HashSet usage with int arrays

Question

As I have an ArrayList of int arrays which contains duplicates, I'd like to use HashSet. Unfortunately, I can't manage to use HashSet as I wish:

System.out.print("\nTESTs\n");
    ArrayList<int[]> list = new ArrayList<int[]>();
    list.add(new int[]{1,2,3});
    list.add(new int[]{5,1,1});
    list.add(new int[]{1,2,3});//duplicate
    list.add(new int[]{5,1,3});

    Set<int[]> set = new HashSet<int[]>(list);
    System.out.println("Size of the set = "+set.size());

    ArrayList<int[]> arrayList = new ArrayList<int[]>(set);
    System.out.println("Size of the arrayList = "+arrayList.size());

    for (int[] array:arrayList){
        System.out.println(Arrays.toString(array));
    }

It results in:

Size of the set = 4
Size of the arrayList = 4
[1, 2, 3]
[1, 2, 3] // duplicate still here
[5, 1, 1]
[5, 1, 3]

Could anybody tell me where I'm wrong ?

Thanks in advance Dominique (java newbie)

Answer 1

Arrays don't override hashCode and equals implemented in Object class, and therefore, two arrays a1 and a2 will be considered as identical to each other by HashSet only if a1==a2, which is false in your case.

If you use ArrayLists instead of arrays, your problem will be solved, since for ArrayLists equality is determined by the equality of the members of the lists (and the order in which they appear).

Answer 2

That's because HashSet uses .equals() to see if a new object is duplicated (and .hashCode() to determine the "bucket").

When you work with arrays, please be aware that new int[]{1,2,3} is NOT "equal to" new int[]{1,2,3}.

The proper way to "deep compare" arrays is via Arrays.equals(a, b) method.

To solve your problem situation effectively you should create a wrapper class which contains your int[] array and then implement .hashCode() and equals() properly.