Export DH record to InvenioRDM record

[fenekku]

Task

Edited Feb 10 2021

This is the epic for the export of a DH record as a whole to an InvenioRDM importable record.

Concretely, it means

1- generating the JSON for creation of a draft:

• shape of payload JSON to create here: https://inveniordm.docs.cern.ch/reference/rest_api/#create-a-record-draft
• details on
  • metadata field here: https://inveniordm.docs.cern.ch/reference/metadata/#metadata
  • access field here: https://inveniordm.docs.cern.ch/reference/metadata/#access-information
  • ext field here https://inveniordm.docs.cern.ch/reference/metadata/#extensions [NOT OPERATIONAL]

2- generating the separate JSON for upload of the file associated with the record. In DH each record has a single file.

• shape of payload JSON to create here: https://inveniordm.docs.cern.ch/reference/rest_api/#start-draft-file-uploads
• This is a very simple JSON containing the key (name of the file)

3- Connecting those 2 JSONs with the actual binary file.

• possible avenues include: nested folders (beware of filesystem limitations) or additional ersatz JSON connecting them together,
• At end of day, importer must be able to connect content of 1- and 2- with appropriate file path. That's what is important.

Subtasks:

1- Subtasks

• Metadata epic Export Invenio Metadata Field #787
  • Creators Export Metadata Subfield Creators #791
  • Contributors Export Metadata Subfield Contributors #796
  • Version Export Metadata Subfield Version #802
  • Sizes Export Metadata Subfield Sizes #800
  • Rights Export Metadata Subfield Rights #803
  • languages Export Metadata Subfield Languages #798
  • funding Export Metadata Subfield Funding References  #805
  • references Export Metadata Subfield References #806
  • identifiers Export Metadata Subfields Identifiers #799
  • related identfiers Export Metadata Subfields Identifiers #799
• Access / ownership
  • Basic access Export Invenio Access Field #784
  • See below and Create list of netid, email, fullname from owners of records #753

2- Subtasks

• "file" export Export Invenio Files Field #783

Additional helpful context

Example shape of full record metadata in here: https://github.com/inveniosoftware/invenio-rdm-records/blob/master/tests/conftest.py

metadata field schema: https://github.com/inveniosoftware/invenio-rdm-records/blob/master/invenio_rdm_records/marshmallow/json.py

Some of the data input are obviously impossible to generate yet (e.g. user id, which is something we have to solve on our side anyway). Adding a field:

"imported_owners": [
  {
    "netid": "...",
    "email": "...",
  },
  ...
]

in the "tie-ing up together JSON" or other should be enough. And is marked as subtask: #753

[saragon02]

For reference: InvenioRDM metadata fields mapped to DigitalHub fields: https://northwestern.box.com/s/1e1ogwnz4vrh7upmrlfuf3nk4vw1ap64

[sharpattack]

Depends on #764