Create a mathlib for ereal and efloat

[Ravenwater]

ereal is a multi-component adaptive precision floating-point, and efloat is a multi-digit adaptive precision floating-point. They are the Oracle number systems within the Universal ecosystem.

We need a set of arbitrary precision function approximations of the standard mathlib.