The algorithms currently implemented have the following best case scenarios for factorizing:
- Fermat's Test (When two prime numbers are close to each other)
- Pollard's Rho (When one prime factor is much smaller than the other)
- Pollard's p-1 (p&q are prime factors -> p-1 divisble by r!, q-1 not divisible by r!, for all r)
These are common methods used to test if a randomly generated RSA public key with two prime numbers is secure enough in today's standards.
Compared to the implemented algorithms, the algorithms I propose to be added to sympy are the general methods that are considered the fastest known to factor a RSA public key.
I believe it is a great addition to Sympy as it would definitely serve as a complement to the current crypto module, specifically the RSA method.