tpop(3.5): Markov Chain Algorithm-Java

Class 1: Prefix.java

import java.util.Vector;

public class Prefix {
    static final int MULTIPLIER = 31; // for hash code
    public Vector pref; //NPREF adjacent words from input

    public Prefix(int n, String str) {
        pref = new Vector();
        for (int i = 0; i < n ; i++)
            pref.addElement(str);

    }
    public Prefix (Prefix p) {
        pref  = (Vector)p.pref.clone();
    }

    public int hashCode() {
        int h = 0;
        for(int i = 0; i < pref.size(); i++)
            h = MULTIPLIER * h + pref.elementAt(i).hashCode();
        return h;
    }

    public boolean equals(Object o) {
        Prefix p = (Prefix) o;
        for (int i = 0; i < pref.size(); i++)
            if(!pref.elementAt(i).equals(p.pref.elementAt(i)))
                return false;
            return true;
    }
}

Class 2: Chain.java

import java.io.BufferedReader;
import java.io.FileReader;
import java.io.IOException;
import java.io.StreamTokenizer;
import java.util.Hashtable;
import java.util.Iterator;
import java.util.Random;
import java.util.Vector;
public class Chain {
    static final int NPREF = 2; //size of prefix
    static final String NONWORD = "#"; //"word" that can't appear
    Hashtable states = new Hashtable(); //key = Prefix, value = suffix Vector

    Prefix prefix = new Prefix(NPREF, NONWORD); //initial prefix @1.1 : [#,#]
    Random rand = new Random();

    public void build(FileReader file) throws IOException {//@2
        StreamTokenizer st = new StreamTokenizer(new BufferedReader(file));
        st.resetSyntax();
        st.wordChars(0, Character.MAX_VALUE);
        st.whitespaceChars(0, ' ');
        try{
            while (st.nextToken() != st.TT_EOF)
                add(st.sval);//@2.1:
            add(NONWORD);//@2.2
        } catch (IOException e){
            System.err.println("st.nextToken() unsucessful");
            throw e;
        }
    }
    /* Chain add: add word to suffix list, update prefix */
     void add(String word)    { //@2.1
        Vector suffix = (Vector) states.get(prefix);
        if (suffix == null) {
            suffix = new Vector();
            states.put(new Prefix(prefix), suffix);//first time: [#,#]
        }
        suffix.addElement(word);
        prefix.pref.removeElementAt(0);
        prefix.pref.addElement(word);
    }


// Chain generate: generate output words
    void generate(int nwords){//@3
        prefix = new Prefix(NPREF, NONWORD);//initial prfix: [#,#]
        for (int i = 0; i < nwords; i++) {
            Vector suffix = (Vector) states.get(prefix);
            if (suffix == null) {
                System.err.println("Markov: internal error: no state");
                System.exit(1);
            }
            int r = Math.abs(rand.nextInt()) % suffix.size();
            String str = (String) suffix.elementAt(r);
            if (str.equals(NONWORD))
                break;
            System.out.println("The prefix "+prefix.pref.toString() + "'s suffix is <" + str+">");
            prefix.pref.removeElementAt(0);
            prefix.pref.addElement(str);
            }
            Iterator it = states.keySet().iterator();
            while(it.hasNext()) {
                Prefix key = (Prefix)it.next();
                System.out.println("The sates are ["+key.pref.toString() +
                                   "]"+states.get(key).toString() );

        }
    }

}

Class 3:Markov.java

import java.io.FileReader;
import java.io.IOException;

import java.util.Random;

public class Markov {
    static final int MAXGEN = 10000; //maxium words generated
    private FileReader file;

    public static void main(String[] args) throws IOException{
        Chain chain = new Chain();//@1
        int nwords = MAXGEN;
        //Markov markov = new Markov(args[0]);
        FileReader file = new FileReader(args[0]);
        chain.build(file);//@2
        chain.generate(nwords);//@3
    }
  }

Input file: input.txt

Here we are here we go

output:

The prefix [#, #]'s suffix is
The prefix [#, here]'s suffix is <we>
The prefix [here, we]'s suffix is
The prefix [we, are]'s suffix is <here>
The prefix [are, here]'s suffix is
The prefix [here, we]'s suffix is <go>
The sates are [[#, #]][here]
The sates are [[we, are]][here]
The sates are [[here, we]][are, go]
The sates are [[we, go]][#]
The sates are [[#, here]][we]
The sates are [[are, here]][we]