Feat/transform bigint limbsize (BitVM#198)

* add transform_limbsize fn

* rename field for clarity

* add a simple manual test. to add more tests.

* add manual tests for all zeros and ones

* add should panic tests for the limbsizes

* add property testing for gen_transform_steps
add assert to ensure the minimum number of bits for transform to be 32

* fix docstring for extract_digits

* add test for random transformations for U256

* add test for extract_digits
refactor extract_digits from inside BigIntImpl

* fix 'assertion msgs'

* use Vec::with_capacity instead of Vec::new
fix typo in fn name
use pop on reversed vec instead of remove at index 0

Author: mukeshdroid