DwC-A handler needs to parse recordedByIDs and identifiedByIDs

[mandrewj]

The Biorepo portal (and symbiota light) now accept ORCID identifiers in the identifiedBy and recordedBy fields: BioKIC/Symbiota@e15726b

However, when these data go to GBIF they are presented as a text string which is somewhat cluttered. See for example:
https://www.gbif.org/occurrence/2620847404

GBIF has 2 fields: gbif:recordedByIDs and gbif:identifiedByIDs - see documentation here:
https://twitter.com/GBIF/status/1247133356724289536
https://gbif.github.io/dwc-api/apidocs/org/gbif/dwc/terms/GbifTerm.html#recordedByID

If an ORCID identifier is found, this identifier should go to the appropriate IDs field and only the remainder of the person's name should be delivered to the recordedBy field.

[dshorthouse]

Bumping this. Am also seeing Wikidata Q numbers here like on https://www.gbif.org/occurrence/4140918829. FWIW, Bionomia strips all these out during its agent parsing routine whereas had there been structured URIs in dwc:recordedByID and dwc:identifiedByID it'd resolve 'em. Likewise, filtering on GBIF proper would struggle with these identifiers embedded in dwc:recordedBy or dwc:identifiedBy.