[EMCAL-689] EMCAL/Geometry: Add function to load alignment matrix fro…

[mhemmer-cern]

…m CCDB

• Add function SetMisalMatrixFromCcdb to set the missalignment matrices for the EMCal via the CCDB. This way they are not laoded from the GeoManager. The function expects a path inside the ccdb, which by default is set to "Users/m/mhemmer/EMCAL/Config/GeometryAligned" and the timestamp. Currently in the CCDB we only have the old Run 2 alignment object. Once we have the new alignment objects we can uplaod them and via the correct time stamp load them fittingly.

[github-actions]

REQUEST FOR PRODUCTION RELEASES:
To request your PR to be included in production software, please add the corresponding labels called "async-" to your PR. Add the labels directly (if you have the permissions) or add a comment of the form (note that labels are separated by a ",")

+async-label <label1>, <label2>, !<label3> ...

This will add <label1> and <label2> and removes <label3>.

The following labels are available
async-2023-pbpb-apass4
async-2023-pp-apass4
async-2024-pp-apass1
async-2022-pp-apass7
async-2024-pp-cpass0
async-2024-PbPb-apass1
async-2024-ppRef-apass1
async-2024-PbPb-apass2
async-2023-PbPb-apass5

[shahor02]

In principle, fine, but why do you need a data member for the ccdbManager?
The latter is anyway a singleton, it is enough to do all operations on the ...::instance().

[mhemmer-cern]

In principle, fine, but why do you need a data member for the ccdbManager? The latter is anyway a singleton, it is enough to do all operations on the ...::instance().

Ah good point, I will change this quickly. Thanks @shahor02

[alibuild]

Error while checking build/O2/fullCI_slc9 for 3404588 at 2025-05-05 19:25:

No log files found

Full log here.

[shahor02]

I had the impression that this was just a temporary hack to collect data for the real alignment. @mhemmer-cern, is that not so?
Otherwise, I don't see the point in loading non-standard alignment objects (matrices) from the private CCDB repo.

[mhemmer-cern]

I had the impression that this was just a temporary hack to collect data for the real alignment. @mhemmer-cern, is that not so? Otherwise, I don't see the point in loading non-standard alignment objects (matrices) from the private CCDB repo.

No you are right, this is just temporary solution till we have an EMCal alignment object that we trust and with what we would want to create a new official AlignmentObject.

[alibuild]

Error while checking build/O2/fullCI_slc9 for 56894e1 at 2025-05-13 23:59:

## sw/BUILD/arrow-latest/log
/sw/BUILD/5c423fbe2eb3b52187e935043b0d4602548f9a00/arrow/src_tmp/cpp/src/gandiva/precompiled/arithmetic_ops.cc:18:10: fatal error: 'cmath' file not found
ninja: build stopped: subcommand failed.

Full log here.

[mhemmer-cern]

I just force pushed to restart the build. In general I think this can be merged now. The CCDB access is now done in an isolated way. The actual usage of this needs to be implemented in O2Physics, for which we need this PR first merged.

[jokonig]

Looks good to me