Xtext: Customize inferred genmodel

0
votes

I'm trying to upgrade an Xtext project to version 2.9 but I cannot figure out how to post process a meta model (inferred from Xtext grammar) when using v. 2.9. In previous versions it was possible to just create a ModelnamePostProcessor.ext file with the necessary custom code and ensure that the MWE2 file contained the necessary setup like so:

fragment = ecore.EMFGeneratorFragment auto-inject {
    emfRuntimeVersion = "2.10"
}

However, when using Xtext 2.9 it appears that the extension doesn't get picked up at all. I didn't make any changes to the .ext file, but the MWE2 workflow looks a bit different:

//...
language = StandardLanguage {
    name = "org.xtext.example.mydsl.MyDsl"
    fileExtensions = "mydsl"

    fragment = ecore.EMFGeneratorFragment2 auto-inject {
        emfRuntimeVersion = "2.9"
    }
    serializer = {
        generateStub = false
    }
}

Does anyone know what's wrong here or is aware of some documentation/example on how to post-process meta models with Xtext 2.9?

1
eclipsextext

1 Answers

1
votes

this is no longer possible. you may have a look at, comment and follow the workaround in https://bugs.eclipse.org/bugs/show_bug.cgi?id=483209