Split B2ppbarhh into four different channels

patch generated by https://gitlab.cern.ch/lhcb/Moore/-/jobs/37133956

Using a bandwith snippet: https://gitlab.cern.ch/-/snippets/2694 the size of obtained data sample for all three lines is 2.0MB, while 28 events are triggered based on 50K events. The test sample is "expected_2024_min_bias_hlt1_filtered" ** Calculated bandwidth of all three divided lines 0.0478 GB/s **

Consider the large rate caused by noPID lines, #717 (comment 7667283)

then try with split them and perform PID for h. Some data following. The rate at least decrease to one third while the effect on efficiency is acceptable. Rate with PID

B→pppipi: (0.023000000 +- 0.000999995)%
B→ppKpi: (0.02100000 +- 0.00458209)%
B→ppKK:(0.007000000 +- 0.00264566)%

noPID:

(0.1800000 +- 0.0423882)% 

Efficiency with PID

B→ppkpi:  (0.6471260 +- 0.110622)%, B→pppiK: (0.6471260 +- 0.110622)%
(B→ppKpi+pppiK: ( 1.199086 +- 0.150162)%)
B→pppipi: (0.9229349 +- 0.145254)%  (0.7152746 +- 0.128007))%
B→ppKK: (0.9505703 +- 0.149582)%
···

without PID

B→pppipi:( 1.338256 +- 0.174542)% B→ppKK: ( 1.259506 +- 0.171913)% B→ppKpi:( 1.332318 +- 0.158178)%