The main goal of this project is to provide a DSL that allows take easily advantage of aggregate computing, which is a paradigm that takes full opportunity of the high availability of computational devices by the means of computational fields.
Using Kotlin Multiplatform allows the DSL to be used in Native, JS and JVM environments.
First of all, add the following plugin to your build.gradle.kts file:
plugins {
id("it.unibo.collektive.collektive-plugin") version "<latest version>"
}This plugin is used to apply the Kotlin compiler plugin used for the alignment of aggregate operator. Then, add the dependency to the project:
dependencies {
implementation("it.unibo.collektive:dsl:<latest version>")
}Currently, the project is composed by three submodules:
- plugin: it is divided in:
- compiler-plugin: the compiler plugin is used modify a data structure used to keep track of the stack at runtime. For each function, when its called, its name is registered in the data structure;
- gradle-plugin: necessary plugin that a gradle project uses in order to include the compiler plugin;
- dsl: the actual DSL, where the logic is implemented and that exposes the operator of the aggregate computing;
- collektive-test: this folder contains a test that allows to calculate the gradient between nodes using the simulation environment offered by Alchemist Simulator.