MapToScalarField() added for Banderwagon points

[advaita-saha]

Fixes #277

Refere the issue #277 for the detailed specs

• Write the mapToBaseField() function for computing $\frac{x}{y}$ and returning as an $F_p$ element
• MapToScalarField() for converting the $F_p$ element to $F_r$, in between it is done by first converting $F_p$ -> Bytes -> $F_r$
• Tests for mapToBaseField() & MapToScalarField()
• Commenting and linking specs documents in the code
• Adding the BatchMapToField(), this does the same operations but for a batch of Banderwagon points together
• Tests for the Batch operations
• Final checks, fixes and commenting the codebase

[mratsim]

OK overall,

I would change the array input to use openArray instead because:

• This is more flexible, many times the input size isn't known at compile-time
• Code size, this creates a proc per N (monomorphization), which isn't worth it for such a slow code.

batchAffine has an example on how to do this properly:

• tag the function {.noInline.} so that the allocStack doesn't lead to stack overflow if the function is called in a loop and inlined for some reason (LTO ...)
• if copy from a temporary, use for loop, it will be optimized by the compiler. (alternatively copyMem https://nim-lang.org/docs/system.html#copyMem%2Cpointer%2Cpointer%2CNatural )

[advaita-saha]

I have also added the tests to nimble file

[mratsim]

Perfect, thanks!