petertrr / merge-data-class

Kotlin Symbol Processor to construct data classes from incomplete sources while preserving nullability cheks

merge-data-class KSP processor

This is a Kotlin Symbol Processor implementation that is intended to ease some tasks related to data classes by generating boilerplate code that helps to preserve type-safe nullability when working with incomplete data sources.

Sometimes there are cases when data comes from different sources, e.g. some fields are read from configuration file and others are set by user on the command line. After application startup those two sets of values sum up into a data class where all fields have to be present, i.e. they are non-nullable in terms of Kotlin type system. So, to construct such a class one might want to read all values manually, or create a special logic for iterating over class properties etc. This approach however leads to a lot of boilerplate.

Or there might be another approach - to make properties nullable, construct data class instances from all data sources and then merge them. However, this passes nullable properties down the whole application and introduces a huge number of unnecessary !! all over the code base.

merge-data-class serves as a pretty simple code generation solution for this case. Based on a data class it would generate a class with nullable properties and a merge method, which would return an instance of the original class while also performing nullability checks.

Consider the code below:

@BuildFromPartial
data class Foo(
    val field1: Type1,
    val field2: Type2?,
)

merge-data-class will generate the following code:

data class FooPartial(
    val field1: Type1?,
    val field2: Type2?,
) {
    fun merge(other: FooPartial): Foo {
        return Foo(
            field1 ?: other.field1,
            field2 ?: other.field2
        )
    }
}

Using it in a gradle project

This plugin can be used as any other KSP processor: apply the KSP plugin and add required dependencies:

plugins{
    kotlin("ksp") version "1.6.21"
}

## Using it in a KMP project
As described in [Kotlin docs](https://kotlinlang.org/docs/ksp-multiplatform.html), there are some differences
in using KSP in Multiplatform proejct over a single-platform project.
kotlin
plugins{
    id("com.google.devtools.ksp") version <ksp version>
}

kotlin {
    sourceSets {
        commonMain {
          implementation("io.github.petertrr.ksp:merge-data-class-annotations:0.1.0")
        }
    }
}

dependencies {
    "ksp"("io.github.petertrr.ksp:merge-data-class-annotations:0.1.0")
    "kspCommonMainMetadata"("io.github.petertrr.ksp:merge-data-class-ksp:0.1.0")
}
  
dependencies {
    compileOnly("io.github.petertrr:merge-data-class-annotations:0.1.0")
    ksp("io.github.petertrr:merge-data-class-ksp:0.1.0")
}