Newer translation tables; tables with no dedicated stops

[Artoria2e5]

src/geneticcode.cc currently only goes up to table 24. There are now a handful of new tables in gc.prt.

Of particular note are tables 27, 28, 31, which have no dedicated stop codons. For the ciliates of 27 and 28, there is a messy position-dependent behavior (https://doi.org/10.1016/j.cell.2016.06.020) that doesn't bode well with a fixed stop codon probability. https://doi.org/10.1093/molbev/msad064 explicitly mentions that the lack of support is messing with BUSCO predictions of completeness. I don't know how 31 decides on whether to stop.