

Corrected for the predicate select/3 in r4402. There might be other predicates that should have local definitions instead of using the ones inherited from "list". Let me know. Not sure about the name change.