Is Java TreeSet and HashSet expected to give different result with the same data?

Question

The Following program prints one|two one|two when HashSet is used. If I change it to TreeSet it prints only one|two

Is it not expected to give the same result with the same set of data ?What am I doing wrong here? What in the program makes the second record duplicate in case of TreeSet?

import java.util.HashSet; import java.util.Set;

public class SetTest {

public static void main(String arg[]) {
    Set<Bk> bookList = new HashSet<Bk>();
            //Set<Bk> bookList = new TreeSet<Bk>();
    bookList.add(new Bk("one","two"));
    bookList.add(new Bk("one","two"));
    for(Bk book: bookList){
        System.out.println(book);
    }
}
}

class Bk implements Comparable<Bk> {


public String name;
public String author;

public String toString(){
    return name+"|"+ author;
}

Bk(String name, String author) {
    this.name = name;
    this.author = author;

}

@Override
public int compareTo(Bk that) {

    int author = this.author.compareTo(that.author);
    int name = 0;
    if (author == 0) {
        name = this.name.compareTo(that.name);
        return name;
    } else {
        return author;
    }

}

}

Answer 1

The HashSet class will use hashCode and equals to determine if there is a duplicate already in the set. The TreeSet class will use the fact that it's Comparable (or alternatively, a Comparator) to order items and determine if there's a duplicate.

You have Bk as Comparable<Bk>, so TreeSet will work properly with it. But, you still need to override hashCode and equals so HashSet will work properly with it also.

Answer 2

If you want consistent behaviour with a HashSet you would have to implement hashCode() and equals(), as per the Javadoc, and in such a way as to be consistent with your compareTo() method.