Thread: performing range checks on variable values

I mean checking if the value of variable is within the permissible range other wise simply report error message and quit. For eg. int is always between INT_MIN and INT_MAX. Is there any elegant way to do this ?

If the intent of the question is checking if input data fits in an int, then ....

strtol() [for conversion of a string to a long] returns zero if the conversion is not possible. If the data in the string cannot be represented in a long (eg value too big) the function returns LONG_MIN or LONG_MAX (depending on sign) and sets errno to the value of ERANGE.

You might also wish to look up functions like atoi() which, except for behaviour on error, are equivalent to strtol() but simpler to use.

Similar functions exist for converting a string to other types.

Originally Posted by grumpy

If the intent of the question is checking if input data fits in an int, then ....

strtol() [for conversion of a string to a long] returns zero if the conversion is not possible. If the data in the string cannot be represented in a long (eg value too big) the function returns LONG_MIN or LONG_MAX (depending on sign) and sets errno to the value of ERANGE.

You might also wish to look up functions like atoi() which, except for behaviour on error, are equivalent to strtol() but simpler to use.

Similar functions exist for converting a string to other types.

what about underflow, overflow and divide by zero errors which often happen ? Should I write a library of my own or is there some C implementation out there that I can use.

Originally Posted by broli86

what about underflow, overflow and divide by zero errors which often happen ? Should I write a library of my own or is there some C implementation out there that I can use.

Check for that possibility before you do the operation.
Ex. make sure the denominator is != 0

Originally Posted by broli86

what about underflow, overflow and divide by zero errors which often happen ?

Those things do not happen with functions like strtol() -- these functions catch the overflow by parsing the input string, and then return a values and/or assign error status if required.

It would be a fairly oddball parsing technique that could produce a divide by zero error when attempting to interpret contents of a string as an integral value.

Originally Posted by grumpy

Those things do not happen with functions like strtol() -- these functions catch the overflow by parsing the input string, and then return a values and/or assign error status if required.

Oh that's great. In that I case I will be safer with strtol than using scanf and scanf looks very ugly for some reason.

Originally Posted by grumpy

It would be a fairly oddball parsing technique that could produce a divide by zero error when attempting to interpret contents of a string as an integral value.

So you are saying that using strtol can cause divide by zero error ?

Originally Posted by laserlight

No, grumpy stated that "those things do not happen with functions like strtol()", which is true.

Ok . What about when reading double precision data from a file ? Is it safer to use fgets or fscanf ? The file format is as follows:

3.456789 0.000000 -1.234567
4.567889 0.000001 -10.22222
..................................................
..................................................

If I use fgets, I presume I will be reading 3 strings and then converting each to double precision using strtol ? But I think fgets only allows one string at a time.

You may want to read up on strtol

http://cboard.cprogramming.com/showthread.php?t=104680

The intended purpose is for conversion. I would not base my calculations for boundary checking on this.

I would recommend to use assert() if at all possible if you want to check for a condition and exit. If you are debugging.


http://www.cprogramming.com/tips/sho...ount=30&page=0

Hello, I looked up the strto* family of functions and I applied it my problem of scanning a size_t variable. Here's my program and something seems to be going wrong :

Code:

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <math.h>
#include <limits.h>

int main(void)
{
   size_t a; /* size_t variable */
   char a_str[50]; /* String to be converted */
   char *ptr = NULL;
   char *endptr;
   unsigned long l;

   if (fgets(a_str, 50, stdin) == NULL)
   {
     fprintf(stderr, "Error while entering the stirng\n");
     return (1);
   }

   if ((ptr = strchr(a_str, '\n')) != NULL)
   {
     *ptr = '\0';
   }

   errno = 0;

   l =  strtoul(a_str, &endptr, 0); /* Convert string to long unsigned first */

   if (errno == ERANGE)
   {
     if (abs(l) == HUGE_VAL)
     {
       fprintf(stderr, "OVERFLOW\n");
       return (1);
     }
     else
       if (l == 0)
       {
         fprintf(stderr, "UNDERFLOW\n");
         return (1);
       }
   }
   else
   {
     if (l == 0.0)
     {
       fprintf(stderr, "Conversion failed\n");
       return (1);
     }
   }

   if (l <= UINT_MAX) /* size_t typedefed as uint on my tcc */
   {
     a = (size_t)l;
     printf("\n%lu", (unsigned long)a);
   }
   else
   {
     printf("OVERFLOW");
   }
   return (0);

}

The program gives proper output until you enter a negative number for
which, ideally, it should print "UNDERFLOW" but it seems to print
"Conversion failed".

Originally Posted by slingerland3g

You may want to read up on strtol


http://cboard.cprogramming.com/showthread.php?t=104680

The intended purpose is for conversion. I would not base my calculations for boundary checking on this.

I would recommend to use assert() if at all possible if you want to check for a condition and exit. If you are debugging.


http://www.cprogramming.com/tips/sho...ount=30&page=0

Well I would have used assert, but I don't really want to exit on all situations. Only some fatal error situations like memory allocation failure.