Improve how allen.py takes tag arguments

allen.py (used in running Allen offline) takes tags in the form

--tags=tag1,tag2 or --tags=dd4hep:tag1,tag2|detdesc:tag3,tag4

This MR adds more checks to the given tags as well as allowing tags in the form --tags=dd4hep:tag1,tag2 and --tags=detdesc:tag1,tag2

This stems from the following conversation: !1585 (comment 7929384)

added 1 commit

• ed668b11 - Fixed formatting

Compare with previous version

changed target branch from master to 2024-patches

changed the description

added RTA label

requested review from @raaij

Hi, may I ask what is the status here, do we expect more development? Or maybe we could do a ci test after rebase? Thanks a lot!

added 284 commits

• ed668b11...6a104d2b - 282 commits from branch 2024-patches
• c7043f42 - Improve how allen.py takes tags
• da7d4228 - Fixed formatting

Compare with previous version

assigned to @anmorris

requested review from @dovombru

mentioned in issue Moore#870 (closed)

approved this merge request

mentioned in issue Moore#889 (closed)

mentioned in issue Moore#890 (closed)

Dear @anmorris, this MR may need rebasing to master (2024 data taking is finished).

mentioned in issue Moore#896 (closed)

added 85 commits

• da7d4228...b4c32658 - 24 commits from branch 2024-patches
• b4c32658...1e090bda - 51 earlier commits
• 10353b4f - Merge branch 'bokutsen_NominalBeamline_qmtest' into 'master'
• f1cf5256 - Update References for: !1831 (merged) based on lhcb-master-mr/12131
• f5636aae - Use the TestFileDB everywhere instead of hardcoded paths on eos
• c1ef838b - Merge branch 'ra-remove-old-options' into 'master'
• 23674f36 - Merge branch 'ref_bot_Allen1831' into 'master'
• cc4cca74 - Merge branch 'fix-spurious-mcvertex-copy' into 'master'
• 7de9480b - Property refactor v2
• 5f1bcc75 - Merge branch 'ahennequ_property_refactor' into 'master'
• 31bacd8b - Improve how allen.py takes tags
• 6d2405ae - Fixed formatting

Compare with previous version

changed target branch from 2024-patches to master

added 1 commit

• 83d676b4 - Fix rebase mistake

Compare with previous version

/ci-test

Started integration test build. Once done, check the results or the comparison to a reference build.

This is clearly broken, I'll debug

On closer inspection I'm not sure this is broken. The tests which failed were due to timeouts, and running the same tests (same build) locally they seem fine. Would you mind giving this another ci-test? @sponce

I do not think we need a new test. I agree this is all fine. What is more worrying is the failure in the ci-test concerning physics-efficiencies. Can you check ?

It's interesting, I can't believe that it has anything to do with this MR - but it appears that a single long track is missed (out of >1e5) for a single graphics card. This is a minor enough change that as a first check I'll just rerun the job and see if that helps

Issue persists, this is very strange and unexpected

Is that problem with 2080Ti or different card? I am just traveling and hard to read logs on a phone.

Just TI2080, trying rerunning the 'run full' step for that card as well to see if that helps

That one was making issues last month, but is supposed to be fine now.

@cagapopo @dovombru Any idea if 2080Ti is misbehaving again?

Discussing with @cagapopo, we believe this to be related to #568 (closed) - not caused by this MR.

Should I then merge ? The ci-test were all fine.

ping

@cagapopo can this be merged in your opinion?

A gentle remind @cagapopo

added ci-test-triggered label