Iterating over three lists simultaneously [duplicate]

Question

The following code is using indexed for loop:

List<A> a = ..;
List<B> b = ..;
List<C> c = ..;

// a.size() == b.size() == c.size() - fundamental business logic assumption

int size = a.size();
for (int i = 0; i < size; i++) {
  doSthWith(a.get(i));
  doSthWith(b.get(i));
  doSthWith(c.get(i));
}

Is this possible to further simplify the code and not use the index to iterate over three lists simultaneously?

Answer 1

Since you are referring to the list by its interface, you should always prefer using an Iterator as opposed to get(), because you do not know whether the latter will be equally efficient.

Iterator<A> aIterator = aList.iterator();
Iterator<B> bIterator = bList.iterator();
Iterator<C> cIterator = cList.iterator();

while (aIterator.hasNext()) {
    assert bIterator.hasNext();
    assert cIterator.hasNext();
    A a = aIterator.next();
    B b = bIterator.next();
    C c = cIterator.next();
}
assert !bIterator.hasNext();
assert !cIterator.hasNext();

You could create your own Iterator which encapsulates the above logic:

for (Three<A, B, C> three : ThreeIterator.iterate(aList, bList, cList)) {
    A a = three.getA();
    B b = three.getB();
    C c = three.getC();
}

public class ThreeIterator<A, B, C> implements Iterator<Three<A, B, C>> {
    private final Iterator<A> aIterator;
    private final Iterator<B> bIterator;
    private final Iterator<C> cIterator;

    public ThreeIterator(Collection<A> aCollection, Collection<B> bCollection, Collection<C> cCollection) {
        aIterator = aCollection.iterator();
        bIterator = bCollection.iterator();
        cIterator = cCollection.iterator();
    }

    @Override
    public boolean hasNext() {
        boolean result = aIterator.hasNext();
        if (result) {
            assert bIterator.hasNext();
            assert cIterator.hasNext();
        } else {
            assert !bIterator.hasNext();
            assert !cIterator.hasNext();
        }
        return result;
    }

    @Override
    public Three<A, B, C> next() {
        return new Three<>(aIterator.next(), bIterator.next(), cIterator.next());
    }

    @Override
    public void remove() {
        aIterator.remove();
        bIterator.remove();
        cIterator.remove();
    }

    public static <A, B, C> Iterable<Three<A, B, C>> iterate(List<A> aList, List<B> bList, List<C> cList) {
        return new Iterable<Three<A, B, C>>() {
            @Override
            public Iterator<Three<A, B, C>> iterator() {
                return new ThreeIterator<>(aList, bList, cList);
            }

            @Override
            public Spliterator<Three<A, B, C>> spliterator() {
                int size = aList.size();
                assert bList.size() == size;
                assert cList.size() == size;
                return Spliterators.spliterator(iterator(), size, 0);
            }
        };
    }
}

public class Three<A, B, C> {
    private final A a;
    private final B b;
    private final C c;

    public Three(A a, B b, C c) {
        this.a = a;
        this.b = b;
        this.c = c;
    }

    public A getA() {
        return a;
    }

    public B getB() {
        return b;
    }

    public C getC() {
        return c;
    }
}

Answer 2

If all three lists are to be iterated together as you expect, that means they are tightly related and qualify to be an object of some type. It involves a design simplification beforehand.

Class MyClass {
   A a;
   B b;
   C c;
}

Then I'll be having a List<MyClass> instead of three lists with related data. This will simplify your code in long run as well.

And, to answer your question, whatever you are using currently seems very naive but the most readable way of doing this if you are not taking the toil of making a class.