while (char[ arraySize - 1] != '/0' ) Why can't I do this?

I want to input into a char array and then have a while loop determine where the input terminates. I don't understand why I can't use the above statement though. My book was really vague about the char data type.


6.33.c:11:38: warning: multi-character character constant
6.33.c: In function `main':
6.33.c:11: warning: comparison is always true due to limited range of data type

Code:

---

#include <stdio.h>

int testPalindrome(char[], int, int);

int main()
{
  char string[ 15 ];
  int arraySize = 15, nSize = 15;
  scanf("&#37;s", &string);

  while( string[ arraySize - 1 ] != '/0' ) {
      --arraySize;
      printf("arraySize: %d\n", arraySize);
  }

  --arraySize;
  nSize = arraySize - 1;

  printf("arraySize: %d nSize: %d\n", arraySize, nSize);
  printf("%d\n", testPalindrome(string, nSize, arraySize) );


return 0;
}



int testPalindrome(char array[], int n, int size)
{
  if (size % 2 == 1 && (size - 1) / 2 == n ) {
      // TEST STATEMENT
      printf("if 1 n: %d\n", n);
      return 1;
  }
  else if (size % 2 == 0 && (size - 1) / 2 == n) {
      // TEST STATEMENT
      printf("if 2 n: %d\n", n);
      if (array[ n ] == array[ size - n ])
        return 1;
      else
        return 0;
  }
  else {
      // TEST STATEMENT
      printf("if 3 n: %d\n", n);
      if (array[ n ] == array[ size - n - 1 ])
        return testPalindrome(array, n - 1, size);
      else
        return 0;
  }

}

---

Code:

---

while( string[ arraySize - 1 ] != '\0' )

---

You made a mistake there.

ssharish2005

You are using the wrong slash, you should use '\0', not '/0' - the latter is two characters, the first is one (with the value zero).

However scanning bacwards in an undefined string for a non-zero value is probably not a super-good idea. Use strlen() instead. It's a standard function, it's most likely faster than the way you would come up with anyways. And it's doing exactly what you want it to do: counting the number of letters in your string.

--
Mats

It appears that you're using gcc. I highly recommend always using at least the -Wall flag when building with gcc. It would have caught one other problem in your code: you should pass `string' or `&string[0]' to scanf(), not `&string'. Admittedly this is a fairly pedantic correction to an almost certainly benign bug, but a bug nevertheless. The main point, though, is that -Wall will provide you a lot more warnings, almost all of them useful.

--> matsp
Do you think it's a good idea to be going beyond my book at this point though?


--> cas
What exactly does -wall do compared to pedantic?

Quote:

---

Originally Posted by yougene

Do you think it's a good idea to be going beyond my book at this point though?

---

If it is teaching you bad habits, you might even consider finding a different one.

Quote:

---

Originally Posted by yougene

--> cas
What exactly does -wall do compared to pedantic?

---

-pedantic will warn about some gcc extensions; it makes the compiler a little more standards compliant. For example, `a ? : b' is valid in GNU C, but not standard C; -pedantic will warn you about that.

-Wall will warn about all sorts of potential problems, even if the C standard doesn't necessarily require a diagnostic. For example, -Wall will check to see if your arguments to printf() match up with your conversion specifiers.

If you're striving toward ANSI/ISO C compliance, a good set of options is: -Wall -std=c99 -pedantic

You might prefer c89, tastes vary.

> If you're striving toward ANSI/ISO C compliance, a good set of options is: -Wall -std=c99 -pedantic

GCC still doesn't have full C99 support (though it's getting close). So at present, C90 is the only C standard it's capable of complying with. On the other hand, it's good practice to write C code that complies with both C90 and C99 (the first because it's most common now, the second because it's the future). So -std=c99 might be useful in testing code to pick up some cases of C99 noncompliance, but is probably not a good idea to use by default.

http://gcc.gnu.org/c99status.html

I use

Code:

---

gcc -W -Wall -ansi -pedantic ...

---

-W is the same as -Wextra.

I usually stick in -g for debugging information as well.

Quote:

---

Originally Posted by robatino

>So -std=c99 might be useful in testing code to pick up some cases of C99 noncompliance, but is probably not a good idea to use by default.

---

Hence "tastes vary". =)

Although gcc doesn't fully conform to C99, I unabashedly use some of the features that it does currently support. I've grown fond of compound literals, designated initializers, variable argument list macros, snprintf() (admittedly a library issue), stdbool, exact-width integers, among others.

It really depends on what your code targets. If you're guaranteed a recent gcc or other C99-esque compiler (such as Intel's), using C99 features may not be a bad thing. Of course, if you want to be maximally portable, C89 is still the proper target. I just don't care much about C89 compliance when it comes to code for my own personal use.

That still leaves a lot of C89 compilers out there, and just a few which pre-date the dinosaurs ;)

We try to stick to C89 here, so that we don't have to care which compiler someone is using, and anyone wandering by has a good chance of using the information.

It is also fine to suggest C99 features if that would make the solution better, but it would be nice to add a "you need C99 for this" type disclaimer along side it.

So as a beginner it would be smart for me to stick with C89?

It would be smart of you to remain aware that there is a difference, and be aware of the language features you're using in your programs.

In the end, you can do what you like. But hopefully, it will be an informed choice rather than just blindly going with whatever your compiler will let you get away with.