Item identifiers including filename extensions

[ghukill]

In a recent PR, a question + response comment touches on this question: #105 (comment).

Two interlated questions emerge for me:

1. should item identifiers in DSC ever contain filename extensions (e.g. ".pdf")?
2. if they do include them, do these directly become identifiers anywhere in DSpace?

I understand that stakeholders, at least in the past, have included filename extensions in some kind of Item Identifier column in CSVs used for uploads. If the answer to #2 above is "No", then it probably doesn't matter; it sounds like it's a 100% internal identifier that is gone after DSC and DSS ingest the item.

Another question though, how does this scale for items that may have multiple files? Suppose it's a SimpleCSV style workflow, and there are 3 PDFs? Is this when we get into juggling _01 suffixes in the filenames? and is that somehow stripped from the item identifier?

[ehanson8]

Ahhh, I get your concern now. # 2 is a no, it is purely for DSC and DSS and it is never written to the metadata.

create_dspace_metadata even has this line to ensure it never goes in:

        for field_name, field_mapping in self.metadata_mapping.items():
            if field_name not in ["item_identifier"]:

                field_value = item_metadata.get(field_mapping["source_field_name"])

And the item identifier in the metadata CSV can be D123.pdf but it is just as often D123 when multiple files are expected (e.g. D123_01.pdf, D123_verso.pdf). Checking that the item identifier is "in" the file name just happens to work in both cases

Does that make sense?