kmm-job-queue

Coroutine job scheduler for Kotlin Multiplatform projects. Run & repeat tasks. Rebuild them from disk. Fine tune execution with rules.

The library depends on kotlinx-serialization for the persistence of the jobs.

⚠️ The project is still work in progress and shouldn't be used in a production project.

Rules

• Delay
• Timeout
• Retry
• Periodic
• Unique
• Network

Capabilities

• Cancellation (all, by id)
• Start & stop scheduling
• Restore from disk (after start)

Example

Define a Task (or DataTask<T>), customize its body and limit when it should repeat.

⚠️ Make sure the data you pass into the task is serializable.

@Serializable
data class UploadData(val id: String)

class UploadTask(data: UploadData): DataTask<UploadData>(data) {
    override suspend fun body() { /* Do something */ }
    override suspend fun onRepeat(cause: Throwable): Boolean { cause is NetworkException } // Won't retry if false
}

Create a single instance of the job queue on app start. To start enqueuing jobs run jobQueue.start().

⚠️ You have to provide the polymorphic serializer of your custom task if you want to persist it.

You can pass a custom Queue.Configuration or JsonStorage to the job queue.

val jobQueue = JobQueue(serializers = SerializersModule {
    polymorphic(Task::class) {
        subclass(UploadTask::class, UploadTask.serializer())
    }
})
jobQueue.start()

You can customize the jobs life cycle during schedule by defining rules.

val data = UploadData(id = "123456")
        
jobQueue.schedule(UploadTask(data)) {
    unique(data.id)
    retry(RetryLimit.Limited(3), delay = 30.seconds)
    persist()
}

You can subscribe to life cycle events (e.g. for logging).

jobQueue.listener.collect { event ->
    when (event) {
        is JobEvent.DidFail -> Logger.error(event.error)
        else -> Logger.info(event)
    }
}