Use GBIF library to generate DwcA

[cgendreau]

We should use GBIF's library to generate our dynamic DarwinCore Archive files.

[cgendreau]

@dshorthouse
Do we want to keep the 'acceptedNameUsageID' with pipe separated values for synonyms?
The alternative is maybe to use the 'Resource Relationship' extension like we plan for hybrids but it makes the whole archive a little bit more complicated.

[peterdesmet]

The pipe symbol is the most simple solution and doesn't seem to break anything as far as I know. Resource Relationship is a good alternative, but indeed more work.

[dshorthouse]

Not sure I follow. Why do you need more than one value for synonyms partially represented by acceptedNameUsageID? Each synonym should have its own record shouldn't it? acceptedNameUsageID = "An identifier for the name usage..."

[cgendreau]

No, could be the synonym of 2 accepted names.

[cgendreau]

See http://data.canadensys.net/vascan/taxon/9946?lang=en

[dshorthouse]

This may not be a nomenclatural issue, but a taxon concept issue. The only way to represent this in DwC is to create a new, dummy entry for an identically named synonym with the same scientificNameID as the other synonym. The synonyms would have different acceptedNameUsageIDs. And, if you want to indicate concepts, you'd also use taxonConceptID; the dummy synonym and one accepted name gets the same taxonConceptID and the second, original synonym & the other accepted name gets a different taxonConceptID. We have to ask Luc for more details about this example and possibly others that have been represented the same way.

[cgendreau]

This would also mean to have 2 records with the same taxonID which is not possible.

[dshorthouse]

No it doesn't, you'd need a new taxonID.

[cgendreau]

Well we should create a new issue since the title of this one is "Use GBIF library to generate DwcA".