Seg fault out of no where

**mart_man00** · 02-25-2003

for some reason this gives me a seg fault, i cant see why. all my pointers are assigned to valid vars or NULL.

Code:

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

int main() {
	char temp[] = "test<abc>data</cba>test";
	remove_tag(temp);
	
	return 0;
}

int remove_tag(char *data) {
	char value[255];
		strcpy(value, data);
	int x = 0;

	int y = 0;
	
	char *start = strchr(data, '<'); /* Find the start of a tag. */
	char *end = strchr(data, '>');   /* Find the end of a tag. */
		end = strchr(data, '>');

	while(x < *start) { /* Copy everything before the tag. */
		data[x] = value[y];

		++x;
		++y;
	}

	x = *end + 1;

	while(data[x] != '\0') { /* Copy everything after the tag*/
		data[x] = value[y];

		++x;
		++y;
	}	

	strcpy(data, value);
	
	return 0;		
}

**quzah** · 02-25-2003

while(x < *start) { /* Copy everything before the tag. */

This is probably not what you think. You pass it a character array, which you dereference character by character. You are comparing the ascii value stored in said character with the integral value of x.

This loop will continue until hit hits a null, or some other character such as a new line, or something, whose ascii value is less than your counter.

I really doubt that this is what you want.

Quzah.

**mart_man00** · 02-25-2003

with while(x < *start), is there a way i can use start as a subscript? i mean like data[start].

**quzah** · 02-25-2003

char *end = strchr(data, '>'); /* Find the end of a tag. */
end = strchr(data, '>');

These lines also do the exact same thing. Remove the second one.

I'm not sure I understand exactly wat you're trying to do. Basicly, I'd do this:

ptr1 = strchr( data, '<' );
ptr2 = strchr( ptr1, '>' );

This will make sure that the second pointer comes after the first in the list. Naturally, you will actually want to do error checking before you do the second call.

Anyway, then you just do this:

strncpy( buffer, ptr1+1, (ptr2-ptr1)-1 );

That should be all you need. If my reasoning is correct, this is what should happen:

ptr1 points to the <
ptr1+1 points to whatever is directly past the <
ptr2 - ptr1 - 1 is the length of the string excluding the > and the <.

That should suffice for copying the contents between the < and the > into a buffer.

Is that what you're trying to do?

Quzah.

**mart_man00** · 02-25-2003

Is that what you're trying to do?

nope. if i run what is in main(with char temp[] = "test<abc>data</cba>test") i want to get back "testtest". i want to just delete the html tags.

since this is supposed to work on real html tags, i need that 2nd end = strchr(data, '>');. since its <tag>some_value</tag> i want to get the 2nd > so i can end it. that seems alittle weird calling the same function 2 times, any better ideas?

**quzah** · 02-25-2003

Your best bet is actually to write a recursive function. This will resolve problems with nested tags. I honestly don't see the need to strip out HTML tag enclosures. Consider the following:

Code:

{html}
{head}
    Some text goes here.
{/head}
{body}
    {p}
        Some test here.
    {/p}
{/body}
{/html}

Your code does not allow for this scenario. What you should be doing is probably using a recursive function, that, when you encounter a <, it calls itself. When it encounters a > it returns. Or something similar.

Furthermore, what would your function expect to return in this scenario? It should according to your outline, return nothing, which is basicly what will happen any time you ever pass your function an HTML file.

Note: I used {} rather than <> so the forum wouldn't eat my HTML.

Quzah.

Thread: Seg fault out of no where

Thread Tools

Search Thread

Display

Seg fault out of no where

Similar Threads

Getting a seg fault

Seg Fault in Compare Function

seg fault at vectornew

weird seg fault

Seg Fault Problem