Question

Hi I am trying to use the custom Binary Integer division method:
Source: http://www.informit.com/guides/content.aspx?g=dotnet&seqNum=642

public static void DivMod (Int128 dividend, Int128 divisor, out Int128 quotient, out  Int128 remainder)
{
// Determine the sign of the results and make the operands positive.
int remainderSign = 1;
int quotientSign = 1;
if (dividend < 0)
{
    dividend = -dividend;
    remainderSign = -1;
}
if (divisor < 0)
{
    divisor = -divisor;
    quotientSign = -1;
}
quotientSign *= remainderSign;

quotient = dividend;
remainder = 0;
for (int i = 0; i < 128; i++)
{
    // Left shift Remainder:Quotient by 1
    remainder <<= 1;
    if (quotient < 0)
        remainder._lo |= 1;
    quotient <<= 1;

    if (remainder >= divisor)
    {
        remainder -= divisor;
        quotient++;
    }
}

// Adjust sign of the results.
quotient *= quotientSign;
remainder *= remainderSign;
}

• However I have 2 problems:

1) I would like to use it for 32 bit integers not Int128. so I assume that the Int128 should be replaced by int, and the (int i = 0; i < 128; i++) should be replaced by i < 32;. Correct?

2) remainder._lo |= 1 -> this line does not work at all in C#. I suppose it has to do with that custom 128bit int struct they use, and I have no idea what it is meant to do. Can somebody help me out with this one, and translate it so that it works with int32?

EDIT: just to clarify I know what the bitwise operators do, the problem part is this:
remainder._lo. I dont know what this property refers to, and not sure of the purpose of this line, and how it would be translated to an int32?

Answer 1

1. To use it with 32 bit integer (System.Int32) you can replace Int128 with int and the 128 in the for loop with 32 – So this is correct.

2. The _lo property is just the lower 64 bits of the 128 bits number. It is used because the biggest integer type in .NET is 64 bits (System.Int64) – So with 32 bits you can just omit the property:
   remainder |= 1;

If you follow the link you gave in your question and go back a few pages you will find the actual implementation of the Int128 struct. It starts here.