[PWGJE] adding reduced tables for tracks, particles, collisions, mccollisions

[louisemillot]

PWGJE: add tacks, particles, collisions, mccollisions tables and producer

Description:

• Add AOD::SlimTracks with collision index and kinematic columns (pt, eta, phi, dcaXY)

• Add AOD::SlimCollisions with posZ, centFT0M, weight, ...

• AOD::SlimMcCollisions with posZ, centFT0M, weight, accepted, ptHard

• AOD::SlimParticles with Mccollision index and kinematic columns (pt, eta, Px, ... etc )

• Implement SlimTablesProducer to fill these tables from input AOD

• Add PROCESS_SWITCH for each processing method

Test:

• Builds locally: yes

[github-actions]

O2 linter results: ❌ 1 errors, ⚠️ 0 warnings, 🔕 0 disabled

[mhemmer-cern]

Dear @louisemillot
seeing this long list of commits trying to fix formatting and linter, I would suggest to look at https://aliceo2group.github.io/analysis-framework/docs/tools/#pre-commit-hooks.
This way you can make sure that there are no linter or clang-tidy or any other problems before opening a PR.

[louisemillot]

Dear @mhemmer-cern
Thank you for the suggestion! This is actually my first commit, so I really appreciate your guidance. I’ll make sure to set up the pre-commit hooks to avoid linter issues in the future

[louisemillot]

Hi,
just to clarify: the linter errors shown in this CI report are not caused by the new files added in this PR (SlimTables.h, slimTablesProducer.cxx).

All errors come from existing workflow names in PWGJE/TableProducer/CMakeLists.txt, unrelated to this PR.
I ran the O2 linter locally on the two files introduced here and both of them pass without any issues.

[nzardosh]

Hi I am not sure how this builds locally as you are trying to include subscriptions to tables that the task produces itself. The logic is completely wrong, as you are not subscribing to tables that have those columns in them.

Also the functionality of producing derived data already exists in the JE framework. It is of course fine if you want to write a task that further reduces the data but you should start with the JE derived data and reduce from there. This avoids running on the full dataset (unless there is a good reason to do so)

[vkucera]

Hi @louisemillot ,

Test: Builds locally: yes

That's a very low standard to claim that your code works. Follow the contribution guidelines before you make a PR and when you do, make it clear what your goal is, not just what you did.