Add extra particles and more hadron combinations to radinclusive lines

@yafan @alobosal The MR includes to main points:

• Add more decay tree possibilities. Besides having charged hadrons, we may have K_s⁰ or Λ⁰ in their place, with the restriction that two neutrals cannot be directly combined (only charged with charged or charged with neutral).
• The BDTs remain unchanged for the time being. We need neutral MCmatching in Moore reco mode to have a proper signal proxy for channels containing K_s⁰ or Λ⁰. But we have the setup ready for when that is possible.
• As for the extra particles saved, we not only used MinBias to determine how many extraparticles are being saved on average, but also used signal MC to cross-check that the efficiency (in terms of particles already saved by the main decay trees) is >=60%. A summary of the performance of the "extra selections" is as follows:

Particle	Cut proposed	Channel tested	Signal eff	Multiplicity
ExtraHadron	K_PT>1200 && K_MINIPCHI2 > 150	K1G_Cocktail_Down (12203271)	0.59	5.1
ExtraGamma	gamma_PT > 3300	PhiG_Down (13102202)	0.65	1.25
ExtraKs0	Ks0_PT > 1200 && Ks0_MINIPCHI2 > 5	KstIsoG_Down (12203303)	0.67	1.07
ExtraLambda	Lambda_PT > 1200&& Lambda_MINIPCHI2 > 5	LambdaPG (15102307)	0.61	0.93
ExtraPi0Merged	pi0_PT > 3300	-	-	0.96
ExtraPi0Resolved	pi0_PT > 3300	-	-	0.75

The pi^0 cuts have not been tested on signal because the main decay trees of the lines do not contain any.

In addition, the extra selections for the 4 radinclusive lines are exactly the same, and that will probably not change in the future. However, we don't want to restrict this simmetry by design. How do you suggest we should rearrange the "extra selection" code? Should it be moved to a different file so that the lines just import the filtered selections and add them to extra_outputs?