DataverseNO feed on UiT Research Data Portal

[philippconzett]

We'd like to have a DataverseNO feed on the UiT Research Data Portal (https://uit.no/forskningsdata) including metadata about the latest published datasets in UiT Open Research Data (https://opendata.uit.no/).

[philippconzett]

Cassandra has contacted the UiT communication department and got the following information on what is needed to get this feed implemented:

Når det kommer til hva vi behøver så er det altså XML data fra Dataverse for at vi skal få til en datastrøm som gjør at vi kan importere det/vise det på nettsiden. Så, hvis de har anledning å dele slikt med oss, kan vi prøve å se om vi får til en lignende løsning som vi har til Munin i dag.

På Munin feeden så har vi en lenke til en XML fil: https://munin.uit.no/open-search/discover?query=(type:doktorgradsavhandling)&format=rss&rpp=2&sort_by=2&order=desc'
Det er en slik vi trenger fra Dataverse.

[philippconzett]

@ubkm I guess we can use an equivalent/similar way to get metadata in XML from DataverseNO:

As ddi:
https://dataverse.no/api/datasets/export?exporter=ddi&persistentId=doi:10.18710/XFYDFL

As oai_datacite:
https://dataverse.no/api/datasets/export?exporter=oai_datacite&persistentId=doi:10.18710/XFYDFL

Or is there another, easier way? See the Dataverse API Guide:
https://guides.dataverse.org/en/latest/api/index.html

[philippconzett]

@ubkm Does the Munin feed only contain metadata about the latest published items? In DataverseNO, we could get this with a call like this:

https://dataverse.no/api/search?q=data&subtree=uit&subtree=uit&type=dataset&per_page=10&sort=date&order=desc

But the result is in JSON, not XML.

[ubkm]

@philippconzett The Munin feed contains the latest published doctoral theses, sorted by date in descending order. The search API in Dataverse seems to return only JSON, net XML. It might be OK for the UiT communication department to use JSON instead of XML, it is just a different format with the same content. We should ask them about that and if it's OK we could use the search API.

If they don't support it we might have to explore the export functionality.