PWGHF: fix index dereference toXiPi

[ZFederica]

Fix index dereference in treeCreatorToXiPi

[fgrosa]

I might be wrong, but I don't think that a template argument is needed, you can probably just use candidate.collision_as<MyEventTable>(). If you already tried and it does not work, we can add it.

[ZFederica]

Hi @fgrosa, I used the template following the suggestion Fabio gave me for the variables

O2Physics/PWGHF/TableProducer/treeCreatorToXiPi.cxx

Lines 315 to 316 in 247837f

	candidate.template piFromCharmBaryon_as<TMyTracks>().isGlobalTrackWoDCA(),
	candidate.template piFromCharmBaryon_as<TMyTracks>().itsNCls(),

(this was the PR #4672, I thought the two cases were equivalent).
If I remember correctly, without the template I was getting a compilation error (see https://ali-ci.cern.ch/alice-build-logs/AliceO2Group/O2Physics/4672/6e3e6e5301b2b64e122cf8001c069a08baa72da7/build_O2Physics_o2/pretty.html )

[fgrosa]

In the PR that you linked me there is no such implementation, i.e. no additional template parameter is added (you indeed use directly aod::TrackSelection), so maybe you were referring to another PR?
I am asking because here in principle there is no need of a template parameter, since you have always one type of collisions (but in any case it should probably be candidate.template collision_as<MyEventTable>(), since candidate is instead a template parameter)

[ZFederica]

I thought the point was how to do the dereferencing of the index in a template function. Yes, in that PR I used directly TrackSelection, then I changed in this PR https://github.com/AliceO2Group/O2Physics/pull/4709/files

[fgrosa]

Thanks @ZFederica!