Thread: Here I go again...arrays....char...

Ok, this is the same thing. You can see that code I have written up is working, but now i want to do it this way...

Code:

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

int main()
{
	int sranje;
	char niz[50];

	printf("Rijec: ");
	fgets(niz,sizeof(niz),stdin);

	funk(niz);
	scanf("%d", &sranje);
}

int funk(char niz[]){

	char *niz2, *niz3;
	int i,j=0,d=0;
	niz2=malloc(sizeof(char));
	niz3=malloc(sizeof(char));

	for(i=0;i<strlen(niz)+1;i++){

		if(niz[i]>=65 && niz[i]<= 122){

			if(niz[i]=='A' || niz[i]=='E' || niz[i]=='I' || niz[i]=='O' || niz[i]=='U' ||
				niz[i]=='a' || niz[i]=='e' || niz[i]=='i' || niz[i]=='o' || niz[i]=='u'){
				
				niz2[j]=niz[i];
				j++;}
			else
				niz3[d]=niz[i];
				d++;
				}
			}
	printf("%s", niz2);
	printf("\n%s", niz3);
}

It's the same thing but little shorter. Everything works, well not really, if i input "Zadar", output is " aa""" ", instead of " is 2, and for niz3 is " z"d"r " and I want it to be aa, zdr. Any suggestions are most welcome.

You may want to do

Code:

niz2[j] = 0;
niz3[d] = 0;

before you try to print your strings.

I would probably do something like this:

Code:

    switch(tolower(niz[i]) {
       case 'a':
       case 'e':
       case 'i':
       case 'o':
       case 'u':
           niz2[..] = ....;
           break;
       default:
           niz3[...] = ...;
           break;
       }

--
Mats

Works fine for niz2[j], but niz3 is still messed up! What niz2[j]=0; does? And when I dont enter anything it still prints out some signs for niz3.

Originally Posted by shardin

Works fine for niz2[j], but niz3 is still messed up! What niz2[j]=0; does? And when I dont enter anything it still prints out some signs for niz3.

Strings need to be terminated with a NULL character, otherwise the printf routine doesn't know where to stop outputting strings. It can't know after all, how much memory you've allocated. If you don't have a NULL terminator, then printf will continue reading memory past the end of your allocated memory until it encounters the first NULL char. That could be three chars from the end, it could be at the end of that page file or there could even not be any NULLs in memory.
So you should always set the last char in your string to NULL, by doing:

Code:

niz2[j] = 0;
/* or */
niz2[j] = '\0';

QuantumPete

Ok, can someone explain to me why this ...

Code:

for(i=0;i<strlen(niz)+1;i++){

		if(niz[i]>=65 && niz[i]<= 122){

			if(niz[i]=='A' || niz[i]=='E' || niz[i]=='I' || niz[i]=='O' || niz[i]=='U' ||
				niz[i]=='a' || niz[i]=='e' || niz[i]=='i' || niz[i]=='o' || niz[i]=='u'){
				
				niz2[j]=niz[i];
				j++;}
			else
				niz3[d]=niz[i];
				d++;
				}
			}

...is not working, and this...

Code:

	for(i=0;i<strlen(niz)+1;i++){
		if(niz[i]>=65 && niz[i]<= 122){
                      
				if(niz[i]=='A' || niz[i]=='E' || niz[i]=='I' || niz[i]=='O' || niz[i]=='U' ||
				niz[i]=='a' || niz[i]=='e' || niz[i]=='i' || niz[i]=='o' || niz[i]=='u')
              
				niz2[j++]=niz[i];
			else
		niz3[d++]=niz[i];
				}
			 }

...works perfectly! In the upper code niz3 is messed up. Only difference is in...

Code:

{
				
				niz2[j]=niz[i];
				j++;}
			else
				niz3[d]=niz[i];
				d++;
				}

...and...

Code:

				niz2[j++]=niz[i];
			else
		niz3[d++]=niz[i];

Only one of your strings is going to get a \0 written to the end of it.

Rearrange the code, and you get a different kind of luck.

> if(niz[i]>=65 && niz[i]<= 122)
How about
if(niz[i]>='A' && niz[i]<= 'z')

In this:

Code:

for(i=0;i<strlen(niz)+1;i++){

		if(niz[i]>=65 && niz[i]<= 122){

			if(niz[i]=='A' || niz[i]=='E' || niz[i]=='I' || niz[i]=='O' || niz[i]=='U' ||
				niz[i]=='a' || niz[i]=='e' || niz[i]=='i' || niz[i]=='o' || niz[i]=='u'){
				
				niz2[j]=niz[i];
				j++;}
			else
				niz3[d]=niz[i];
				d++;
				}
			}

Check the braces - particularly on the else-branch. If you use Emacs, mark the region and do "M-x indent-region", or in MSVC under Edit->Advanced->Format Selection (again after marking the region).

Always using braces on ALL parts of if/else/while statements avoids this problem.

--
Mats

> if(niz[i]>=65 && niz[i]<= 122)
How about
if(niz[i]>='A' && niz[i]<= 'z')

Or even

Code:

#include <ctype.h>

if(isalpha(niz[i]))

assuming that the intent of the code was to see if niz[i] was a lower or uppercase letter.

tolower() might also be called for -- then you don't have to check for 'E' and 'e', 'I' and 'i'.

> Always using braces on ALL parts of if/else/while statements avoids this problem.
Sound advice indeed!

Also:

Code:

for(i=0;i<strlen(niz)+1;i++){

Calculating strlen() is expensive, you might want to do it outside of the loop. And that's the same as

Code:

for(i=0;i<=strlen(niz);i++){

which might be easier to read, or it might not.

I'm not allowed to use string functions in exam! Only strlen!

No, the thing is in braces and 'else' statement! If I use niz[i]=niz2[a], i++, then I need to use braces, if I use niz[i++]=niz2[a], then I dont need braces and then affects 'else' statement, yes I just realized! It doesnt have anything with upper, lower...

I'm not allowed to use string functions in exam! Only strlen!

I wasn't suggesting that you use a different function. I was suggesting that you calculate strlen() before the loop and save the result in a variable.

Code:

size_t len = strlen(niz);
for(i=0;i<len+1;i++){

It doesnt have anything with upper, lower...

Your code also matches [, ], and other characters that are between 'Z' and 'a'. Is that what you wanted? Probably not. And your code isn't even portable. It could break with different character sets.

All in all, isalpha() is a very good idea.

The tolower() suggestion would make your code a lot easier.

Code:

			if(niz[i]=='A' || niz[i]=='E' || niz[i]=='I' || niz[i]=='O' || niz[i]=='U' ||
				niz[i]=='a' || niz[i]=='e' || niz[i]=='i' || niz[i]=='o' || niz[i]=='u'){

->

Code:

			char c = tolower(niz[i]);
			if(niz[i]=='a' || niz[i]=='e' || niz[i]=='i' || niz[i]=='o' || niz[i]=='u'){

No, no, no....I think you are missing my point.

No, the thing is in braces and 'else' statement! If I use niz[i]=niz2[a], i++, then I need to use braces, if I use niz[i++]=niz2[a], then I dont need braces and then affects 'else' statement, yes I just realized!

I realize that you identified the problem as being a case of missing curly braces, and that you can resolve this by using a single, more concise statement.

My suggestions are just that -- suggestions. They won't affect the compilation or execution (much) of your program, except to perhaps speed it up (in the strlen() case) or not count [ and ] etc as letters (in the isalpha() case). And the tolower() one makes the code shorter.

If you're not allowed to use functions from <ctype.h>, then . . . oh well. That would be ridiculous.