pods4k

Performance-oriented data structures for Kotlin

A collection of performance-oriented data structures providing performance and efficiency that might appear to be impossible at first glance.

If you're enjoying this library, please give it a and share it with others.

Dependency

Gradle:

dependencies {
    implementation("com.danrusu.pods4k:pods4k:0.4.0")
}

Maven:

<dependency>
    <groupId>com.danrusu.pods4k</groupId>
    <artifactId>pods4k</artifactId>
    <version>0.4.0</version>
</dependency>

Alternatively, you can pick & choose individual components of this library.

Immutable Arrays

Immutable Arrays are a safer and more efficient alternative to read-only lists while maintaining familiar operations.

Unlike regular arrays, the elements cannot be re-assigned, so immutable arrays can be safely shared without needing immutable collections:

val names = immutableArrayOf("Dan", "Bob")
println(names[0]) // "Dan"

names[1] = "Jane" // Compile error: No set method providing array access

On top of the core efficiency improvements, immutable arrays are even more efficient when dealing with the 8 base types as they use between 3 to 24 times less memory than lists and performance can improve by over 10 times!

Surprisingly, immutable arrays are even more efficient than regular arrays as they have hundreds of optimized specializations for common operations when working with or transforming into one of the 8 base types:

val people = immutableArrayOf(
    Person(name = "Dan", age = 3),
    Person(name = "Bob", age = 4),
) // ImmutableArray<Person>

// Since age is a non-nullable Int, this creates an efficient
// ImmutableIntArray storing primitive int values
val ages = people.map { it.age }
performStatisticalAnalysis(ages)

See Immutable Arrays for more specifics along with a detailed comparison against regular arrays, read-only lists, and immutable lists.

Java Compatibility

Requires JDK 11 or later. Automated tests are run on JDK LTS releases 11, 17, & 21.

Some data structures use techniques which are only supported by the Kotlin compiler. While they can be used in a mixed Java & Kotlin codebase to store one of these data structures in a Java-defined collection, referencing these data structures by their class name is only supported within Kotlin files.