kasparsklavins / bigint Goto Github PK

I write this code:

#include"bigint.h"
#include<iostream>
using namespace std;
using namespace Dodecahedron;
int main(){
	Bigint a,b;
	cin>>a>>b;
	a-=b;
	cout<<a<<'\n'<<-1%10;
	return 0;
}

I get this output after #9 #ab6c858

IN:
5 6
OUT:
-999999999
-1

Before #9 at #1b840c7, everything is right

IN:
5 6
OUT:
-1
-1

So may "subtraction fix" make subtraction fail?

get 0 to Bigint n from cin.
but n==0 is not true.
code is below.

int main() {
    Dodecahedron::Bigint n;
    while(cin >> n){
        //if(Dodecahedron::to_string(n)=="0"){//ok
        //if(n==n*0){//ok
        if(n==0){//do not work
            break;
        }
        cout << n << "\n";
    }
    return 0;
}

Bigint a("m");

The following code gives a loop.

Hi there!
You clearly need to handle power of zero since in your code you've got endless recursion.

In file bigint.cpp around line 73, I found this code:

Bigint &Bigint::operator+=(Bigint const &b)
{
    if (!b.positive) {
        return *this -= b;
    }
    if (!b.positive && positive) {
        positive = false;
    }

The first if make it returned when !b.positive, then the second if can't be true, right? So it should be dead code

Problem occurs only in debug version. But everything is ok in release. I can't figure out why.

Here is the detail information:

bigint a = 159753;
cout << a.pow(6) << endl;

I found division operation has added in #9e3967c but has removed in #e093d1e. Is it because of bugs or just for the commit #7159412? If it's the second, why not get division operation back? I think division operation is also used widely

Bigint a("-10");
Bigint b("-9");
Bigint c = a + b;

Bigint a("-10");
Bigint b("9");
Bigint c = a + b;

hi,

thank you for this source code,
i am trying to use bigint with a class, but its not working, please help me to solve this problem.

maht_test.zip

regards

#include<iostream>
#include<Bigint.h>
using namespace std;
using namespace Dodecahedron;
int main()
{
    Bigint a("4");
    Bigint b("100000000000000000000000000");
    Bigint c;
    c=a*b;
    cout<< c <<endl;
}

Given Output : 0
Correct Answer : 400000000000000000000000000

It fails when multiplying smaller number with large number, but not vice versa.

typedef long double ldbl;

const int imax = numeric_limits< int >::min();
ldbl xdbl = pow(static_cast< ldbl >(imax), 5);
Dodecahedron::Bigint ibig = imax;
Dodecahedron::Bigint xbig = ibig.pow(5);

imax = -2147483648
xdbl = -4.56719e+046
xbig = 45671926166590716193865151022383844364247891968

The sign is incorrect for xbig.

The same incorrect result is obtained with
Dodecahedron::Bigint xbig = ibig * ibig * ibig * ibig * ibig;

Just a reminder to eventually add test coverage.

how to (XOR, OR, AND...)?

This class has addition ,subtraction and multiplication right now, but just out of curiosity, is there any reason that (integer) division and modulus are missing?

the current code is:

    //Multiplication
    Bigint operator*(Bigint const &);
    Bigint &operator*=(Bigint const &);
    Bigint operator*(long long const &);
    Bigint &operator*=(int const &);

why not be "Bigint &operator*=(long long const &);" ?

In function Bigint::pow there are two indentical recruisive calls, that could be replaced by just one call
Consider changing

Bigint Bigint::pow(int const &power, std::map<int, Bigint> &lookup)
{
    if (power == 1) return *this;
    if (lookup.count(power)) return lookup[power];

    int closestPower = 1;
    while (closestPower < power) closestPower <<= 1;
    closestPower >>= 1;

    //here pow(power / 2, lookup) is called twice
    if (power == closestPower) lookup[power] = pow(power / 2, lookup) * pow(power / 2, lookup);
    else lookup[power] = pow(closestPower, lookup) * pow(power - closestPower, lookup);

    return lookup[power];
}

to

Bigint Bigint::pow(int const &power, std::map<int, Bigint> &lookup)
{
    if (power == 1) return *this;
    if (lookup.count(power)) return lookup[power];

    int closestPower = 1;
    while (closestPower < power) closestPower <<= 1;
    closestPower >>= 1;

    if (power == closestPower)
    {
        auto tmp = pow(power / 2, lookup);
        lookup[power] = tmp * tmp;
    }
    else
    {
        lookup[power] = pow(closestPower, lookup) * pow(power - closestPower, lookup);
    }

    return lookup[power];
}