Topology tolerance [AVD-1723]

[pp-mo]

🚀 Pull Request

Description

Allow "wrong" topology-dimension on a UGRID mesh, deduce "correct" value from attached connectivites.

This is just one of the pending "tolerant loading" ideas which we listed elsewhere (AVD-1723).
But this one will help with the forthcoming internal release demo.

---

Consult Iris pull request check list

[stephenworsley]

Looks good, just a few comments to address.

[pp-mo]

Thanks @stephenworsley
I hope the latest commit addresses your suggestions
Please re-review !

N.B. I'm not sure if the broken docs links are transient : I have re-spun a couple of times and they seem to consistently fail.
Depending on the current state of the target branch, we can either consider merging this without fixing that, or perhaps we will need a separate fix on the target branch.
I am currently re-running this test on the latest mesh-data-model branch commit : That succeeded ~1 week ago when merged, but I suspect will fail now...

[stephenworsley]

Looks good, I think this is good to merge.

[pp-mo]

Back, with requested changes to warnings/logging.

@trexfeathers and/or @stephenworsley : Please re-review !

[pp-mo]

Thanks @trexfeathers for pointing out the existing functionality for logging tests.
I was very close to committing effort to writing that myself !

[trexfeathers]

Thanks @pp-mo! Couple of little things to address.

[pp-mo]

Thanks for suggestion. Done that, I hope ?
BTW I also moved the support vars for the synthetic data calls into the inside of the logging context block, which I think reads better.

Please re-review @trexfeathers

[trexfeathers]

Merging despite the link check failures. Once #4104 is in, we'll do a merge-back to get that fix into mesh-data-model.

[pp-mo]

Many thanks, significantly better for your suggestions @stephenworsley @trexfeathers !
💐