Question

Given a list of objects (all of the same type), how can I make sure that it contains only one element for each value of a certain attribute, even though equals() may return false for such elements due to more attributes being checked? In code:

private void example() {
    List<SomeType> listWithDuplicates = new ArrayList<SomeType>();

    /*
     * create the "duplicate" objects. Note that both attributes passed to 
     * the constructor are used in equals(), though for the purpose of this 
     * question they are considered equal if the first argument was equal
     */
    SomeType someObject1 = new SomeObject1("hello", "1");
    SomeType someObject2 = new SomeObject1("hello", "2");

    List<SomeType> listWithoutDuplicates = removeDuplicates(listWithDuplicates)
    //listWithoutDuplicates should not contain someObject2
}

private List<SomeType> removeDuplicates(List<SomeType> listWithDuplicates) {
    /*
     * remove all but the first entry in the list where the first constructor-
     * arg was the same
     */
}

Answer 1

Could use a Set as an intermediary placeholder to find the duplicates as Bozho suggested. Here’s a sample removeDuplicates() implementation.

private List<SomeType> removeDuplicates(List<SomeType> listWithDuplicates) {
    /* Set of all attributes seen so far */
    Set<AttributeType> attributes = new HashSet<AttributeType>();
    /* All confirmed duplicates go in here */
    List duplicates = new ArrayList<SomeType>();

    for(SomeType x : listWithDuplicates) {
        if(attributes.contains(x.firstAttribute())) {
            duplicates.add(x);
        }
        attributes.add(x.firstAttribute());
    }
    /* Clean list without any dups */
    return listWithDuplicates.removeAll(duplicates);
}