Repeating columns for Dublin Core CSV files.

[RussellMcOrmond]

We currently only support repeated Dublin Core terms using the special sequence "||" to separate them.

This makes managing the fields harder. This ticket is to enhance the Dublin Core handling to support multiple columns in the same way that Archivematica does for the import metadata feature.

That documentation page includes an example csv table with all 15 Dublin Core terms, with 3 dc.subject columns.

In our case we have an objid" column and optional "label" column, rather than a "filename" column.

It is also easy to support both the "dc." prefix Archivematica uses, as well as the "dc:" prefix we currently use. We only support dc.element and not dcterms.element as well.

[RussellMcOrmond]

For testing I took the example from the Import Metadata documentation and changed the "filename" field to "objid".

archivematica-dc-example.csv

This seems to work well.

I would now want to get feedback form @nataliemacdonald and @BeS4 on how this works for them. The "Preview Metadata" feature of https://access.canadiana.ca/dmd/b885edadd03daf7b06f2670d21ceaeee shows what the Dublin Core XML will look like.

[RussellMcOrmond]

Enhancement has been in use for a few weeks with no concerns or bug reports, so merged the code into main.

• This does not replace existing functionality, and provides a migration path for staff processes:
  • method of using || to separate multiple values retained
  • using dc:element is supported as well as dc.element