An attempt to transform an asyncapi spec to a target schema.

Ideally this should:

  • parse an asyncapi spec, available in the "com.al333z" %% "asyncapi-gen-core" % "x.y.z" dependency
  • be able to generate the schema for the models in:
    • protobuf: able to output a .proto file content, available in the "com.al333z" %% "asyncapi-gen-protobuf" % "x.y.z" dependency
    • json
    • avro
  • be able to generate the scala code for the models (models, serdes)
    • protobuf: done via scalapb, available in the "com.al333z" %% "asyncapi-gen-protobuf" % "x.y.z" dependency
    • json
    • avro
  • be able to offer a set of utils which will simplify the consumer/producer client code (e.g. serdes, ...) for various broker/formats:
    • kafka/protobuf, available in the "com.al333z" %% "asyncapi-gen-kafka" % "x.y.z" dependency


This can be useful if in your organization there's a definition-first approach to define event-streaming platforms.

You can attach this codegen tool to a PR merge, commit or any relevant event of your event definition lifecycle, so that the events configured will auto-magically be transformed in all the sources needed from consumer/producer applications. After the generation, a good approach would be to actually tag the generated artifacts with a version and upload to a registry.

The reason behind this is that we should only be defining our contracts once, and let the machine generate the low level details such as value classes, serdes, etc.


  • Check and run the main in protobuf-kafka-example/src/main/scala/gen/Gen.scala. This will generate the schema (.proto) and the sources (java and scala) for the models and for the serdes, and will return a Topics companion object which will offer the user a nice way to consume/produce from/to a configured topic (e.g. user_events).
 object Topics {
   def userEvents: Topic[Int, gen.UserSignedUp] = ???
  // ...
  • run docker-compose -f "protobuf-kafka-example/docker-compose.yml" up, to turn on locally a working kafka env.
  • To see how the generated sources can be used, uncomment and run the SampleConsumer.scala and SampleProducer.scala.
  • You should see messages flowing, and thus, everything just working (de/serialization, topic consumption/production, etc...).

NB: the example here is not how you should use this tool. See Rationale section for a more principled approach.


No docs provided yet. The best things you can do to see how it works is checking/playing with tests.