Please document/describe how datacite-v*.json files are generated

[yarikoptic]

Is your feature request related to a problem? Please describe.

Thank you for taking care about maintaining jsonschema serializations of the datacite schema.

In light of the removal of "identifiers" from the jsonschema in 4.3 to 4.5 (just discovered #80), and

• https://github.com/datacite/schema/issues/149

we decided to look into how such files were generated. Unfortunately, we failed to find anything relevant in the 160250d6c7ebd39f24c93aa5dfdb4dea3169754b commit comment, or

• 94

Describe the solution you'd like

A clear and concise description of how jsonschema json files are produced -- scripts which are run. And also when producing, which specifically datacite schema (4.5.1? 4.5.2? ... and ideally with the commit if not from versioned tag) was converted.

Describe alternatives you've considered

one feasible and "complimentary alternative" could be to use datalad run command to encode in the commit message what command produced the changes.

[tmorrell]

There is more discussion in https://github.com/inveniosoftware/datacite/issues/81 of why there are differences, but yes we should script the generation of the example files. I believe for 4.5 this was all manual. This would be a good to tackle when 4.6 comes out.

[yarikoptic]

Thank you @tmorrell for the reply, all your work and pointing to the details!