michaeljclark / bignum

C++ bignum with support for arbitrary precision integer arithmetic.

bignum

C++ bignum with support for arbitrary precision integer arithmetic.

• wideint arbitrary fixed width signed two's complement or unsigned integer.
• bignum arbitrary variable width signed two's complement or unsigned integer.
• supports static or dynamic width.
• supports arbitrary precision signed and unsigned arithmetic.
• supports operator overloads for C++ math, logical and bitwise operators.
• multiplication and division algorithm from Hacker's Delight.
• divide and conquer algorithm for radix 10 conversion to string.

Build

The bignum library standard build process uses cmake.

GCC or Clang on Linux or macOS

mkdir build
cd build
cmake -G "Unix Makefiles" ..

Microsoft Visual Studio 2019 on Windows

mkdir build
cd build
cmake -G "Visual Studio 16 2019" ..

Examples

Example wideint arithmetic

#include <wideint.h>

typedef wideint<256> int256_t;

int main()
{

	int256_t a = 2;
	int256_t b = a | 3;
}

Example code for C++ bignum arithmetic

#include <iostream>

#include <bignum.h>

int main()
{
	bignum x = 2;
	bignum y = x.pow(4096);
	std::cout << y.to_string(10) << std::endl;
}

Equivalent code using GMP (GNU Multiple Precision Arithmetic Library)

#include <iostream>

#include <gmp.h>

int main()
{
	mpz_t result, base;

	mpz_init(base);
	mpz_init(result);
	mpz_set_ui(base, 2);
	mpz_pow_ui(result, base, 4096);

	std::cout << mpz_get_str(NULL, 10, result) << std::endl;

	mpz_clear(base);
	mpz_clear(result);
}

bignum API

bignum implements the following operators and methods:

• bignum(unsigned int n)
• bignum(std::string str, size_t radix)
• bignum& operator+=(const bignum &operand)
• bignum& operator-=(const bignum &operand)
• bignum& operator<<=(int shamt)
• bignum& operator>>=(int shamt)
• bignum& operator&=(const bignum &operand)
• bignum& operator|=(const bignum &operand)
• bignum operator+(const bignum &operand) const
• bignum operator-(const bignum &operand) const
• bignum operator<<(int shamt) const
• bignum operator>>(int shamt) const
• bignum operator&(const bignum &operand) const
• bignum operator|(const bignum &operand) const
• bool operator==(const bignum &operand) const
• bool operator<(const bignum &operand) const
• bool operator!=(const bignum &operand) const
• bool operator<=(const bignum &operand) const
• bool operator>(const bignum &operand) const
• bool operator>=(const bignum &operand) const
• bool operator!() const
• bignum operator*=(const bignum &operand)
• bignum operator/=(const bignum &divisor)
• bignum operator%=(const bignum &divisor)
• bignum operator*(const bignum &operand) const
• bignum operator/(const bignum &divisor) const
• bignum operator%(const bignum &divisor) const
• void from_string(std::string, size_t radix = 0 /autodetect/)
• std::string to_string(size_t radix = 10) const

bignum's Issues

Any Benchmarks?

Hello,

I am seeking a fast arbitrary floating point and integer package for a Windows x64 project.

Bignum seems to be easy to use as compared to GMP/MPZ but I am wondering if it has any benchmarks against these or native calculations to get an idea of speed relevance as well.

Any thoughts?
Thanks and have a great day