Just a toy project of big int, even without division(now).

Contains a non-trivial allocator(though very memory-consuming),a karatsuba implementation,and a balanced toom-3 implementation. Correctness hasn't been fully tested; but a squaring is performed in the test.cpp with correct output anyway, though the testcase is very weak (due to convenience).

(Yesterday for that testcase running karatsuba is roughly nlogn..)

FFT multiplication. Higher-order toom-cook&inbalanced splitting. Division by inverse.

The substraction routine is very tricky, only being correct under some case(most usable case).

Division by 3 is implemented with inverse element on char[2] field.

Since then I made many commits but no futher work is done for the bigint part..(It this repo just a grocery storage for me?I don't know either.It is better for management and searching than gist and ubuntu pastebin anyway.)

I added a template function library(for computing within compile time), a number theory template library coded in Python(Python is very convenient in multiprecision integer handling,okay I am just petrified of multiprecision division that is a excruciating task for me..), and zemmud, some code for RNGs..(zemmud is not a existing word but it just looks well..). zemmud should be compiled with -std=c++11 because I used variadic template..(And zemmud itself contains a simplified version of CPPMPL so it can be built independently)