use newer mappings for pycsw

[ricardogsilva]

Is your feature request related to a problem? Please describe.
Up until recently, pycsw, which GeoNode uses to provide compatibility with OGC CSW, only supported integration with downstream projects by means of using a single DB table as its source of metadata. This meant it was hard to integrate with projects that store metadata in a group of DB tables (AKA in a normalized DB way). This meant it was difficult for GeoNode to provide information on stuff like a datast's owner organization, related tags, etc via CSW.

Describe the solution you'd like
With the recent merge of geopython/pycsw#750 it is now possible to define a SQL view that can be used to present all GeoNode-dataset metadata as if it was stored in a single DB table (but instead of a table we use a view).

As such, it would be nice if GeoNode would revamp its current integration with pycsw to make use of this feature.

Describe alternatives you've considered
The alternative is to keep the status quo and not change the pycsw integration at all. It kind of works OK, with the caveats mentioned above.

Additional context
A nice bonus of modernizing the pycsw integration is that GeoNode would gain instant support for other metadata-related standards which are already supported by pycsw, specifically the new OGC-API: Records (the successor to the CSW standard) and STAC.

[t-book]

Great 👍

[gannebamm]

@ricardogsilva I think this requires quite some effort and is not a simple task. Do you provide funding for this feature request? The Thünen-Institute is still finalizing some specification documents for a tender process and one of our ideas is to extend the current metadata schema to better fit research data. This pycsw mapping would be very handy to map those new fields to the csw responses.

[ricardogsilva]

@gannebamm

Yeah, I realize the work would be substantial, but doable. At this stage I am not volunteering to do this work, nor do I have any funding for it, just signaling that it would be a nice feature to have 😄

[gannebamm]

PSC: @t-book @afabiani @giohappy @francbartoli
Do you think this should be extended to a GNIP? Maybe we have enough funding to get this done later this year.

[afabiani]

@etj

[mattiagiupponi]

We open an issue to do the upgrade to version 3.0 of PyCSW: #12540