Add cvfpufit, pfopufit and mhtpufit chains to the menu

This MR adds three new algorithms to the trigger code. They are covered in an overview talk to the TGM here and in more detail here.

I've added 4 chains and their corresponding output containers

• one pfopufit chain
• one cvfpufit chain
• two mhtpufit chains, one with EMTopo jets and one with EMPFlow jets, both using the subjesgscIS calibrations

A note on optimisation: there are a fair few variants that we could test here beyond these four chains. One option would be to try and implement the configuration for these in the menu code and put a few options in the menu, however I think in the long run this will bloat the trigger EDM file which I'd prefer to avoid. I think if we produce some AODs with the AODFULL option then everything we need to reproduce these algorithms offline is output as part of the AODs*. This would then mean that with sufficient stats we can redo the studies linked above with the online objects.

Another point for future optimisation is that the CVF calculation does a very similar operation to part of the PFO reconstruction, namely extrapolating tracks to the calorimeter and associating them to clusters. Long term it may be a good idea to see if we can share this information, but I don't think it's worth doing the necessary refactoring until we know if we will actually use cvfpufit.

Tagging @bcarlson, @okumura, @jpanduro, @bernius, @jmasik and @okuprash given this involves menu and EDM changes.

---

• The only exception to this is the xAOD::EventShape container used in the jet area correction which isn't written out. This is used inside the mhtpufit algorithm. One option might be to add this to the output container as auxdata.