Akka

The goal of the Akka translator is to generate all the infrastructure code needed to support the domain(s) defined in the RIDDL input. While this does produce a working system, it also relieves developers of much of the repetitive, detail oriented infrastructure that is relatively boring compared to the business logic of the system.

This translator recognizes that its output will be co-mingled with the business logic output that developers are writing and keeps them separate. When an expected developer file does not exist, it will create it, but otherwise uses Scala inheritance, or other mechanisms, to indicate the portions that ought to be written.

Output File Types

The Akka translator doesn’t just generate Scala code. It aims to generate complete projects that should be familiar to those famliar with Scala and Akka projects. This means generating a mixture of file types, as described in the following sections

sbt build files

build.sbt project/xxx.scala project/build.properties project/plugins.sbt

protobuffers definitions

Specifically processed by akka-grpc (via scalapb via protoc plugins)

Scala code

Build Time Dependencies

sbt 1.6 or later
sbt-riddl plugin
akka-grpc sbt plugin
buildinfo sbt plugin

Runtime Dependencies

Akka 2.6.17 (or later) including streams, http, cluster, persistence, …
Alpakka 3.x
Kafka

Comparison To Akka Serverless

The Akka Translator has an objective to support both the advanced Akka Scala developer and less knowledgeable developers. It does that by using Scala’s abstraction features but without completely disconnecting from Akka’s powerful features. For example:

Actors are used, but in a fill-in-the-blank way. Advanced programmers can still do more complicated things, neophytes will find it easy.
Does not use GRPC as the interface like Akka Serveless does. Scala is the interface and there is no side-car aspect in the runtime.