Resolve more taxonID suppliers than WoRMS

[mdoering]

WoRMS LSIDs can be used in dwc:taxonID, dwc:taxonConceptID or dwc:scientificNameID to link an occurrence to the GBIF taxonomy. It currently get's used by pipelines to interpret the taxon in preference over the other verbatim taxonomy terms like dwc:scientificName.

There are other important and already often used reference taxonomies with stable identifiers that we should interpret. Consider to support the following:

• COL: ^col:([2-9BCDFGHJKLMNPQRSTVWXYZ]+)
• ITIS: ^(?:itis|tsn:)(\\d+)
• GBIF: ^gbif:(\\d+)
• IPNI: ^(?:ipni|urn:lsid:ipni.org:names):(\\d+-[1-3])
• UNITE: ^(?:unite:)(SH\\d+\.\\d\\dFU)
• BOLD: ^(BOLD:[A-Z0-9]+)

For non global identifiers we should follow the CURIE format (prefix:local-identifier) which is the recommended way to use taxonomic identifiers with ColDP.

[mdoering]

This should also be documented, see gbif/tech-docs#98