Figure out if / how we move the NARCIS link from the legacy RSD to the new one

[ewan-escience]

We should have look at:

• https://www.narcis.nl/inclusion/Language/en
• https://www.narcis.nl/repository/repository/RSD-NLeSC/Language/en
• https://harvester.dans.knaw.nl/

[ewan-escience]

There is a harvester in the legacy RSD that we might be able to adapt

[jmaassen]

As far as I can tell, the harvester hardly uses any data in the RSD. It only retrieves a list of conceptDOIs for all published software, translates these into Zenodo IDs (using a DOI lookup on zenodo and extracting the resulting ID) and then asks Zenodo for the metadata of this conceptDOI in the datacite4 format.

It then caches this data on disk somehow. Not sure how it is served back to NARCIS

[jmaassen]

Example that works for Zenodo:

https://zenodo.org/oai2d?verb=GetRecord&identifier=oai:zenodo.org:5171937&metadataPrefix=datacite4 https://zenodo.org/oai2d?verb=GetRecord&identifier=oai:zenodo.org:344475&metadataPrefix=datacite4

Example that works for the RSD:

https://research-software.nl/oai-pmh?verb=ListRecords&metadataPrefix=datacite4

[jmaassen]

In the legacy RSD, application.py, line 337 contains the code to respond to the OAI-PMH call. Depending on the verb that is provided (ListRecords or GetRecord), it either returns the listrecords.xml or record-XXXX.xml file (where XXXX is a zenodo ID) which are produced by the harverster.

[ewan-escience]

The new data is live on Narcis, see https://act.narcis.nl/search/coll/dataset/Language/EN/meta_repositorygroupid/RSD-NLeSC