[BE] Support CSV metadata

[ioggstream]

I expect

• support CSV metadata

Instead

• they are just loaded and published as csv

Notes

ControlledVocabularyPathProcessor uses CsvParser. A more structured library such as frictionless data could support metadata via datapackages

[ioggstream]

When I load a CSV, I cannot attach metadata to it (e.g., the publishing API does not specify the x-jsonld-context and x-jsonld-type).

In order for an API, (e.g., https://schema.gov.it/api-docs?vocabIri=https%3A%2F%2Fw3id.org%2Fitalia%2Fcontrolled-vocabulary%2Fclassifications-for-people%2Feducation-level ) to be annotated with x-jsonld-context and x-jsonld-type
the CSV should map field data to classes.

Annotating a CSV with metadata allows to provide the semantic information in the REST API.

This issue is part of a broader architectural analysis aimed at serving data and OpenAPI schemas in a secure ad efficient way, without impacting on the harvesting infrastructure.