Choosing List Implementation in Java: ArrayList, LinkedList, CopyOnWriteArrayList, List.Of

Luqman Rom — Wed, 05 Feb 2025 23:32:44 +0000

To choose which list you need you to use requires some basic understanding of each of the implementations.

Summary

Implementation	Underlying Structure	Thread-Safe?	Remarks
ArrayList	Dynamic Array	No	General-purpose, random access.
LinkedList	Doubly Linked List	No	Frequent insertions/deletions. Slower access (O(n)).
CopyOnWriteArrayList	Dynamic Array	Yes	Thread-safe, read-heavy workloads.
List.of() / Immutable	Fixed Array	Yes	Unmodifiable data.

When to Use Which?

ArrayList: Default choice for most cases.
LinkedList: Frequent insertions/deletions at both ends (e.g., queues).
CopyOnWriteArrayList: Thread-safe reads with infrequent writes.
Immutable Lists (List.of()): Data that never changes.

Thread-Safe Demo

Here is examples to see how ArrayList is not thread safe

package org.example;

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

public class ArrayListThreadSafetyDemo {
    public static void main(String[] args) throws InterruptedException {
        List<Integer> unsafeList = new ArrayList<>();

        // Create two threads that add elements to the same list
        Thread thread1 = new Thread(() -> addElements(unsafeList));
        Thread thread2 = new Thread(() -> addElements(unsafeList));

        thread1.start();
        thread2.start();

        // Wait for both threads to finish
        thread1.join();
        thread2.join();

        // Expected size: 2000 (1000 elements per thread)
        // The size is often less than 2000, and the program may occasionally throw exceptions like ArrayIndexOutOfBoundsException.

        /*
            Race Condition: Two threads might both try to increment the size field and write to the same array index, causing one write to be overwritten.
            Internal Array Corruption: The internal elementData array might resize inconsistently, leading to index errors.
            Lost Updates: The size counter may not update atomically, resulting in fewer elements than expected.
         */
        System.out.println("Final list size: " + unsafeList.size());
    }

    private static void addElements(List<Integer> list) {
        for (int i = 0; i < 1000; i++) {
            list.add(i); // Concurrent modification
        }
    }
}

To have a thread safe list, choose CopyOnWriteArrayList instead. Since it is coming from same List interface, most methods are the same.

package org.example;

import java.util.List;
import java.util.concurrent.CopyOnWriteArrayList;

public class CopyOnWriteArrayListThreadSafetyDemo {
    public static void main(String[] args) throws InterruptedException {
        List<Integer> unsafeList = new CopyOnWriteArrayList<>();

        // Create two threads that add elements to the same list
        Thread thread1 = new Thread(() -> addElements(unsafeList));
        Thread thread2 = new Thread(() -> addElements(unsafeList));

        thread1.start();
        thread2.start();

        // Wait for both threads to finish
        thread1.join();
        thread2.join();

        // Expected size: 2000 (1000 elements per thread)
        System.out.println("Final list size: " + unsafeList.size());
    }

    private static void addElements(List<Integer> list) {
        for (int i = 0; i < 1000; i++) {
            list.add(i); // Concurrent modification
        }
    }
}

You can also use to have a thread-safe list.

List<Integer> safeList = Collections.synchronizedList(new ArrayList<>());

Checking version updates in package.json

Luqman Rom — Mon, 23 Oct 2023 01:10:49 +0000

There is a nice tool to check the project dependencies automatically. As for now, it reaches 8k+ stars on Github, and 300k+ weekly downloads. Well liked by the community.

Here are the steps:

npm install -g npm-check-updates
ncu
ncu -u

DEV Community: Luqman Rom

Choosing List Implementation in Java: ArrayList, LinkedList, CopyOnWriteArrayList, List.Of

Summary

When to Use Which?

Thread-Safe Demo

Checking version updates in package.json