What is the difference between *variable and *variable-'0'?

Question

I’m writing a Sudoku game with C and ncurses, below follows the code of a function that receives a pointer to a variable (ch) which stores typed key with function getch(), if the number already exists horizontally, vertically or in the block returns false, if not true.

Doubt:

• In the code, if you replace *ch-'0' for *ch, it stops working correctly, returns false only in numbers added by user, ignoring the standards, the behavior also changes according to the cursor position. Why? What is the difference?

Code:

bool
num(int *ch)
{
    // g.y e g.x Armazenam a posição do cursor no array
    // g.r_board Array de referência
    if (g.r_board[g.y][g.x] > 0)
    {
        hide_banner(); // Esconde frase da tela
        show_banner("Número fixo"); // Mostra uma frase na tela
        show_cursor(); // Move o cursor pra posição original
        return false;
    }

    // Verifica se tem número igual na horizontal
    // g.board Array que armazena números do jogo
    for (int i = 0; i < 9; i++)
        if (g.board[g.y][i] == (*ch-'0'))
        {
            hide_banner();
            show_banner("Número igual horizontal");
            show_cursor();
            return false;
        }

    // Verifica se tem número igual na vertical
    for (int i = 0; i < 9; i++)
        if (g.board[i][g.x] == (*ch-'0'))
        {
            hide_banner();
            show_banner("Número igual vertical");
            show_cursor();
            return false;
        }

    // Verifica se tem número igual dentro do bloco
    for (int i = (g.y/3)*3, j = i+3; i < j; i++)
        for (int k = (g.x/3)*3, l = k+3; k < l; k++)
            if (g.board[i][k] == (*ch-'0'))
            {
                hide_banner();
                show_banner("Número igual bloco");
                show_cursor();
                return false;
            }

    return true;
}

Answer 1

This is the same thing as doing

*ch - 48

I put in the Github for future reference.

already '0' is a character displaying text of a zero and equivalent to code 48 in decimal of ascii table. He’s not worth 0 as you might think at first. He used the type char because it is more semantic for this code, but arithmetic is defined in binary, no matter if it has a number a character in its source code.

If the code asks for a subtraction and you remove the result will go wrong.

It’s taking the received code and subtracting the character from zero to make the value reset, that is if you typed the character 0, you want the number 0, then if the character 0 is worth 48, you have to subtract 48 to reach 0, which is what that expression does most simply, you don’t even need to know that the character 0 is worth 48 in the ASCII table. The same goes for the other characters, if you type the character 1, of course it will take the code 49, minus 48 of the character 0 that is there in the expression and will give the number 1.