Improve handling of `packageId` input errors

[gremau]

There are some functions in EDIutils that fail silently or with confusing error messages when an incomplete packageId parameter is supplied. For example, leaving out the package version for EDIutils::get_provenance_metadata('knb-lter-jrn.202') gives a confusing Internal Server Error message. I think a fairly simple error catch with some useful user feedback ("The packageId is incomplete") would be very helpful for debugging input errors, especially when EDIutils is being called as a dependency. Would utilities/parse_packageId be a good place to put this?

[clnsmth]

@gremau, thanks so much for spotting this issue and suggesting a fix! You're right, those silent failures and confusing error messages can be a pain.

Your suggestion to add an error catch in utilities/parse_packageId is a good one, but there are a few things to consider. About 25% of the functions that take a packageId argument don't actually use parse_packageId, so it wouldn't cover all cases.

We could create another utility or validation check and apply it higher up in the functions, but our preferred approach, and the one we typically use, is to let the error messages from the EDI repository itself trickle down to EDIutils. This helps preserve the original error handling and prevents EDIutils from accidentally introducing conflicting messages.

I'm going to raise this issue with the repository team to see if we can address it there. If not, then your original suggestion to add the error catch in utilities/parse_packageId definitely makes sense to pursue further.

Thanks again for your help on this!