class AliTOFT0: public TTask

 This is a TTask that made the calculation of the Time zero using TOF.
 Description: The algorithm used to calculate the time zero of
 interaction using TOF detector is the following.
 We select in the MonteCarlo some primary particles - or tracks in
 the following - that strike the TOF detector (the larger part are
 pions, kaons or protons).
 We choose a set of 10 selected tracks, for each track You have the
 length of the track when the TOF is reached (a standard TOF hit
 does not contain this additional information, this is the reason
 why we implemented a new time zero dedicated TOF hit class
 AliTOFhitT0; in order to store this type of hit You have to use the
 AliTOFv4T0 as TOF class in Your Config.C. In AliTOFv4T0 the
 StepManager was modified in order to fill the TOF hit branch with
 this type of hits; in fact the AliTOF::AddT0Hit is called rather
 that the usual AliTOF::AddHit), the momentum at generation (from
 TreeK) and the time of flight given by the TOF detector.
 (Observe that the ctor of the AliTOF class, when the AliTOFv4T0
 class is used, is called with the "tzero" option: it is in order
 create the fHits TClonesArray filled with AliTOFhitT0 objects,
 rather than with normal AliTOFhit)
 Then Momentum and time of flight for each track are smeared
 according to known experimental resolution (all sources of error
 have been token into account).
 Let consider now only one set of 10 tracks (the algorithm is the
 same for all sets).
 Assuming the (mass) hypothesis that each track can be AUT a pion,
 AUT a kaon, AUT a proton, we consider all the 3 at 10 possible
 cases.
 For each track in each (mass) configuration
 (a configuration can be
 e.g. pion/pion/kaon/proton/pion/proton/kaon/kaon/pion/pion)
 we calculate the time zero (we know in fact the velocity of the
 track after the assumption about its mass, the time of flight given
 by the TOF, and the corresponding path travelled till the TOF
 detector). Then for each mass configuration we have 10 time zero
 and we can calculate the ChiSquare for the current configuration
 using the weighted mean over all 10 time zero.
 We call the best assignment the mass configuration that gives the
 minimum value of the ChiSquare.
 We plot the weighted mean over all 10 time zero for the best
 assignment, the ChiSquare for the best assignment and the
 corresponding confidence level.
 The strong assumption is the MC selection of primary particles. It
 will be introduced in the future also some more realistic
 simulation about this point.

 Use case:
 root [0] AliTOFT0 * tzero = new AliTOFT0("galice.root")
 Warning in <TDatabasePDG::TDatabasePDG>: object already instantiated
 root [1] tzero->ExecuteTask()
 root [2] tzero->ExecuteTask("tim")
             // available parameters:
             tim - print benchmarking information
             all - print usefull informations about the number of
                   misidentified tracks and a comparison about the
                   true configuration (known from MC) and the best
                   assignment

-- Author: F. Pierella