Update derivative generation to use derivative rodeo to skip *TN.jpg and .READER.pdf files

[jeremyf]

In the <2023-03-14 Tue> conversation with Katharine, we have the following situation:

• We have archival PDFs for everything
• Some works have secondary PDFs

We can and should skip derivative generation for PD for those secondary PDFs. All secondary PDFs have the suffix .READER.pdf (make sure to test in a case insensitve manner). Example: =32000812.READER.pdf=

For a reference implementation (albeit with different rules):

• Adding Hyrax::ConditionalDerivativeDecorator utk-hyku#353

Related to:

• Conditional Derivatives: Inject module into registered to override function utk-hyku#343

We also do not want to create derivatives for TN.jpg files.

Testing Instructions

• In the UI, create a work
• Attach a simple PDF with unique text (something you could search on) that has the file suffix .READER.pdf (e.g. my-file.READER.pdf)
• Once the file is attached and processed, attempt to search for the text of that PDF. The created work should not be found.

[KatharineV]

Team, can we exclude PDFs with .READER.pdf and also .pdf-r.pdf? I recently found a big batch of material our digitization center processed with the .pdf-r.pdf file naming convention at some point in the past. We'd like to exclude these files from the viewer, as they are Reader files (but just didn't receive the correct name).

[jeremyf]

Absolutely going to add this bit of logic.

[jeremyf]

I want to de-prioritize this as the derivative work that I’m doing this week should resolve/supercede the changes that I’ve made to attempt to address this ticket.

(The importer process I'm working through will allow for significant improvements but is a complete re-architecture of the approach)

Duplicated/replaced by:

• EPIC: Import Optimization with Out of Band Processing #421

[laritakr]

We also do not want to create derivatives for TN.jpg files.

[jillpe]

dependent on derivative rodeo work