Collections

The collections framework in java.util provides a number of generic classes for sets of data with functionality that can’t be provided by regular arrays.

Collections framework contains interfaces for Collection<O>, with main sub-interfaces List<O> and Set<O>, and mapping collection Map<K,V>. Collections are the root interface and are being implemented by many other collection frameworks.

Removing items from a List within a loop

It is tricky to remove items from a list while within a loop, this is due to the fact that the index and length of the list gets changed.

Given the following list, here are some examples that will give an unexpected result and some that will give the correct result.

List<String> fruits = new ArrayList<String>();
fruits.add("Apple");
fruits.add("Banana");
fruits.add("Strawberry");

INCORRECT

Removing in iteration of for statement Skips “Banana”:

The code sample will only print Apple and Strawberry. Banana is skipped because it moves to index 0 once Apple is deleted, but at the same time i gets incremented to 1.

for (int i = 0; i < fruits.size(); i++) {
    System.out.println (fruits.get(i));
    if ("Apple".equals(fruits.get(i))) {
         fruits.remove(i);
    }
}

Removing in the enhanced for statement Throws Exception:

Because of iterating over collection and modifying it at the same time.

Throws: java.util.ConcurrentModificationException

for (String fruit : fruits) {
    System.out.println(fruit);
    if ("Apple".equals(fruit)) {
        fruits.remove(fruit);
    }
}

CORRECT

Removing in while loop using an Iterator

Iterator<String> fruitIterator = fruits.iterator();
while(fruitIterator.hasNext()) {
    String fruit = fruitIterator.next();
    System.out.println(fruit);
    if ("Apple".equals(fruit)) {
        fruitIterator.remove();
    }
}

The Iterator interface has a remove() method built in just for this case. However, this method is marked as “optional” in the documentation, and it might throw an UnsupportedOperationException.

Throws: UnsupportedOperationException - if the remove operation is not supported by this iterator

Therefore, it is advisable to check the documentation to make sure this operation is supported (in practice, unless the collection is an immutable one obtained through a 3rd party library or the use of one of the Collections.unmodifiable...() method, the operation is almost always supported).

While using an Iterator a ConcurrentModificationException is thrown when the modCount of the List is changed from when the Iterator was created. This could have happened in the same thread or in a multi-threaded application sharing the same list.

A modCount is an int variable which counts the number of times this list has been structurally modified. A structural change essentially means an add() or remove() operation being invoked on Collection object (changes made by Iterator are not counted). When the Iterator is created, it stores this modCount and on every iteration of the List checks if the current modCount is same as and when the Iterator was created. If there is a change in the modCount value it throws a ConcurrentModificationException.

Hence for the above-declared list, an operation like below will not throw any exception:

Iterator<String> fruitIterator = fruits.iterator();
fruits.set(0, "Watermelon");
while(fruitIterator.hasNext()){
    System.out.println(fruitIterator.next());
}

But adding a new element to the List after initializing an Iterator will throw a ConcurrentModificationException:

Iterator<String> fruitIterator = fruits.iterator();
fruits.add("Watermelon");
while(fruitIterator.hasNext()){
    System.out.println(fruitIterator.next());    //ConcurrentModificationException here
}

Iterating backwards

for (int i = (fruits.size() - 1); i >=0; i--) {
    System.out.println (fruits.get(i));
    if ("Apple".equals(fruits.get(i))) {
         fruits.remove(i);
    }
}

This does not skip anything. The downside of this approach is that the output is reverse. However, in most cases where you remove items that will not matter. You should never do this with LinkedList.

Iterating forward, adjusting the loop index

for (int i = 0; i < fruits.size(); i++) {
    System.out.println (fruits.get(i));
    if ("Apple".equals(fruits.get(i))) {
         fruits.remove(i);
         i--;
    }
}

This does not skip anything. When the ith element is removed from the List, the element originally positioned at index i+1 becomes the new ith element. Therefore, the loop can decrement i in order for the next iteration to process the next element, without skipping.

Using a “should-be-removed” list

ArrayList shouldBeRemoved = new ArrayList();
for (String str : currentArrayList) {
    if (condition) {
        shouldBeRemoved.add(str);
    }
}
currentArrayList.removeAll(shouldBeRemoved);

This solution enables the developer to check if the correct elements are removed in a cleaner way.

In Java 8 the following alternatives are possible. These are cleaner and more straight forward if the removing does not have to happen in a loop.

Filtering a Stream

A List can be streamed and filtered. A proper filter can be used to remove all undesired elements.

List<String> filteredList =
    fruits.stream().filter(p -> !"Apple".equals(p)).collect(Collectors.toList());

Note that unlike all the other examples here, this example produces a new List instance and keeps the original List unchanged.

Using removeIf

Saves the overhead of constructing a stream if all that is needed is to remove a set of items.

fruits.removeIf(p -> "Apple".equals(p));

Constructing collections from existing data

Standard Collections

Java Collections framework

A simple way to construct a List from individual data values is to use java.utils.Arrays method Arrays.asList:

List<String> data = Arrays.asList("ab", "bc", "cd", "ab", "bc", "cd");

All standard collection implementations provide constructors that take another collection as an argument adding all elements to the new collection at the time of construction:

List<String> list = new ArrayList<>(data); // will add data as is
Set<String> set1 = new HashSet<>(data); // will add data keeping only unique values
SortedSet<String> set2 = new TreeSet<>(data); // will add data keeping unique values and sorting
Set<String> set3 = new LinkedHashSet<>(data); // will add data keeping only unique values and preserving the original order

Google Guava Collections framework

Another great framework is Google Guava that is amazing utility class (providing convenience static methods) for construction of different types of standard collections Lists and Sets:

import com.google.common.collect.Lists;
 import com.google.common.collect.Sets;
 ...
 List<String> list1 = Lists.newArrayList("ab", "bc", "cd");
 List<String> list2 = Lists.newArrayList(data);
 Set<String> set4 = Sets.newHashSet(data);
 SortedSet<String> set5 = Sets.newTreeSet("bc", "cd", "ab", "bc", "cd");

Mapping Collections

Java Collections framework

Similarly for maps, given a Map<String, Object> map a new map can be constructed with all elements as follows:

Map<String, Object> map1 = new HashMap<>(map);
SortedMap<String, Object> map2 = new TreeMap<>(map);

Apache Commons Collections framework

Using Apache Commons you can create Map using array in ArrayUtils.toMap as well as MapUtils.toMap:

import org.apache.commons.lang3.ArrayUtils;
 ...
 // Taken from org.apache.commons.lang.ArrayUtils#toMap JavaDoc

 // Create a Map mapping colors.
 Map colorMap = MapUtils.toMap(new String[][] {{
     {"RED", "#FF0000"},
     {"GREEN", "#00FF00"},
     {"BLUE", "#0000FF"}});

Each element of the array must be either a Map.Entry or an Array, containing at least two elements, where the first element is used as key and the second as value.

Google Guava Collections framework

Utility class from Google Guava framework is named Maps:

import com.google.common.collect.Maps;
 ...
 void howToCreateMapsMethod(Function<? super K,V> valueFunction,
           Iterable<K> keys1,
           Set<K> keys2,
           SortedSet<K> keys3) {
     ImmutableMap<K, V> map1 = toMap(keys1, valueFunction); // Immutable copy
     Map<K, V> map2 = asMap(keys2, valueFunction); // Live Map view
     SortedMap<K, V> map3 = toMap(keys3, valueFunction); // Live Map view
 }

Using Stream,

Stream.of("xyz", "abc").collect(Collectors.toList());

or

Arrays.stream("xyz", "abc").collect(Collectors.toList());

Declaring an ArrayList and adding objects

We can create an ArrayList (following the List interface):

List aListOfFruits = new ArrayList();

List<String> aListOfFruits = new ArrayList<String>();

List<String> aListOfFruits = new ArrayList<>();

Now, use the method add to add a String:

aListOfFruits.add("Melon");
aListOfFruits.add("Strawberry");

In the above example, the ArrayList will contain the String “Melon” at index 0 and the String “Strawberry” at index 1.

Also we can add multiple elements with addAll(Collection<? extends E> c) method

List<String> aListOfFruitsAndVeggies = new ArrayList<String>();
aListOfFruitsAndVeggies.add("Onion");
aListOfFruitsAndVeggies.addAll(aListOfFruits);

Now “Onion” is placed at 0 index in aListOfFruitsAndVeggies, “Melon” is at index 1 and “Strawberry” is at index 2.

Iterating over Collections

Iterating over List

List<String> names  = new ArrayList<>(Arrays.asList("Clementine", "Duran", "Mike"));

names.forEach(System.out::println);

If we need parallelism use

names.parallelStream().forEach(System.out::println);

for (String name : names) {
    System.out.println(name);
}

for (int i = 0; i < names.size(); i++) {
    System.out.println(names.get(i));
}

//Creates ListIterator which supports both forward as well as backward traversel
ListIterator<String> listIterator = names.listIterator();

//Iterates list in forward direction
while(listIterator.hasNext()){
    System.out.println(listIterator.next());
}

//Iterates list in backward direction once reaches the last element from above iterator in forward direction
while(listIterator.hasPrevious()){
    System.out.println(listIterator.previous());
}

Iterating over Set

Set<String> names = new HashSet<>(Arrays.asList("Clementine", "Duran", "Mike"));

names.forEach(System.out::println);

for (Iterator<String> iterator = names.iterator(); iterator.hasNext(); ) {
    System.out.println(iterator.next());
}

for (String name : names) {
    System.out.println(name);
}

Iterator iterator = names.iterator();
while (iterator.hasNext()) {
    System.out.println(iterator.next());
}

Iterating over Map

Map<Integer, String> names = new HashMap<>();
names.put(1, "Clementine");
names.put(2, "Duran");
names.put(3, "Mike");

names.forEach((key, value) -> System.out.println("Key: " + key + " Value: " + value));

for (Map.Entry<Integer, String> entry : names.entrySet()) {
    System.out.println(entry.getKey());
    System.out.println(entry.getValue());
}

// Iterating over only keys
for (Integer key : names.keySet()) {
    System.out.println(key);
}
// Iterating over only values
for (String value : names.values()) {
    System.out.println(value);
}

Iterator entries = names.entrySet().iterator();
while (entries.hasNext()) {
    Map.Entry entry = (Map.Entry) entries.next();
    System.out.println(entry.getKey());
    System.out.println(entry.getValue());
}

Immutable Empty Collections

Sometimes it is appropriate to use an immutable empty collection. The Collections class provides methods to get such collections in an efficient way:

List<String> anEmptyList = Collections.emptyList();
Map<Integer, Date> anEmptyMap = Collections.emptyMap();
Set<Number> anEmptySet = Collections.emptySet();

These methods are generic and will automatically convert the returned collection to the type it is assigned to. That is, an invocation of e.g. emptyList() can be assigned to any type of List and likewise for emptySet() and emptyMap().

The collections returned by these methods are immutable in that they will throw UnsupportedOperationException if you attempt to call methods which would change their contents (add, put, etc.). These collections are primarily useful as substitutes for empty method results or other default values, instead of using null or creating objects with new.

Unmodifiable Collection

Sometimes it’s not a good practice expose an internal collection since it can lead to a malicious code vulnerability due to it’s mutable characteristic. In order to provide “read-only” collections java provides its unmodifiable versions.

An unmodifiable collection is often a copy of a modifiable collection which guarantees that the collection itself cannot be altered. Attempts to modify it will result in an UnsupportedOperationException exception.

It is important to notice that objects which are present inside the collection can still be altered.

import java.util.ArrayList;
import java.util.Collections;
import java.util.List;

public class MyPojoClass {
    private List<Integer> intList = new ArrayList<>();

    public void addValueToIntList(Integer value){
        intList.add(value);
    }

    public List<Integer> getIntList() {
        return Collections.unmodifiableList(intList);
    }
}

The following attempt to modify an unmodifiable collection will throw an exception:

import java.util.List;

public class App {

    public static void main(String[] args) {
        MyPojoClass pojo = new MyPojoClass();
        pojo.addValueToIntList(42);

        List<Integer> list = pojo.getIntList();
        list.add(69);
    }
}

output:

Exception in thread "main" java.lang.UnsupportedOperationException
    at java.util.Collections$UnmodifiableCollection.add(Collections.java:1055)
    at App.main(App.java:12)

Pitfall: concurrent modification exceptions

This exception occurs when a collection is modified while iterating over it using methods other than those provided by the iterator object. For example, we have a list of hats and we want to remove all those that have ear flaps:

List<IHat> hats = new ArrayList<>();
hats.add(new Ushanka()); // that one has ear flaps
hats.add(new Fedora());
hats.add(new Sombrero());
for (IHat hat : hats) {
    if (hat.hasEarFlaps()) {
        hats.remove(hat);
    }
}

If we run this code, ConcurrentModificationException will be raised since the code modifies the collection while iterating it. The same exception may occur if one of the multiple threads working with the same list is trying to modify the collection while others iterate over it. Concurrent modification of collections in multiple threads is a natural thing, but should be treated with usual tools from the concurrent programming toolbox such as synchronization locks, special collections adopted for concurrent modification, modifying the cloned collection from initial etc.

Sub Collections

List subList(int fromIndex, int toIndex)

Here fromIndex is inclusive and toIndex is exclusive.

List list = new ArrayList();
List list1 = list.subList(fromIndex,toIndex);

1. If the list doesn’t exist in the give range, it throws IndexOutofBoundException.
2. What ever changes made on the list1 will impact the same changes in the list.This is called backed collections.
3. If the fromnIndex is greater than the toIndex (fromIndex > toIndex) it throws IllegalArgumentException.

Example:

List<String> list = new ArrayList<String>();
List<String> list = new ArrayList<String>();
list.add("Hello1");
list.add("Hello2");
System.out.println("Before Sublist "+list);
List<String> list2 = list.subList(0, 1);
list2.add("Hello3");
System.out.println("After sublist changes "+list);

Output:
Before Sublist [Hello1, Hello2]
After sublist changes [Hello1, Hello3, Hello2]

Set subSet(fromIndex,toIndex)

Here fromIndex is inclusive and toIndex is exclusive.

Set set = new TreeSet();
Set set1 = set.subSet(fromIndex,toIndex);

The returned set will throw an IllegalArgumentException on an attempt to insert an element outside its range.

Map subMap(fromKey,toKey)

fromKey is inclusive and toKey is exclusive

Map map = new TreeMap();
Map map1 = map.get(fromKey,toKey);

If fromKey is greater than toKey or if this map itself has a restricted range, and fromKey or toKey lies outside the bounds of the range then it throws IllegalArgumentException.

All the collections support backed collections means changes made on the sub collection will have same change on the main collection.

Join lists

Following ways can be used for joining lists without modifying source list(s).

First approach. Has more lines but easy to understand

List<String> newList = new ArrayList<String>();
newList.addAll(listOne);
newList.addAll(listTwo);

Second approach. Has one less line but less readable.

List<String> newList = new ArrayList<String>(listOne);
newList.addAll(listTwo);

Third approach. Requires third party Apache commons-collections library.

ListUtils.union(listOne,listTwo);

Using Streams the same can be achieved by

List<String> newList = Stream.concat(listOne.stream(), listTwo.stream()).collect(Collectors.toList());

References. Interface List

Removing matching items from Lists using Iterator.

Above I noticed an example to remove items from a List within a Loop and I thought of another example that may come in handy this time using the Iterator interface.
This is a demonstration of a trick that might come in handy when dealing with duplicate items in lists that you want to get rid of.

Note: This is only adding on to the Removing items from a List within a loop example:

So let’s define our lists as usual

   String[] names = {"James","Smith","Sonny","Huckle","Berry","Finn","Allan"};
    List<String> nameList = new ArrayList<>();

    //Create a List from an Array
    nameList.addAll(Arrays.asList(names));

    String[] removeNames = {"Sonny","Huckle","Berry"};
    List<String> removeNameList = new ArrayList<>();

    //Create a List from an Array
    removeNameList.addAll(Arrays.asList(removeNames));

The following method takes in two Collection objects and performs the magic of removing the elements in our removeNameList that match with elements in nameList.

private static void removeNames(Collection<String> collection1, Collection<String> collection2) {
      //get Iterator.
    Iterator<String> iterator = collection1.iterator();

    //Loop while collection has items
    while(iterator.hasNext()){
        if (collection2.contains(iterator.next()))
            iterator.remove(); //remove the current Name or Item
    }
}

Calling the method and passing in the nameList and the removeNameListas follows removeNames(nameList,removeNameList);
Will produce the following output:

Array List before removing names: **James Smith Sonny Huckle Berry Finn Allan**
Array List after removing names: **James Smith Finn Allan**

A simple neat use for Collections that may come in handy to remove repeating elements within lists.

Collections and Primitive Values

Collections in Java only work for objects. I.e. there is no Map<int, int> in Java. Instead, primitive values need to be boxed into objects, as in Map<Integer, Integer>. Java auto-boxing will enable transparent use of these collections:

Map<Integer, Integer> map = new HashMap<>();
map.put(1, 17); // Automatic boxing of int to Integer objects
int a = map.get(1); // Automatic unboxing.

Unfortunately, the overhead of this is substantial. A HashMap<Integer, Integer> will require about 72 bytes per entry (e.g. on 64-bit JVM with compressed pointers, and assuming integers larger than 256, and assuming 50% load of the map). Because the actual data is only 8 bytes, this yields a massive overhead. Furthermore, it requires two level of indirection (Map -> Entry -> Value) it is unnecessarily slow.

There exist several libraries with optimized collections for primitive data types (that require only ~16 bytes per entry at 50% load, i.e. 4x less memory, and one level of indirection less), that can yield substantial performance benefits when using large collections of primitive values in Java.

Creating your own Iterable structure for use with Iterator or for-each loop.

To ensure that our collection can be iterated using iterator or for-each loop, we have to take care of following steps:

The stuff we want to iterate upon has to be `Iterable` and expose `iterator()`.

1. Design a `java.util.Iterator` by overriding `hasNext()`, `next()` and `remove()`.

I have added a simple generic linked list implementation below that uses above entities to make the linked list iterable.

package org.algorithms.linkedlist;

import java.util.Iterator;
import java.util.NoSuchElementException;


public class LinkedList<T> implements Iterable<T> {

    Node<T> head, current;

    private static class Node<T> {
        T data;
        Node<T> next;

        Node(T data) {
            this.data = data;
        }
    }

    public LinkedList(T data) {
        head = new Node<>(data);
    }

    public Iterator<T> iterator() {
        return new LinkedListIterator();
    }

    private class LinkedListIterator implements Iterator<T> {

        Node<T> node = head;

        @Override
        public boolean hasNext() {
            return node != null;
        }

        @Override
        public T next() {
            if (!hasNext())
                throw new NoSuchElementException();
            Node<T> prevNode = node;
            node = node.next;
            return prevNode.data;
        }

        @Override
        public void remove() {
            throw new UnsupportedOperationException("Removal logic not implemented.");
        }
    }

    public void add(T data) {
        Node current = head;
        while (current.next != null)
            current = current.next;
        current.next = new Node<>(data);
    }

}

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

        LinkedList<Integer> list = new LinkedList<>(1);
        list.add(2);
        list.add(4);
        list.add(3);

        //Test #1
        System.out.println("using Iterator:");
        Iterator<Integer> itr = list.iterator();
        while (itr.hasNext()) {
            Integer i = itr.next();
            System.out.print(i + " ");
        }

        //Test #2
        System.out.println("\n\nusing for-each:");
        for (Integer data : list) {
            System.out.print(data + " ");
        }
    }
}

Output

using Iterator:
1 2 4 3
using for-each:
1 2 4 3

This will run in Java 7+. You can make it run on Java 5 and Java 6 also by substituting:

LinkedList<Integer> list = new LinkedList<>(1);

with

LinkedList<Integer> list = new LinkedList<Integer>(1);

or just any other version by incorporating the compatible changes.

Remarks

Collections are objects that can store collections of other objects inside of them. You can specify the type of data stored in a collection using Generics.

Collections generally use the java.util or java.util.concurrent namespaces.

Java 1.4.2 and below do not support generics. As such, you can not specify the type parameters that a collection contains. In addition to not having type safety, you must also use casts to get the correct type back from a collection.

In addition to Collection<E>, there are multiple major types of collections, some of which have subtypes.

• List<E> is an ordered collection of objects. It is similar to an array, but does not define a size limit. Implementations will usually grow in size internally to accomodate new elements.

`Set` is a collection of objects that does not allow duplicates.

- `SortedSet` is a `Set` that also specifies element ordering.
• SortedMap<K,V> is a Map<K,V> that also specifies element ordering.

Java 5 adds in a new collection type:

• Queue<E> is a collection of elements meant to be processed in a specific order. The implementation specifies whether this is FIFO or LIFO. This obsoletes the Stack class.

Java 6 adds in some new subtypes of collections.

• NavigableSet<E> is a Set<E> with special navigation methods built in.
• NavigableMap<K,V> is a Map<K,V> with special navigation methods built in.
• Deque<E> is a Queue<E> that can be read from either end.

Note that the above items are all interfaces. In order to use them, you must find the appropriate implementing classes, such as ArrayList, HashSet, HashMap, or PriorityQueue.

Each type of collection has multiple implementations that have different performance metrics and use cases.

Note that the Liskov Substitution Principle applies to the collection subtypes. That is, a SortedSet<E> can be passed to a function expecting a Set<E>. It is also useful to read about Bounded Parameters in the Generics section for more information on how to use collections with class inheritance.

If you want to create your own collections, it may be easier to inherit one of the abstract classes (such as AbstractList) instead of implementing the interface.

Prior to 1.2, you had to use the following classes/interfaces instead:

• Vector instead of ArrayList
• Dictionary instead of Map. Note that Dictionary is also an abstract class rather than an interface.
• Hashtable instead of HashMap

These classes are obsolete and should not be used in modern code.