Add a new tuple tool that emulates TOS-matching on MC particles

FYI @rmatev @sstahl @mvesteri

For HLT efficiencies studies using the HltEfficiencyChecker package (see MooreAnalysis project), we'd like to be able to emulate the TOS matching of offline reconstruction hits to the hits of the selected tracks in the SelReports that is usually done in TupleToolTISTOS via DaVinci. However, since we're writing tuples straight out of Moore, we have no offline reconstruction to match the SelReports to, and so instead we emulate this by matching the LHCbIDs of MC particles to the selected tracks from the SelReports. This tuple tool does this, giving us the ability to write out emulated TOS decisions to a tuple, and we can then calculate "TOS" efficiencies.

TODO:

• TOS match in each subdetector, as detailed in the Run I trigger performance paper?
• MR in MooreAnalysis to print out/make some plots of TOS efficiencies.
• Tidy up/make small attempt to make the code more performant.
• Decide whether to use lhcbIDsFlattened() or the function I have written (appear to do the same thing, with the difference of 1 if statement that I am yet to trigger...) -> added an assertion to catch when they give different results.
• Figure out why/if it's a problem that the fill method seems to get called twice / MCParticle.

Required by MooreAnalysis!33 (merged).