How do I extract data from a Java object?

Question

I am C programmer and now I moved to Java. I am trying to convert C program in Java program. C programs simply calculate term frequency and inverse document frequency (tf/idf).

I created one data class

public class Data {
    private String fileName,fileText;
    private int fileId;
    private float value;

    public void addData(String fileName, String fileText, float value){
        this.fileName = fileName;
        this.fileText = fileText;
        this.value = value;
    }

    public int getFileId(){
        return this.fileId;
    }


    public String getFileName(){
        return this.fileName;
    }

    public String getFileText(){
        return this.fileText;
    }

    public float getValue(){
        return this.value;
    }
}

This class is responsible to store file name, file text, and Value (tf value or idf value).

The following class is responsible to store data:

public class main {
    public static void main(String[] args) {

        HashMap<String, Data> map = new HashMap<String, Data>();
        Data dt = new Data();

        dt.addData("abc.txt", "some contents", 2);
        map.put("1",dt);
        dt.addData("w", "some more contents in second file", 3);
        map.put("2",dt);

        System.out.println(map);

    }

}

When I print map, it gives me some weird values. I think, I have to declare array of data class? I don't know how many files are there, therefore I can not put any static array number.

Also, how can I calculate TF and IDF based on this data structure?

In a C program, I simply read files, count the words divide by total number of words to get TF and a word divided by total occurrence of that word in all files to get IDF. I do not know how to do it using above data structure.

I get weird values. Maybe these are objects:

{2=test2.Data@19821f, 1=test2.Data@19821f}

Is there any way to get a specific value from Data class using getFileName etc. functions?

Answer 1

For question one, unless you override toString(), you are unlikely to get any meaningful output just by printing objects directly to stdout. The 'test2.Data@19821f' is what Object.toString() returns - class name followed by object hash. In this case, it quite helpfully shows that both your values are the same objects.

You can open/read files using java.io.File and java.io.FileInputStream. A map from strings to integers java.util.Map<String,Integer> will probably help with counting words in those files.

There doesn't seem to be much need for your data class for this simple application. You've already described the algorithm to follow, it's just a case of writing it in Java syntax.

Welcome to the wonderful world of type safety and not worrying about memory leaks.

Answer 2

You're creating only one instance of Data. You probably want to do something more like:

    Data dt = new Data();
    dt.addData("abc.txt", "some contents", 2);
    map.put("1",dt);

    dt = new Data();
    dt.addData("w", "some more contents in second file", 3);
    map.put("2",dt);

Or better yet, change Data to take the properties in its constructor:

    map.put("1", new Data("abc.txt", "some contents", 2));
    map.put("2", new Data("w", "some more contents in second file", 3));

Answer 3

It's not clear what your question is (see comments below your question), but there's a few things wrong with your code. addData is a misleading name for a method that replaces the data in the object. But the real problem is here:

  dt.addData("abc.txt", "some contents", 2);
  map.put("1",dt);
  dt.addData("w", "some more contents in second file", 3);
  map.put("2",dt);

This results in a map containing two entries, both of which refer to the same Data object, which will contain the values from the last call to addData. Change addData to be a constructor:

public Data(String fileName, String fileText, float value) {

Then change your map code to this:

map.put("1", new Data("abc.txt", "some contents", 2));
map.put("2", new Data("w", "some more contents in second file", 3));

Answer 4

Most likely you are expecting to see the Data string representation.

When you invoke println over any object ( including the map ) the system invokes Object.toString()

In the case of the map, the toString method returns the content of the map, in a format similar to this:

{ key = value, key2, value2 }

That is, print the key,value pairs it has.

Now, the key and value are objects too, so their own toString() method is invoked. For a string the value is it self. But, in the case of Data, since you haven't supplied your own implementation, you'll get default's which is object fully qualified name @ object.hashCode() So you are probably getting something like:

 { 1 = Data@0xa6f2be, 2 = Data@0xa6f2be }

To change this you have to override the toString() method:

 class Data { 
  ... etc. etc. 
     public String toString() { 
       // return something meaningful like:
       return String.format( "Data( fileName = %s, fileText = %s, etc ", this.fileName, this.fileText );
     }
  }

As for the second question, you'll do it basically the same way as you would in C. Perhaps you should create a calculate() method that opens the file, and start the counting. Probably this deserve its own question.