Data model extension

[fenekku]

As a developer-hoster I want to be able to extend the core metadata schema with fields specific to my domain so that my metadata schema addresses my needs.

These custom fields are limited to the following types: array, string, integer and date.

Links of interest
Marshmallow schema

• https://github.com/zenodo/zenodo/blob/d56a7d5005d9982ebcaa7d7ea0091376000f017f/zenodo/modules/records/serializers/schemas/common.py#L498

jsonschema

• https://github.com/zenodo/zenodo/blob/d56a7d5005d9982ebcaa7d7ea0091376000f017f/zenodo/modules/records/jsonschemas/records/record-v1.0.0.json#L611

Inner filling for ES indexing

• https://github.com/zenodo/zenodo/blob/fecfe1c38415266c4dcefe09d6009a31b9bc3663/zenodo/modules/records/indexer.py#L84
• https://github.com/zenodo/zenodo/blob/fecfe1c38415266c4dcefe09d6009a31b9bc3663/zenodo/modules/records/utils.py#L161
• https://github.com/zenodo/zenodo/blob/fe72e36cc6d04102dfd6791ab1082c9b7af75d6d/tests/unit/records/test_records_utils.py#L69

An example of such extension would be added biomedical profile fields:

The "biomedical" profile will have the following extensions to the core metadata (see #1)

Field or equivalent	Notes	Why	Task	Implemented
Language language	Language of the content, optional. Just 1 for now.	filtering, legacy		in main model as language
Presentation location presentation_location	Location where content was presented, optional. Applies to exhibits and presentations. Geolocation values or controlled vocabulary values or both?	filtering, legacy
Content location content_location	Location pertaining to the content itself. (e.g., Uganda for dataset of vaccinated population in Uganda). Optional. Geolocation values or controlled vocabulary values (Feed from MeSH) or both? We may or may not be able to stick this in terms	searching, filtering, legacy		in main model as location
Number in sequence number_in_sequence	Indicates page or order of record in an ensemble. Integer, optional	sorting, in collection record ordering, synergy with part relation type
Private Note private_note	Free text, optional. Used internally for repo managers. SUPER_USER, librarian, owner, proxy can see it.	Need to understand use case better?		in main model as internal_notes
Subject: Name (re-use terms)	Name of person/organization referred in content (e.g. book about someone). Optional. Fed from controlled vocabulary	legacy, searching, filtering		in main model as subjects

Note that acknowledgements are not included for now.
"Abstract" and "content date" will be addressed by core metadata.

[lnielsen]

• Reference issues: records: support custom key/value pairs metadata zenodo/zenodo#1758, search: custom key/value pairs metadata zenodo/zenodo#1766, deposit: custom key/pair value pairs metadata REST API zenodo/zenodo#1765
• Commits/PRs: Custom metadata refactor zenodo/zenodo#1772, commit