Decimal to Binary, strange output

Question

I wrote program to convert decimal to binary for practice purposes but i get some strange output. When doing modulo with decimal number, i get correct value but what goes in array is forward slash? I am using char array for being able to just use output with cout <<.

// web binary converter: http://mistupid.com/computers/binaryconv.htm

#include <iostream>
#include <math.h>
#include <malloc.h> // _msize
#include <climits>
#define WRITEL(x) cout << x << endl;
#define WRITE(x) cout << x;

using std::cout;
using std::endl;
using std::cin;

char * decimalToBinary(int decimal);
void decimal_to_binary_char_array();

static char * array_main;

char * decimalToBinary(int decimal) // tied to array_main
{
   WRITEL("Number to convert: " << decimal << "\n");
   char * binary_array;
   int t = decimal,     // for number of digits
       digits = 0,      // number of digits
       bit_count = 0;   // total digit number of binary number
   static unsigned int array_size = 0;
   if(decimal < 0) { t = decimal; t = -t; }  // if number is negative, make it positive
   while(t > 0) { t /= 10; digits++; }       // determine number of digits
   array_size = (digits * sizeof(int) * 3);  // number of bytes to allocate to array_main
   WRITEL("array_size in bytes: " << array_size);
   array_main = new char[array_size];
   int i = 0;  // counter for number of binary digits
   while(decimal > 0)
   {
      array_main[i] = (char) decimal % 2 + '0';
      WRITE("decimal % 2 = " << char (decimal % 2 + '0') << "  ");
      WRITE(array_main[i] << "  ");
      decimal = decimal / 2;
      WRITEL(decimal);
      i++;
   }
   bit_count = i;
   array_size = bit_count * sizeof(int) + 1;
   binary_array = new char[bit_count * sizeof(int)];
   for(int i=0; i<bit_count+1; i++)
      binary_array[i] = array_main[bit_count-1-i];
   //array_main[bit_count * sizeof(int)] = '\0';
   //WRITEL("\nwhole binary_array: "); for(int i=0; i<array_size; i++) WRITE(binary_array[i]); WRITEL("\n");
   delete [] array_main;
   return binary_array;
}

int main(void)
{

   int num1 = 3001; 
           // 3001    = 101110111001
           // 300     = 100101100
           // 1000    = 1111101000
           // 1200    = 10010110000
           // 1000000 = 11110100001001000000
           // 200000  = 110000110101000000
   array_main = decimalToBinary(num1);
   WRITEL("\nMAIN: " << array_main);

   cin.get();
   delete [] array_main;
   return 0;
}

The output:

Number to convert: 3001

array_size in bytes: 48
decimal % 2 = 1  /  1500
decimal % 2 = 0  0  750
decimal % 2 = 0  0  375
decimal % 2 = 1  1  187
decimal % 2 = 1  /  93
decimal % 2 = 1  1  46
decimal % 2 = 0  0  23
decimal % 2 = 1  1  11
decimal % 2 = 1  1  5
decimal % 2 = 1  1  2
decimal % 2 = 0  1  1
decimal % 2 = 1  1  0

MAIN: 1111101/100/

What are those forward slashes in output (1111101/100/)?

Answer 1

I haven't tried to analyze all your code in detail, but just glancing at it and seeing delete [] array_main; in two places makes me suspicious. The length of the code makes me suspicious as well. Converting a number to binary should take about two or three lines of code; when I see code something like ten times that long, I tend to think that analyzing it in detail isn't worth the trouble -- if I had to do it, I'd just start over...

Edit: as to how to do the job better, my immediate reaction would be to start with something on this general order:

// warning: untested code.
std::string dec2str(unsigned input) {
    std::deque<char> buffer;
    while (input) {
        buffer.push_front((input & 1)+'0');
        input >>= 1;
   }
   return std::string(&buffer[0], &buffer[0]+buffer.size());
}

While I haven't tested this, it's simple enough that I'd be surprised if there were any errors above the level of simple typos (and it's short enough that there doesn't seem to be room to hide more than one or two of those, at very most).

Answer 2

Your problem is here:

array_main[i] = (char) decimal % 2 + '0';

You are casting decimal to char and it is swiping off the high-order bits, so that in some cases it becomes negative. % applied to a negative number is negative, hence you get one character before 0 in the ASCII chart, which is /.

I would also like to say that I think your macros WRITEL and WRITE qualify as preprocessor abuse. :-)

Answer 3

It must be array_main[i] = (char) (decimal % 2 + '0'); (note the parentheses). But anyway, the code is horrible, please write it again from scratch.