Kotlin PreviewGenerator (PrevGen)

Kotlin PreviewGenerator is a lightweight compiler plugin writ ten with Kotlin Symbol Processing (KSP). Preview Generator creates preview instances for data classes, which can be used in Jetpack Compose view previews.

Navigation

Setup
Usage
Examples
Roadmap
Feedback and contact us
License
Contribute
Configurations

Setup

Add the KSP plugin in the main build.gradle:

plugins {
    // ...
    id("com.google.devtools.ksp") version "1.9.21-1.0.15"
}

Add the Sonatype Maven URL:

repositories {
    // ...
    mavenCentral()
}

Add KSP and the module into Gradle dependencies:

dependencies {
    implementation("io.github.vram-voskanyan.kmp:PreviewGenerator:1.0.0") // take latest from Maven central
    ksp("io.github.vram-voskanyan.kmp:PreviewGenerator:1.0.0")
}

Sync the project, and you are good to go.

Usage

Add a @DataPreview annotation to the class that needs to have a generated preview.
Rebuild the project. It will generate a Preview class inside build\ksp\[BUILD_VARIANT]\ClassName.kt.

Examples

@DataPreview
data class DummyClass(val name: String, val age: Int, val loginDate: Long)

val previewValue = dummyClassPreview  

// dummyClassPreview is a generated class which is: DummyClass(name = "Ryan", age = 85, loginDate = 1705600601029)
// Note: `loginDate` is a Timestamp ;)

Do Not Do This

These are examples that will not work:

    class A(items: List<A>)

In this case, it will attempt to create an endless items.

    data class B(enumItem: EnumValue)

Currently, enums and sealed classes are not supported. Make the type nullable in this case:

    data class B(enumItem: EnumValue?)

For more configurations and setups, please refer here: Configurations.md

Example for Jetpack Compose

Medium Demo

Ongoing and Future Works

Since this project was developed only during non-office hours, not everything is covered. For example, the following types are not currently supported: Non-Data classes, Enum, Map, List (in progress).

We also have some exciting features in mind for implementation, depending on community interest:

Customized Data Size: Introduce an option to edit the generated data size for checking how the view behaves with large or small datasets. This may involve adding a parameter option to annotations.
Flexible Data Generation: Currently, the generator relies on an argument name-checker to pick values close to the parameter type. For the next step, we aim to provide an option to pass values from outside, allowing customization such as non-English values.
Adding caching functionality: Actually, the script is already fast, but there are plans to implement caching for pregenerated data. ;)

Contact Author

Email: vram.arm@gmail.com
LinkedIn: Vram Voskanyan

License

Apache License 2.0

Contributing

If you would like to contribute code you can do so through GitHub by forking the repository and sending a pull request. When submitting code, please make every effort to follow existing conventions and style in order to keep the code as readable as possible.

Vram-Voskanyan / PrevGen