check whether we are using the identifier schema correctly

[surchs]

The DATS identifier schema has two main fields:

"identifier": {
  "description": "A code uniquely identifying an entity locally to a system or globally.",
  "type" : "string"
},
"identifierSource": {
  "description": "The identifier source represents information about the organisation/namespace responsible for minting the identifiers. It must be provided if the identifier is provided.",
  "type" : "string"
  }

Currently, we are using "identifier" to store the non-dereferencable string name of an identifier and "identifierSource" to store a dereferencable IRI that points to the identifier. Here is an example.

I believe that the original intent of DATS was to do the opposite, i.e. to

• have "identifier" be the dereferencable IRI and
• have "identifierSource" be the string name of the organization / entity maintaining the identifier

Apart from the description in the schema, this also fits with the fact that "identifier" used to require a URI format (see datatagsuite/schema@f02264d).

We should discuss this and make an issue in the conp-dataset repo

[cmadjar]

Based on today's discussion at the CONP dev call, we move the URL of the species taxonomy to the identifier field instead of the identifierSource field.

@cmadjar will send a PR that modifies the validator to check that the URL is provided in identifier and not identifierSource so we can get a list of affected datasets.

@emmetaobrien will go through the list of datasets to modify and will modify the datasets that we can modify ourselves.

[cmadjar]

@surchs I created a new issue in conp-dataset. Would you mind taking a look at the issue and see if I forgot something?

CONP-PCNO/conp-dataset#712