ReactiveState for Kotlin Multiplatform and Android¶

Easy reactive state management and ViewModels for Kotlin Multiplatform. No boilerplate. Compatible with Android.

ReactiveState-Kotlin provides these foundations for building multiplatform ViewModels and lower-level logic:

reactive programming: everything is recomputed/updated automatically based on straightforward code
demand-driven programming: resource-consuming computations and values are allocated on-demand and disposed when not needed
multiplatform: share your ViewModels and reactive state handling logic between all platforms
event handling: simple events based on interfaces (more composable and less boilerplate than sealed classes)
automatic error catching: no more forgotten try-catch or copy-pasted error handling logic all over the place
coroutine-based unit tests: worry no more about passing around CoroutineDispatchers everywhere
lifecycle handling
state restoration

See the ReactiveState documentation for more details.

Supported platforms¶

android, jvm, ios, tvos, watchos, macosArm64, macosX64, mingwX64, linuxX64

Installation¶

Add the package to your build.gradle’s dependencies {}:

dependencies {
    // Add the BOM using the desired ReactiveState version
    api platform("com.ensody.reactivestate:reactivestate-bom:VERSION")

    // Leave out the version number from now on:
    implementation "com.ensody.reactivestate:reactivestate"

    // Utils for unit tests that want to use coroutines
    implementation "com.ensody.reactivestate:reactivestate-test"
    // Note: kotlin-coroutines-test only supports the "jvm" target,
    // so reactivestate-test has the same limitation
}

Also, make sure you’ve integrated the Maven Central repo, e.g. in your root build.gradle:

subprojects {
    repositories {
        // ...
        mavenCentral()
        // ...
    }
}

Quick intro¶

The following two principles are here to give you a quick idea of the reactive programming aspect only. The “Guide” section in the documentation describes how to work with the more advanced aspects like multiplatform ViewModels, lifecycle handling, etc.

Note: While the discussion is about StateFlow, you can also use LiveData or even implement extensions for other observable values.

Observing StateFlow¶

Imagine you have an input form with first and last name and want to observe two StateFlow values at the same time:

isFirstNameValid: StateFlow<Boolean>
isLastNameValid: StateFlow<Boolean>

This is how you’d do it by using the autoRun function:

autoRun {
    submitButton.isEnabled = get(isFirstNameValid) && get(isLastNameValid)
}

With get(isFirstNameValid) you retrieve isFirstNameValid.value and at the same time tell autoRun to re-execute the block whenever the value is changed. That code is similar to writing this:

lifecycleScope.launchWhenStarted {
    isFirstNameValid
        .combine(isLastNameValid) { firstNameValid, lastNameValid ->
            firstNameValid to lastNameValid
        }
        .conflate()
        .collect { (firstNameValid, lastNameValid) ->
            try {
                submitButton.isEnabled = firstNameValid && lastNameValid
            } catch (e: CancellationException) {
                throw e
            } catch (e: Throwable) {
                onError(e)
            }
        }
}

Reactive StateFlow / reactive data¶

The same principle can be used to create a derived, reactive StateFlow:

val isFormValid: StateFlow<Boolean> = derived {
    get(isFirstNameValid) && get(isLastNameValid)
}

Now you can use autoRun { submitButton.isEnabled = get(isFormValid) } in the rest of your code.

Going even further, isFirstNameValid itself would usually also be the result of a derived computation. So, you can have multiple layers of reactive derived StateFlows.

Relation to Jetpack Compose / Flutter / React¶

Reactive UI frameworks like Jetpack Compose automatically rebuild the UI whenever e.g. a StateFlow changes. So, in the UI layer autoRun can usually be replaced with a Composable.

However, below the UI your data still needs to be reactive, too. Here ReactiveState provides derived to automatically recompute a StateFlow based on other StateFlows. This pattern is very useful in practice and provides the perfect foundation for frameworks like Jetpack Compose which primarily focus on the UI aspect. ReactiveState’s derived and autoRun provide the same reactivity for your data and business logic.

In Jetpack Compose you even have derivedStateOf which is very similar to derived. So, you can choose whether you want to build your business logic based on the official coroutines library (StateFlow/derived) or Jetpack Compose (State/derivedStateOf). However, the coroutines library has the advantage that it’s available for more platforms and it’s fully independent of any UI frameworks. Finally, most open-source non-UI libraries will probably be based on coroutines, so StateFlow based code might also be better for compatibility/interoperability.

In other words, the combination of both solutions used together results in a fully reactive, multiplatform codebase - which improves code simplicity and avoids many bugs.

Moreover, Jetpack Compose currently doesn’t provide any multiplatform ViewModel support or any large-scale architecture. So, this library solves that by providing BaseReactiveState for ViewModels. It also comes with a lifecycle-aware event system (eventNotifier) and loading state handling (so you can track one or multiple different loading indicators based on coroutines that you launch).

License¶

Copyright 2024 Ensody GmbH, Waldemar Kornewald

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

   http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.