Currently, if fields are in the wrong order, or omitted, etc. for the attribute definitions and id definitions for any of the kinds, this will lead to invalid schema being generated (e.g. schema transaction data with nils, missing required fields like cardinality, etc.) -- this can be handled with simple spec based collection regex validation. A spec for the Unify compatible schema generation DSL will also serve as documentation for other Clojurists. (Though more natural language documentation of that spec should be provided as well).
Currently, if fields are in the wrong order, or omitted, etc. for the attribute definitions and id definitions for any of the kinds, this will lead to invalid schema being generated (e.g. schema transaction data with
nils, missing required fields like cardinality, etc.) -- this can be handled with simple spec based collection regex validation. A spec for the Unify compatible schema generation DSL will also serve as documentation for other Clojurists. (Though more natural language documentation of that spec should be provided as well).