Here is a simple Prolog program to exhibit a problem that I currently have with my real code.
I am trying to write a Brainf*ck interpreter in Prolog, but it is not returning the right Output
I am running bf("+.", Output)
Expected Return – Output = [1]
Actual Return – Output = [].
Output looses its value when it returns to iterateProg(Prog,Output) in bf2.
Program is as follows:
bf2(Prog,Output):-
iterateProg(Prog,Output).
iterateProg([],_):- !.
iterateProg(Prog, Output):-
checkInstruction(Prog,Output, NewProg, NewOutput),
iterateProg(NewProg, NewOutput).
checkInstruction([Plus|ProgTail],Output,ProgTail,Output):-
char_code('+',Plus)
%increase memory by 1 for printing later.
.
checkInstruction([Period|ProgTail], Output, ProgTail, NewOutput):-
char_code('.',Period),
%take the value in memory from addition just now, we will assume it is 1
append(Output,[1],NewOutput).
Your predicate
bf2(Prog, Output)has the meaning of: this programProgcorrespond with this outputOutput– so theOutputin thebf2predicate and also initerateProgshould contains the full output.So in
checkInstructionpredicate, the first 2 arguments should contain the full program and the full output, and the last 2 arguments should contain the partial program and the corresponding partial output.Therefore, you should append the
[1]withNewOutput, to produce the final output inOutputfor the second rule ofcheckInstructionpredicate. You should rename theNew...with a different name – since it is actually the partial/”tail” program and corresponding output.Note the order (I was inaccurate in the comment).
[1]is the output for the current instruction, and it should precede the output of the instruction inProgTail.By the way,
bf2predicate is unnecessary – since it calliterateProgpredicate with exact same arguments.Another thing is
iterateProg([],_)should beiterateProg([],[])since an empty program should correspond with empty output. (The converse is not true, so you should never put a cut!here, if you have learned about it).