Add support for storing 16 byte types in Simple8b. For this to work we need to have enough leading and/or trailing zeros to be able to fit the value in our Simple8b buckets.

We want to generalize the Simple8b class by making it a template with the two supported combinations being uint64_t and uint128_t.

Integrate uint128_t from boost multiprecision (cpp_int backend). We also need to make functions that can calculate leading and trailing zeros on this type.