Open tillmo opened 9 years ago
alternatively, the EBNF could be related to the abstract syntax (i.e. the MOF metamodel). This should be easier, because the MOF metamodel has been automatically derived from the EBNF. @MGlauer, could you briefly describe how this automatic construction works? (i.e.: each EBNF non-terminal leads to a MOF metaclass, etc.)
Given that MOF metamodel is normative and the EBNF is informative, I believe it would be more appropriate to describe it the other way round. E.g., describe the mapping from the MOF to the EBNF -- which should be easier, since the MOF contains more information.
agreed
We could split the EBNF in parts that mirror the package structure in our metamodel. This yields a glossary-like overview of Classes to their corresponding MOF picture.
good idea
Annex J: it’s not clear how this EBNF relates to the Concrete Syntax scattered through Section 9