CYK algorithm implementation java

Question

I'm trying to implement the CYK algorithm based on wikipedia pseudocode. When I test the string "a b" for the grammar input:

S->A B

A->a

B->b

It gives me false, and I think it should be true. I have an arraylist called AllGrammar that contains all the rules. For the example above it would contain:

[0]: S->A B
[1]: A->a
[2]: B->b

For the example S->hello and the input string hello it gives me true as it should. More complex tests (more productions) gives me false :S

public static boolean cyk(String entrada) {
    int n = entrada.length();
    int r = AllGrammar.size();
    //Vector<String> startingsymbols = getSymbols(AllGrammar);
    String[] ent = entrada.split("\\s");
    n = ent.length;
    System.out.println("length of entry" + n);
    //let P[n,n,r] be an array of booleans. Initialize all elements of P to false.
    boolean P[][][] = initialize3DVector(n, r);
    //n-> number of words of string entrada, 
    //r-> number of nonterminal symbols

    //This grammar contains the subset Rs which is the set of start symbols
    for (int i = 1; i < n; i++) {
        for(int j = 0; j < r; j++) {
            String[] rule = (String[]) AllGrammar.get(j);
            if (rule.length == 2) {
                if (rule[1].equals(ent[i])) {
                    System.out.println("entrou");
                    System.out.println(rule[1]);
                    P[i][1][j + 1] = true;
                }
            }
        }
    }
    for(int i = 2; i < n; i++) {
        System.out.println("FIRST:" + i);

        for(int j = 1; j < n - i + 1; j++) {
            System.out.println("SECOND:" + j);

            for(int k = 1; k < i - 1; k++) {
                System.out.println("THIRD:" + k);
                for(int g = 0; g < r; g++) {
                    String[] rule = (String[]) AllGrammar.get(g);
                    if (rule.length > 2) {
                        int A = returnPos(rule[0]);
                        int B = returnPos(rule[1]);
                        int C = returnPos(rule[2]);
                        System.out.println("A" + A);
                        System.out.println("B" + B);
                        System.out.println("C" + C);
                        if (A!=-1 && B!=-1 && C!=-1) {
                            if (P[j][k][B] && P[j + k][i - k][C]) {
                                System.out.println("entrou2");
                                P[j][i][A] = true;
                            }
                        }
                    }
                }
            }
        }
    }

    for(int x = 0; x < r; x++) {
        if(P[1][n][x]) return true;
    }

    return false;
}

AllGrammar is an ArrayList and represents all the productions — user2466704, Apr 06 '14 at 16:39

score 0 · Answer 1 · answered Apr 07 '14 at 14:04

0

As compared to the CYK algorithm:

you have indexing starting at 1, but the arrays would appear to start at 0
the function returnpos() is not defined, and it's not obvious what it does.

It would seem the problems could be fairly basic in the the use of indexes. If you are new to the language, you might want to get a refresher.

answered Apr 07 '14 at 14:04

david.pfx

10,520
3
30
63

I'm not new to the language :p the function returnpos returns the position of the symbol in AllGrammar. For example A->BC, if I want the position of b the function will check for any B->XXXXX and return the position. The indexes start at 1, and the grammar at 0, but there is no difference because I'm incrementing the j in P[i][1][j + 1] = true; which should be P[i][1][j] = true; if the grammar started at 1. And in function returnpos im returning +1, as if I started at 1. – user2466704 Apr 07 '14 at 21:54
Then why do the loops run from 1 to – david.pfx Apr 07 '14 at 22:50

CYK algorithm implementation java

1 Answers1