Add missing radiative inclusive lines

Following !1331 (merged)
Fixed the missing input from the BDT and included the lines:

• BToHHHGamma_Incl_line
• BToHHGamma_GammaToEE_Incl_line
• BToHHHGamma_GammaToEE_Incl_line

Rates:
Current rates on sample upgrade_minbias_hlt1_filtered (50k evts)

Line	Inclusive	Exclusive
Hlt2Topo2BodyDecision *	5.96 +/- 0.3442 kHz	5.72000 +/- 0.3372 kHz
Hlt2RD_BToHHGamma_InclDecision *	1.74 +/- 0.1863 kHz	1.46 +/- 0.1707 kHz
Hlt2RD_BToHHHGamma_InclDecision	1.44000 +/- 0.1695 kHz	1.08 +/- 0.1468 kHz
Hlt2RD_BToHHGamma_GammaToEE_InclDecision	1.0 +/- 0.1413 kHz	0.54 +/- 0.1038 kHz
Hlt2RD_BToHHHGamma_GammaToEE_InclDecision	0.88 +/- 0.1326 kHz	0.38 +/- 0.0871 kHz

*onnly for comparison and to check overlap
Added rate w.r.t. previous: ~2kHz
Rates in the PR tests are a bit lower than here (1.19 instead of 1.74 for HHGamma, so they may be a bit overestimated)

Comments:
BDTs are still the same ones as before, retraining is pending.
Efficiencies may seem low specially for converted photons, but the CanRecoChildren denominator only checks that Kaons and pions are reconstructible so we are not correcting for the non reconstructible electrons.

Efficiencies:
On K1G:

Line	Inclusive
Hlt2RD_BToHHGamma_InclDecision	0.080 +/- 0.010
Hlt2RD_BToHHHGamma_InclDecision	0.066 +/- 0.009
Hlt2RD_BToHHGamma_GammaToEE_InclDecision	0.019 +/- 0.005
Hlt2RD_BToHHHGamma_GammaToEE_InclDecision	0.011 +/- 0.004

On KstG:

Line	Inclusive
Hlt2RD_BToHHGamma_GammaToEE_InclDecision	0.010 +/- 0.003
Hlt2RD_BToHHGamma_InclDecision	0.103 +/- 0.010