Make Subroutine.argnames return lower-case names

[Claude]

Thanks for asking me to work on this. I will get started on it and keep this PR's description up to date as I form a plan and make progress.

---

This section details on the original issue you should resolve

<issue_title>Make Subroutine.argnames return lower-case names</issue_title>
<issue_description>To avoid patterns like this, and because Fortran is case-insensitive by default, we should make Subroutine.argnames return lower-case arg names .

    routine_arg_names = [arg.lower() for arg in routine.argnames]

[No action] We should make the routine.argnames property lower-case by default.

Originally posted by @reuterbal in #671 (comment)
</issue_description>

<agent_instructions>Start by implementing the lower-casing in Subroutine.argnames and make sure all tests are still passing. If tests are failing, fix the implementation to make tests pass.

Subsequently, check for now unnecessary lower-casing in places that use Subroutine.argnames and clean-up these uses.</agent_instructions>

Comments on the Issue (you are @claude[agent] in this section)

[reuterbal]

@claude[agent] The tuple conversion is redundant here.

Suggested change

	trafo_data = {'orig_argnames': tuple(routine.argnames)}
	trafo_data = {'orig_argnames': routine.argnames}

[Claude]

Removed the redundant tuple conversion in commit 39cdc34 since routine.argnames now returns a tuple directly.

[github-actions]

Documentation for this branch can be viewed at https://sites.ecmwf.int/docs/loki/686/index.html

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 94.20%. Comparing base (03c986f) to head (0f46e15).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #686   +/-   ##
=======================================
  Coverage   94.20%   94.20%           
=======================================
  Files         288      288           
  Lines       47546    47546           
=======================================
  Hits        44790    44790           
  Misses       2756     2756           

Flag	Coverage Δ
lint_rules	96.40% <ø> (ø)
loki	94.17% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[reuterbal]

OK, this should be ready for a look @mlange05.

It was meant as a small maintenance task after noticing that argnames was case-sensitive, but it turned out to require more modifications than anticipated because particularly case-sensitive transformation paths (e.g., C transpilation) used that property.

So, the question is whether we want to have the property lower-case, like most other things, and accept that we have to be more explicit and case-aware in the case-sensitive code regions (using [arg.name for arg in routine.arguments] or similar.
Or if we retain current behaviour and use explicit lower-casing in those instances where we don't want to be case-sensitive (e.g., [arg_name.lower() for arg_name in routine.argnames])

[mlange05]

Ok, as discussed offline, this API change has a few pro and few con uses. As the number of "pros" seem to outweigh the "con" uses, I think we should accept this.

I'd like the docstring of the property to make this change explicit - and we might want to bundle with with the other impending API-breaking change. But otherwise GTG from me.

[mlange05]

Could we update the docstring here please to make the shift in API explicit?