Lalinference o2+spectral

This is code that is being used for the fast-track EOS paper. These updates allow for 1) the calculation of a new EOS model, which is a spectral decomposition of the adiabatic index, within lalsismulation, and 2) lalinference sampling in these parameters. While I think the main components of these updates are ready to go, I marked it as a WIP, since I'm assuming folks will have opinions regarding where is best to store the new functions I added, and things like that.

I'm assuming this will need two reviewers, one for the lalinference changes (@vivien? @john-veitch? matthew?) and one for the lalsimulation changes (@jolien-creighton? richard?)

Anyway, I wanted to get this up so that we can start thinking about how best to get this implemented.