Thread: iterating through hex numbers in c

Does your hexadecimal number string begin with '0x' or '0X'? If so, you can use the strtoi function to convert it to an integer. Then you can increment the integer, and use sprintf to print the incremented integer back to a string.

Originally Posted by TheBigH

If so, you can use the strtoi

That should be an l (ell) not an i (eye).

strto*:
d = double
f = float
ld = long double
l = long
ll = long long
ul = unsigned long
ull = unsigned long long
imax - convert strings to maximum sized integers
uimax - "


Quzah.

Originally Posted by quzah

That should be an l (ell) not an i (eye).

strto*:
d = double
f = float
ld = long double
l = long
ll = long long
ul = unsigned long
ull = unsigned long long
imax - convert strings to maximum sized integers
uimax - "


Quzah.

Whoops, you're right. I forgot that strtoi is the only one left out of that family of functions. (I saw it years ago somewhere, but it seems to have died).

Originally Posted by tabstop

Probably because it is completely ludicrous. Why are you storing an integer as a string?

Text files. Why do people use them? Beats me, but they always show up here complaining about flat files being bad. :P


Quzah.

You can use it as-is. You just have to do it yourself manually. For example:

Code:

switch( ... )
{
    case '0': ...
    case '1': ...
    ...
    case 'F': ...
}

You also have to think about what happens if the left most value is an F and you need to increment that. Do you grow your string longer, or do you treat it as overflow and apply some rule that you decide upon?


Quzah.

Originally Posted by tabstop

the answer is probably even worse: "we've always used a string".

Why store something in one byte what you can store it in five!

Code:

char *p = "0xFF";
unsigned char byte = 0xFF;

Quzah.

Originally Posted by f.g

the number has to exceed the largest unsigned int that the chip can contain

So you want a built in simple way to work with numbers too big to work with normally? That's pretty silly if you stop and actually think about what you are asking. You are going to need to make your own method of handling this. Like I said, you need to also decide what happens when you are trying to increment something bigger than what your string allows.


Quzah.

Originally Posted by f.g

so quzah, these numbers would not be too large to work with normally, just in this context

Right, by "bigger than what you can do normally" means "on this particular setup, it's bigger than what you can normally do". If the only thing your implementation of C had was char, then trying to manipulate a two-byte value wouldn't be something you can normally do. In your case, your have something with what, 16 bit ints? So trying to use a 32 bit value is bigger than what it can normally do. Thus, you need a string or an array of said type, or whatever you choose.

Now, you've already said you have to use a string and you can't change that, so obviously:

Originally Posted by f.g

thought that there might be a better way.

There isn't a better way, because you just said you had to use a string. What you could do is reverse the order that they are stored in the string, and don't bother storing 0x (or in your case x0 if it was reversed).

"0123456789abcdef" -> "fedcba907654321"

Code:

for( p = string; *p; p++ )
{
    switch( tolower( *p ) )
    {
        case '0':
        case '1':
        case '2':
        case '3':
        case '4':
        case '5':
        case '6':
        case '7':
        case '8':
        case 'a':
        case 'b':
        case 'c':
        case 'd':
        case 'e': (*p)++; return;
        case '9': *p = 'a'; return;
        case 'f' : *p = '0'; /* now you are carrying, if you run out of string, you need to do something */
    }
}

That looks about right.


Quzah.