Thread: increment IP addresses - newbie Q.

>sounds awfully complex for such a simple task...
IP addressing doesn't exactly make it easy though. Of course, you could just convert the entire address to an array of integers and work with that:

Code:

int i = 0;
int j;
int ip[4];
char hostaddr[16];
char *snip;
  
/* Put an ip string in hostaddr */
for (snip = strtok(hostaddr, "."); snip != NULL; snip = strtok(NULL, "."))
  ip[i++] = atoi(snip);

for (j = 0; j < 4; j++)
  printf("%d%s", ip[j], (j == 3) ? "\n" : ".");

You could do that too, there are a number of approaches to this problem.

There's an even easier way.

Ip's are stored as unsigned long integers. That's 4 bytes of data.
To manipulate these bytes 'easily', the creators of the socket specification came up with this ridiculous structure:

Code:

 struct in_addr 
{
  union 
 {
     struct 
   { 
     u_char s_b1,
                 s_b2,
                 s_b3,
                 s_b4; 
   } S_un_b;

     struct 
   { 
     u_short s_w1,
                  s_w2; 
   } S_un_w;

    u_long S_addr;

 } S_un;
};

Basically just a container for an unsigned long of course.

So to access byte[0] we 'simply' navigate to in_addr::S_un.S_un_b.s_b1 (!)

That's a bit complicated for such a trivial task!

So here's a simpler approach:

Code:

unsigned char 
 get_subnet(unsigned long address, int which)
{
 return ((unsigned char*)&address)[which];
}


unsigned long 
 set_subnet(unsigned long * address, int which, unsigned char what)
{
 ((unsigned char*)address)[which] = what;
 return *address;
}

Of course, the above could be coded as a macro, if necessary.

So here's an example of putting this simple little scheme to work:

Code:

#define MAX 256

int main()
{
 char buff[MAX];
 
 unsigned long address;
 
 int byte;
     
 printf("Enter a valid ip address > ");
 
 fgets(buff, MAX, stdin);
 
    if((address = inet_addr(buff)) != INADDR_NONE)
   {
    printf("Which subnet to manipulate? [0-3] > ");
    
       if((byte = atoi(fgets(buff, MAX, stdin))) < 4)
      {
          for(int i = get_subnet(address, byte)+1; i < MAX; ++i)
         {
          set_subnet(&address, byte, i);
          
          printf("%s\n", inet_ntoa((in_addr&)address));
         }         
      }      
   }
 system("PAUSE");
}

>> int abc = sscanf(ipadd, "%3u.%3u.%3u.%3u%c", &b1, &b2, &b3, &b4, &c);


That's not really necessary. The inet_addr() function takes a dotted decimal string and converts it to an unsigned long.

Then you have the inet_ntoa() function that does just opposite (though you have to cast the unsigned long to an in_addr).

So between the two, you can convert back and forth rather easily, ie:

printf("%s", inet_ntoa(inet_addr(inet_ntoa(inet_addr("127.0.0.1 ")))));

- prints "127.0.0.1", of course.



Anyway, I hope that helps.

Happy coding.