Different way of doing things

I began working out a problem and ran into a brain freeze and I eventually solved it, but it doesn't look like what I originially had in mind. The results are perfect, I will keep the code, but for some reason, I thought the code would look slightly different...but only in the part where I take the value of 500.

Is there a "better" way of doing this? I'm sure it's obvious, but basically I want to separate the values city and id from the string, "new york (500)".


Appreciate all the help!

Code:

#include <stdio.h>
#include <string.h>

int main()

{


        int offset; 
     	char *position; 
     	char *str = "New York (500)";
     	char *search_str = "(" ;
	char city[110];
	char id[110]; // this represents the 500 number
                int IDSIZE = 3;


               position = (char *)strstr(str, search_str);

               offset = (int)(position - str);

               position = (char *)strncpy(city, str, offset);
		
               printf("String is %s , Offset = %d , City is: %s", str, 
               offset, city);
		       
               str = (str + (offset + 1));
               
               position = (char *)strncpy(id, str, IDSIZE);		           
               printf("Branch id is:  %s", id);
                   
               return 0;

}

Well, you could do it like this:

Code:

#include <stdio.h>
#include <string.h>

#define IDSIZE 3

int main(void)
{
  char *str = "New York (500)";
  char city[110];
  char id[IDSIZE + 1];
  char *p;

  p = strchr(str, '(');

  strncpy(city, str, p - str);
  city[p - str] = '\0';

  strncpy(id, p + 1, IDSIZE);
  id[IDSIZE] = '\0';

  printf("City: %s, Branch ID: %s\n", city, id);

  return 0;
}

itsme@itsme:~/C$ ./cityid
City: New York , Branch ID: 500

But it makes some dangerous assumptions and the city string has a trailing space. It also isn't very strict on what the original string needs to look like. It limits the ID to 3 digits and doesn't even care if the string is in the format "city (id)".

This one is a lot better I think:

Code:

#include <stdio.h>
#include <string.h>

int main(void)
{
  char *str = "New York (500)";
  char city[110];
  char id[110];
  char *p, *q;
  int tocopy;

  if(!(p = strchr(str, '(')))
  {
    puts("String syntax validation failed -- Expected '('");
    return 1;
  }

  // Get rid of space(s) between city name and ID field
  for(q = p - 1;q >= str && *q == ' ';--q)
    ;
  q++;

  // Don't forget bounds checking
  tocopy = q - str >= sizeof(city) ? sizeof(city) : q - str;
  strncpy(city, str, tocopy);
  city[tocopy] = '\0';

  for(++p, q = id;q - id < sizeof(id) && *p != ')';++p, ++q)
    *q = *p;
  *q = '\0';

  if(*p != ')')
  {
    puts("String syntax validation failed -- Expected ')'");
    return 1;
  }

  printf("City: %s, Branch ID: %s\n", city, id);

  return 0;
}

itsme@itsme:~/C$ ./cityid2
City: New York, Branch ID: 500

Yet another possible way to do this.

Code:

#include <stdio.h>
#include <string.h>

int main(void)
{
   static const char *line[] = {"New York (500)","Los Angeles(400)"};
   char city[32];
   int id;
   size_t i;
   for ( i = 0; i < sizeof line / sizeof *line; ++i )
   {
      if ( sscanf(line[i], "%31[^(](%d)", city, &id) == 2 )
      {
         size_t last = strlen(city) - 1;
         if ( city[last] == ' ' )
         {
            city[last] = '\0';
         }
         printf("city = \"%s\", id = %d\n", city, id);
      }
   }
   return 0;
}

/* my output
city = "New York", id = 500
city = "Los Angeles", id = 400
*/

And another (well, not really).

Code:

#include <stdio.h>

int main(void)
{
   static const char *line[] = {"New York  (500)","Los Angeles(400)","Dullsville (300)"};
   char city[32];
   int id, last;
   size_t i;
   for ( i = 0; i < sizeof line / sizeof *line; ++i )
   {
      if ( sscanf(line[i], "%31[^(]%n(%d)", city, &last, &id) == 2 )
      {
         while ( city[--last] == ' ' )
         {
            city[last] = '\0';
         }
         printf("city = \"%s\", id = %d\n", city, id);
      }
   }
   return 0;
}

[edit=much later]The following is not true.

It's subtle, but there is UB on this line.

Code:

if ( sscanf(line[i], "%31[^(]%n(%d)", city, &last, &id) == 2 )

[edit2]C99 has the following sentence in its description of the n directive.

If the conversion specification includes an assignment suppressing character or a field width, the behavior is undefined.

And apparently C89 does not have that same sentence.

It'sMe, Dave, Anonymouse:

Thanks a bunch, gentlemen (and ladies if I've made a sexist assumption)! I appreciate it. As I said, I'm just picking up the language, and it helps to see things done differently. I really do appreciate it!

>C99 has the following sentence in its description of the n directive.
Okay, that doesn't explain your statement about undefined behavior though. %n is a valid conversion specification.

Originally Posted by Prelude

>C99 has the following sentence in its description of the n directive.
Okay, that doesn't explain your statement about undefined behavior though. %n is a valid conversion specification.

Hm. Perhaps I've been reading that wrong. I read it as, "If you've got an assignment suppression or a field width" -- in this case the field width for the string -- "and you also use the %n to determine the number of characters consumed, then the behavior is undefined". Should it be read, "You can't do "%*n" or "%5n"?"

>Should it be read, "You can't do "%*n" or "%5n"?"
Yes. The other interpretation is unnecessarily restrictive.

I guess the question is: where is the assignment suppressing character or field width? Because I don't see either, but then again, my eyes aren't that good.

>where is the assignment suppressing character or field width?
That's the misinterpretation. Dave was looking at the field width here:

Code:

sscanf(line[i], "%31[^(]%n(%d)", city, &last, &id)

But the standard only talks about undefined behavior when there's a field width here:

Code:

sscanf(line[i], "%31[^(]%n(%d)", city, &last, &id)

Originally Posted by Prelude

>Should it be read, "You can't do "%*n" or "%5n"?"
Yes. The other interpretation is unnecessarily restrictive.

My bad. And cool -- I liked it that way better!

Whatever day I had been Groups Googling, I thought it was saying that if you are discarding characters, or limiting the field width, that counting them was an issue. Anyways, thanks for the clarification.

>I thought it was saying that if you are discarding characters, or
>limiting the field width, that counting them was an issue.
Discarding characters would be a minor issue, but assignment suppression doesn't discard, it simply fails to assign a matched specification with a variable. Limiting the field width also wouldn't cause a problem because in the end that doesn't affect the character count, it just limits the number of characters from the stream that can be counted.