Stricter array summary in CML output

[ukmo-ccbunney]

🚀 Pull Request

Closes: #6244

Description

---

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

[bjlittle]

@ukmo-ccbunney Lookin' good.

Just some initial comments to consider 👍

[codecov]

Codecov Report

❌ Patch coverage is 70.83333% with 21 lines in your changes missing coverage. Please review.
✅ Project coverage is 90.24%. Comparing base (fa6d61d) to head (7dda361).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/iris/coords.py	65.90%	11 Missing and 4 partials ⚠️
lib/iris/cube.py	78.57%	3 Missing and 3 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6743      +/-   ##
==========================================
- Coverage   90.29%   90.24%   -0.06%     
==========================================
  Files          91       91              
  Lines       24475    24613     +138     
  Branches     4571     4604      +33     
==========================================
+ Hits        22100    22212     +112     
- Misses       1607     1624      +17     
- Partials      768      777       +9     

☔ 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.

[trexfeathers]

A real-life case of CML protecting us, if you were interested:

SciTools/iris-grib#708

[ukmo-ccbunney]

The CML output for testing (generated from _shared_utils.assert_CML now sets the following new formatting options:

• numpy_formatting=False : Don't use numpy formatting, rather use the simplified iris.utils.array_summary formatter in this PR.
• coord_checksum=True : Generate checksum for coord like cube components. This provides a guarantee the coord data is all identical as the array summary will only show by default the first and last 3 elements.

No other formatting options have been enabled by default for the CML generation in the tests (such as mask_count or coord_data_array_stats) as these are not suitable for all existing test output.

Commit 95f3a42 covers all the updated CML files + one change to unit/pandas/test_pandas.py to disable the coord_checksum as some tests have coord data that is stored as an object dtype which doesn't generate a consistent hash.

[ukmo-ccbunney]

BTW: 2 new CML outputs were added as part of 7791944 to provide test coverage or the new formatting options.

[bjlittle]

@ukmo-ccbunney This is an impressively nimble refresh to some thorny historic code 🥇

Just a few minor observations to service, otherwise LGTM 👍