reuploading changes for bounds of derived variables

[HGWright]

🚀 Pull Request

Description

I seem to have royally screwed up the old PR for this. So here are all of the relevant current changes for that.

Will copy over all of the discussions from the PR soon.

---

Consult Iris pull request check list

---

Add any of the below labels to trigger actions on this PR:

• benchmark_this Request that this pull request be benchmarked to check if it introduces performance shifts

[codecov]

Codecov Report

Attention: Patch coverage is 29.33333% with 53 lines in your changes missing coverage. Please review.

Project coverage is 89.60%. Comparing base (da199ee) to head (82be35f).

Files with missing lines	Patch %	Lines
lib/iris/fileformats/cf.py	28.37%	49 Missing and 4 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6350      +/-   ##
==========================================
- Coverage   89.79%   89.60%   -0.20%     
==========================================
  Files          90       90              
  Lines       23554    23608      +54     
  Branches     4391     4411      +20     
==========================================
+ Hits        21150    21153       +3     
- Misses       1662     1710      +48     
- Partials      742      745       +3     

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

[pp-mo]

Note (for my benefit!)
This basically replaces #6190

This comment is a useful summary of the agreed behaviour we should be delivering
-- plus the statement, just previous
"We can also avoid making this save to the best version of CF until a later date"
-- which means, as I read it, no change to saving at all (for now)

[stephenworsley]

Just a couple notes on places to improve the code. As mentioned offline, the major thing this needs now tests. It's worth pushing something up demonstrating expected behaviour even if they're currently failing.

[stephenworsley]

if you reach this point you will have a standard name, so or cf_root_var.long_name is unnecessary.

[stephenworsley]

I feel like this is probably not necessary. Would you not want to do something like bounds_name = getattr(cf_root_coord, "bounds", default=None)?

[stephenworsley]

Ultimately, I think you'll probably want to replace this try-except call. I expect you ought to be able to replace this with something like:

bounds_vars = [...]
if len(bounds_vars) == 1:
  bounds_var = bounds_vars[0]
  ...
else:
  ...