To come to a solution to the issues discussed in #301, I would like to propose that the Biolink build produces a simplified file for prefixes which is just a simple map from prefixes to expansions. This could be prefixes.yaml, prefixes.json, or both. This would give most users what they want—an easy to use map. Then the JSON-LD context would be freed up to become slightly more complex, adding "@prefix": "true" to every prefix object.

I'm hoping that then the RDFLIB JSON-LD library would be using the 1.1 spec, and using the more complicated context file we would get correct prefix expansions in the RDF serialization of Biolink model (@hsolbrig would this work in the current version?)

So what I am proposing are these two items:

• produce simplified prefixes.yaml or prefixes.json
• update JSON-LD context output to intersperse "@prefix": "true" as needed