How does stringstream work when converting a double value into a char variable

Question

I saw a post on here asking how to convert a double variable value into a char array. Someone said to just use stringstream but didn't explain why it works. I tried googling but couldn't find any documentation on specifically how it converts it. I was wondering if someone could explain to me how it works. Here is the code I wrote that converts a double variable value into a char array.

#include <iostream>
#include <sstream>
using namespace std;

int main()
{
   double a = 12.99;
   char b[100];
   stringstream ss;

   ss << a;
   ss >> b;
   cout << b; // it outputs 12.99

   return 0;
}

Answer 1

When you do ss << a; you are inserting the double in the stringstream (let's suppose it holds the value in a string ) and so when you run ss >> b; it just copies the string in the char[] char by char.
Now the only point is to convert double to string , thing that can be achieved with a simple algorithm:

std::string converter(double value){
    char digits[] = {'0', '1', '2', '3', '4', '5', '6', '7', '8', '9' };
    bool is_negative = value < 0;
    std::string integer_to_string;
    value =  is_negative ? value * -1 : value; // make the number positive
    double fract = value - static_cast<unsigned int>(value); // fractionary part of the number
    unsigned int integer = static_cast<int>(value); // integer part of the number
    do{
        unsigned int current = integer % 10; // current digit
        integer_to_string = std::string(1, digits[current]) + integer_to_string; // append the current digit at the beginning
        integer = integer / 10; // delete the current digit
    } while(integer > 0); // do over and over again until there are digits
    integer_to_string = (is_negative ? "-" : "") + integer_to_string; // put the - in case of negative
    std::string fract_to_string;
    if(fract > 0) {
        fract_to_string = ".";
        do {
            unsigned int current = static_cast<int>(fract * 10); // current digit
            fract_to_string = fract_to_string + std::string(1, digits[current]); // append the current digit at the beginning
            fract = (fract * 10) - current; // delete the current digit
        } while (fract > 0);
    }
    return integer_to_string + fract_to_string;
}

Keep in mind that this is a very basic conversion and will have a lot of errors due to the instability of the operator- in floating point operation, and so it's a lot unstable, but it's just an example

NOTE: this absolutely to avoid to use in legacy (and actually not only legacy) code, it's just been done as an example, instead you should use std::to_string() with will perform it faster and without any type of error (check this )