I came back to my float-to-decimal implementation, fixed some inefficiencies, made is C-compatible, and over all easier to use as a standalone library, in case someone is interested.
https://github.com/eternalStudent/float_to_decimal/