Thread: Loop and Side-Effect

Can this be considered as a side-effect, bad practice ?

Code:

char empty[1] = {0};

while (array[i]) {

    // more code

    if (abc()) {
        array = empty;
        i = -1;

    }
    i = 1 + i;
}

Yes!

Why would you do this?

Better:

Code:

#include <stdio.h>

int main(void)
{
  while(1)   // Infinite loop
  {
    if(Some false condition)
    {
      break;
    }
  }
  return 0;
}

I would rather avoid break;

Originally Posted by kodax

I would rather avoid break;

while(1), break, continue, are all the normal way to handle loops, and infinite loops, in C programming. If you use other methods, then you are on your own.

Infinite loop is also bad.

If statement is the correct way to do it, although the overhead will increase.

I suggest for your own sanity you stop

Projection.

It really depends exactly what you're trying to do. Based on your last code I might write it thus:

Code:

for (int i = 0; !abc(); ++i) {
    if (array[i]) { // More code
    }
}

One thing this does that yours doesn't is declare and initialize i. That part can be modified or taken out as needed.

What am I missing? Why do you need empty at all? Just use NULL

Code:

void my_function(int array[]) {
   int i = 0;
   while (array && array[i]) {
       // more code   
       if (abc()) {
           array = NULL;
       }
       i = 1 + i;
   }
}

No, on second through I'm agreeing with the others - it is a folly to not use 'break' in this case.


However if you do want to avoid using break that original code isn't the way to go about it. I've put three versions into Compiler Explorer:

Code:

#include <stdlib.h>


int abc(void);


void my_function1(char array[]) {
   int i = 0;
   while (array && array[i]) {
       // more code   
       if (abc()) {
           array = NULL;
       }
       i = 1 + i;
   }
}


 
void my_function2(char array[]) {
   char empty[1] = {0};
   int i = 0;
   while (array[i]) {
      // more code 
      if (abc()) {
         array = empty;
         i = -1;   
      }
      i = 1 + i;
   }
}

This is the resulting code:

Code:

my_function1(char*):
        test    rdi, rdi
        je      .L9
        push    rbx
        mov     rbx, rdi
.L3:
        cmp     BYTE PTR [rbx], 0
        je      .L1
        call    abc()
        add     rbx, 1
        test    eax, eax
        je      .L3
.L1:
        pop     rbx
        ret
.L9:
        ret


my_function2(char*):
        cmp     BYTE PTR [rdi], 0
        je      .L20
        push    rbp
        mov     rbp, rdi
        push    rbx
        xor     ebx, ebx
        sub     rsp, 8
.L14:
        call    abc()
        test    eax, eax
        je      .L23
        add     rsp, 8
        pop     rbx
        pop     rbp
        ret
.L23:
        add     ebx, 1
        movsx   rax, ebx
        cmp     BYTE PTR [rbp+0+rax], 0
        jne     .L14
        add     rsp, 8
        pop     rbx
        pop     rbp
        ret
.L20:
        ret

The top one looks to produce nicer code to me...

What am I missing

1. If block sets array to empty, to break out of the loop.
2. If block does error handling, this will be useful after the loop.

Code:

while (array && array[i])

3. While block will never be executed..

Again, it all depends on exactly what you're trying to do. What is array? I can only deduce that it's a pointer (not an array), since you can't assign to an array. Are you returning that pointer to a caller, or what else do you need to do with it after the loop ends? Also, what does "empty" mean? The empty array is not actually empty--it contains a single element (it is, however, a valid zero-length string, but initializing it with array syntax suggests you're not treating it as a string).

We can all post code that does what we think you're trying to do, based only on the code that you've posted so far, but it's still not clear what you want the code to do.

Originally Posted by kodax

1. If block sets array to empty, to break out of the loop.
2. If block does error handling, this will be useful after the loop.

Code:

while (array && array[i])

3. While block will never be executed..

Can you expand in points 2 and 3, because I can't see what you are getting at.... here's the two implementations, with abc() changed to see if the string has a 'z' in it, and doing 'error handling'. Both implementation produce the same output (except one doesn't segfault if it is passed a NULL).

Code:

#include <stdio.h>
#include <stdlib.h>




int abc(char c) {
   return c == 'z';
}


void my_function1(char array[]) {
   int i = 0;
   while (array && array[i]) {
       // more code
       putchar(array[i]);

       if (abc(array[i])) {
           array = NULL;
       }
       i = 1 + i;
   }
   if(!array) {
      printf("It had a 'z' in it \n");
   }
}


void my_function2(char array[]) {
   char empty[1] = {0};
   int i = 0;
   while (array[i]) {
      // more code
      putchar(array[i]);

      if (abc(array[i])) {
         array = empty;
         i = -1;
      }
      i = 1 + i;
   }
   if(array == empty) {
      printf("It had a 'z' in it \n");
   }
}


int main(void) {
   my_function1("Hello!\n");
   my_function2("Hello!\n");
   return 0;
}

Code:

$ ./a
Hello!
Hello!

(using 'break' is better than either of them... IMO)