[Common] Unified, simple, fully self-configuring mult/cent provider

[ddobrigk]

• adds a task to tentatively replace mult/cent services with a single, self-configuring, lightweight device following the rationale of the strangenessbuilder. Track loops / slices replaced by single operations for faster operation in all processing paths.
• now with run 2 and run 3 modes tested thanks to @romainschotter. Still some tuning to be done with defineDataProcessing autodetect, but will follow in separate PR.

[github-actions]

O2 linter results: ❌ 32 errors, ⚠️ 0 warnings, 🔕 0 disabled

[ddobrigk]

Tagging @romainschotter - whenever you have a moment, please help me with testing

Tagging @iarsene - this should provide a measurable reduction in memory footprint but you should only use this once it is working and has been confirmed to lead to the same output as the legacy tasks

Some further features to be added:

• bc-based centrality
• Run 2 style truncation for anchor point calculation

[ddobrigk]

@fjonasALICE @nstrangm this PR will provide a unified mult/cent task that will include BC-level centrality values for FT0 signals and, while at it, it will simplify the task, using less memory and less CPU. I will finish some testing soon and then move the PR to review, but even then, since this is intended to be a replacement for multiplicityTable and centralityTable, we'll have to commission it a bit. Hope this helps!

[romainschotter]

To make sure that the multiplicity/centrality tasks and the new multCentTable task produce the same centrality distribution, here is a comparison of the centrality distribution produced by those tasks in real data, in MC data and in Run2 converted data:

As you can see, the new multCentTable task produces the same results as the multiplicity/centrality tasks

[ddobrigk]

We're still missing the autodetect at defineDataProcessing but let's move this to review and take care of that extra feature later. Might still do a little cleanup etc before going ahead

[alibuild]

Auto-approving on behalf of @ddobrigk.