search

caltechlibrary / irdmtools

A Go and Python package for working with InvenioRDM repositories.
https://caltechlibrary.github.io/irdmtools
Other
1 stars 1 forks source link

doi2rdm: ISSN mapping #38

Closed tmorrell closed 1 year ago

tmorrell commented 1 year ago

Switch ISSN mapping from alternativeIdentifiers to custom_fields.journal:journal.issn

rsdoiel commented 1 year ago

How to we handle ISSN for electronic version versus print version (some publications have both)?

rsdoiel commented 1 year ago

Fixed in the upcoming v0.0.23 release for both crossref imports and eprint2rdm.

rsdoiel commented 1 year ago

If we have a single ISSN it just puts that in the journal:journal.issn field, otherwise I save an array of ISSN to that field.

tmorrell commented 1 year ago

RDM doesn't like the array for multiple ISSNs. I think for the moment we pass it as an array in a string '[ "1745-2473", "1745-2481"]', and see what folks think of that.

rsdoiel commented 1 year ago

Could we just delimit that with comma and skip the brackets?

rsdoiel commented 1 year ago

Second question, if we save this as either a text of a JSON array or simply delimit with a comma isn't this going to cause idutils to fail validating the ISSN?

tmorrell commented 1 year ago

Ah, validation. That's probably it. Here is a new proposal which is a bit more complicated but probably the right solution.

If CrossRef has two or more ISSN values, either run a query against the journal vocabulary (e.g. https://authors.caltechlibrary.dev/api/vocabularies/caltechjournals?q=id:2053-1583) or look at this file https://github.com/caltechlibrary/ames/blob/main/journal-names.tsv. Ideally this would be against cold in the future, so just pick whatever you prefer.

tmorrell commented 1 year ago

This has been fixed for now by putting the first ISSN in the journal ISSN field and any others in the alternative identifiers field. A fancier implementation will come later