Wildcard fragments

[tfonda-fbk]

Hi, I have recently discovered that one can define a fragment with a wildcard instead of the returns <Type> clause:

fragment MyFragment *:
    feature=STRING;

I observe that the immediate consequence of this definition is that a MyFragment type is not created during Ecore model inference. My question is: has this definition got any other (side-)effect, or were wildcards introduced exclusively for the purpose of suppressing default fragment type generation?
Thanks!
BTW, it looks like the Xtext documentation completely lacks a section about fragments. I will be happy to contribute such a section as soon as I get an answer to this question :)

[szarnekow]

I will be happy to contribute such a section

That's greatly appreciated.

Wildcard fragments serve the following purposes:

• For generated Ecore packages, they allow for suppressing the inference of a type by hoisting the structural features in the caller's type
• With imported Ecore packages, they help to decouple the concrete syntax from the existing package structure since common syntactic rules do not necessarily enforce a common super class in the EPackage