Factoring large integers is a computationally very difficult problem, and there is no general factorization algorithm that is capable of factoring arbitrary integers with more than about 130 -- 150 decimal digits on currently existing computers, but however, there are methods (not algorithms in the sense that it is guaranteed that they will give the desired result after a finite number of steps) for factoring integers with prime factors being far beyond the range where trial division is feasible.
One important class of such methods is based on exponentiation in suitably chosen groups acting on subsets of the k-fold cartesian product of the set of residue classes (mod n), where n denotes the number to be factored. Representatives of this class are the Elliptic Curves Method (ECM), Pollard's p-1 and Williams' p+1. These methods have the important property that they find smaller factors usually considerably faster than larger ones (but with the drawback of being slower for large factors).
The other important class consists of the so-called factor base methods. These methods compute factorizations of perfect squares (mod n) over an appropriately chosen factor base (a set of small prime numbers, or of prime ideals in the case of the Generalized Number Field Sieve) and use them to determine a pair of integers (x,y), such that x2 and y2 are congruent (mod n), but ±x and ±y are not. In this situation, taking Gcd(n,x-y) will yield a non-trivial factor of n. Representatives of this class are the Continued Fraction Algorithm (CFRAC), the Multiple Polynomial Quadratic Sieve (MPQS) and the already mentioned Generalized Number Field Sieve (GNFS), which has got the asymptotically lowest complexity of all factoring methods known today (in respect of ``difficult'' numbers, of course), with the drawback of being superior to the MPQS only for numbers with more than 100 decimal digits, say, which is probably not within the feasible range of such a function implemented in GAP. The first two methods are implemented in this package.
The only method which I am aware of (except of the ``naive'' ones like trial division and some ``historical'' methods) that does not fit into one of the two above-mentioned classes is Pollard's Rho, which is already implemented in the GAP Library.
In respect of the current state-of-the-art in integer factorization, see for example the respective web pages of the RSA Laboratories, e.g. http://www.rsasecurity.com/rsalabs/challenges/factoring/numbers.html
I would like to thank Bettina Eick and Steve Linton for their support and many interesting discussions.
I would be grateful for any bug reports, comments or suggestions,
Stefan Kohl
FactInt manual