String and character pointers

[kzar]

Ok im brand new to c . Been programing in delphi and stuff for a while but at the moment im just using nedit and cc in linux.

Im trying to write a program to read a file, and print out the links it has. Most of the code i have tested and works, the problem is the "p[0] = &a;" line. The idea behind it is, i want to shift the first character in the string up some after ive found a link, so it caries on from where it left off, rather than starting from the start and finding the same link each time.

Googled for a while but im totaly stumped this time!

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

void readfile(FILE *fp, char *p);
void findtag(char *p, char *start, char finish, char *result);

main(int argc, char *argv[])
{
	FILE *fp;
	char c[10000], tag[100];
    
	if ((argc != 1) && ((fp = fopen(*++argv, "r")) != NULL))
	{
        	readfile(fp,c);
        	fclose(fp);
	
		while(c != '\0')
		{
			findtag(c,"<a",'>',tag);
			printf("%s\n",tag);
			/* findtag(tag,"href=",' ',tag); */
		} 
	}
	else
		printf("Arf\n");

}

void readfile(FILE *fp, char *p)
{
	int i;

	i = 0;
	while ((p[i++] = getc(fp)) != EOF)
        	;
	p[i - 1] = '\0';
}

void findtag(char *p, char *start, char finish, char *result)
{
	char *a;
	int i;
	
	i = 0;
	a = strstr(p, start) + strlen(start);
	if (a != '\0')
	{
		while ((result[i++] = *a++) != finish)
			;
		result[i-1] = '\0';
	}
	
	p[0] = &a;
}

[Kip Neuhart]

Why not set a pointer to the beginning of the array in main, then pass a pointer to that pointer to findtag. That way you can update the pointer yet still use it in main for your loop test.

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

void readfile(FILE *fp, char *p);
void findtag(char **p, char *start, char finish, char *result);

main(int argc, char *argv[])
{
  FILE *fp;
  char c[10000], tag[100];
  char *p;

  if ((argc != 1) && ((fp = fopen(*++argv, "r")) != NULL))
  {
    readfile(fp,c);
    fclose(fp);

    p = c;
    while(*p != '\0')
    {
      findtag(&p,"<a",'>',tag);
      printf("%s\n",tag);
      /* findtag(tag,"href=",' ',tag); */
    } 
  }
  else
    printf("Arf\n");

}

void readfile(FILE *fp, char *p)
{
  int i;

  i = 0;
  while ((p[i++] = getc(fp)) != EOF)
    ;
  p[i - 1] = '\0';
}

void findtag(char **p, char *start, char finish, char *result)
{
  char *a;
  int i;

  i = 0;
  a = strstr(*p, start) + strlen(start);
  if (a != '\0')
  {
    while ((result[i++] = *a++) != finish)
      ;
    result[i-1] = '\0';
  }

  *p = a;
}

Please note that I have not checked this thoroughly.

[kzar]

Ive given that a go, it now segmentation faults. (after it lists the links) so its basicaly working.

I dont see why it segmentation faults because pointers still confuse me and i dont understand the thing you have done with **p and things. why two *'s?

[kzar]

Ok ive looked at it for ages, improved the code and tried sticking printf's everywhere. I still cant see why it segfaults at the end!

Also now ive tried to add code to find the "'s around links and remove them but that doesnt work either!

I hope c gets easier, im really banging my head againts a wall at the moment.

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

void readfile(FILE *fp, char *p);
bool findtag(char **p, char *start, char finish, char *result);
void cleanup(char *tag);

main(int argc, char *argv[])
{
	FILE *fp;
	char c[10000], tag[100];
	char *p, *o;

	if ((argc != 1) && ((fp = fopen(*++argv, "r")) != NULL))
	{
		readfile(fp,c);
		fclose(fp);

		p = c;
		while(findtag(&p,"<a",'>',tag))
		{
			o = tag;
			findtag(&o,"href=",' ',tag);
			cleanup(tag);
			printf("%s\n",tag);
		} 	
	}
  	else
    		printf("Arf\n");

}

void readfile(FILE *fp, char *p)
{
	int i;

	i = 0;
	while ((p[i++] = getc(fp)) != EOF)
		;
	p[i - 1] = '\0';
}

bool findtag(char **p, char *start, char finish, char *result)
{
	char *a;
	int i;

	i = 0;
  
	if (a = strstr(*p, start) + strlen(start))
	{
		while ((result[i++] = *a++) != finish)
			;
		result[i-1] = '\0';
		*p = a;
		return 1;
	}
	else
		return 0;
}

void cleanup(char *tag)
{
	char *a, temp[100];
	int i;
	
	if (tag[0] == '"')
	{
		i=1;
		*a=tag[1];
		while((*a++ = tag[i++]) != '"')
			temp[i-1] = *a;
		tag = temp;
	}
}

[Dave_Sinkula]

bool findtag(char **p, char *start, char finish, char *result)
{
	/* ... */
	if (a = strstr(*p, start) + strlen(start))
[Warning 613] Possible use of null pointer 'unknown-name' in left argument to operator 'ptr+int'

[edit]
Did you mean to do something like this?

   a = strstr(*p, start);
   if ( a != NULL )
   {
      a += strlen(start);

Also,

void cleanup(char *tag)
{
	/* ... */
		*a=tag[1];
[Warning 530] Symbol 'a' not initialized

[/edit]

[kzar]

   a = strstr(*p, start);
   if ( a != NULL )
   {
      a += strlen(start);

Looks good , i tried somthing similar except i did "a = a + strlen(start)" which gave me an error. That was the final straw and i posted here

Could you explain why you used a += ... instead of a = a + ... ? What is += ?

[Hammer]

a += 10;
is shorthand for
a = a + 10;

[kzar]

ok ive just tried your sugestion with the += and it does what it did when i tried it by doing a = a + ... , it segmentation faults.

i dont really understand the warning you posted about the cleanup function either .

[Dave_Sinkula]

>Could you explain why you used a += ... instead of a = a + ... ? What is += ?

A compound assignment of the form E1 op= E2 differs from the simple assignment expression E1 = E1 op (E2) only in that the lvalue E1 is evaluated only once.

[kzar]

What did you get that warning from by the way? im doing cc in linux to compile but if theres somthing else i could use that gave me more of a clue whats wrong with my programs i might give it a go.

[Dave_Sinkula]

>i dont really understand the warning you posted about the cleanup function either

'a' is a pointer. It is not pointing anywhere. Later on, you do an assignment to where 'a' points ('*a'). So, you are dereferencing some random memory location. Bad -- could cause a seg fault.

>ok ive just tried your sugestion with the += and it does what it did

The += stuff was not the main issue. The main thing was that in 'findtag' the return value of strstr should be checked to see that is was not NULL before attempting to find an offset. Dereferencing a null pointer could cause a seg fault.

>What did you get that warning from by the way?

PC-lint for C/C++ (NT) Ver. 8.00m


[edit]Boy am I typing slow today.

[kzar]

if ( a != NULL ) checks if its nulll doesnt it?

[Dave_Sinkula]

>if ( a != NULL ) checks if its nulll doesnt it?

Yup.

Okay, how about this one.

void readfile(FILE *fp, char *p)
{
	int i;

	i = 0;
	while ((p[i++] = getc(fp)) != EOF)
[Info 734] Loss of precision (assignment) (31 bits to 7 bits)
		;
	p[i - 1] = '\0';
}

FAQ

[Hammer]

>>while ((p[i++] = getc(fp)) != EOF)
buffer overflow anyone?

Protect your buffers at all times, to keep your app safe.

[kzar]

thanks for all the help guys , i looked closely at the code and saw you where right, i forgot to say where the pointer was actauly pointing to. I fixed that and it no longer segmentation faults.

I didnt really understand why that would be a buffer overflow though, and the cleanup function still isnt working.

Heres the code so far

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

void readfile(FILE *fp, char *p);
bool findtag(char **p, char *start, char finish, char *result);
void cleanup(char *tag);

main(int argc, char *argv[])
{
	FILE *fp;
	char c[10000], tag[100];
	char *p, *o;

	if ((argc != 1) && ((fp = fopen(*++argv, "r")) != NULL))
	{
		readfile(fp,c);
		fclose(fp);

		p = c;
		while(findtag(&p,"<a",'>',tag))
		{
			o = tag;
			findtag(&o,"href=",' ',tag);
			cleanup(tag);
			printf("%s\n",tag);
		} 	
	}
  	else
    		printf("Arf\n");

}

void readfile(FILE *fp, char *p)
{
	int i;

	i = 0;
	while ((p[i++] = getc(fp)) != EOF)
		;
	p[i - 1] = '\0';
}

bool findtag(char **p, char *start, char finish, char *result)
{
	char *a;
	int i;

	i = 0;
  	
	a = strstr(*p, start);
	if (a != NULL)
	{
		a += strlen(start);
		while ((result[i++] = *a++) != finish)
			;
		result[i-1] = '\0';
		*p = a;
		return 1;
	}
	else
		return 0;
}

void cleanup(char *tag)
{
	char *a, temp[100];
	int i;
	
	if (tag[0] == '"')
	{
		i=1;
		a=&tag[1];
		while((*a++ = tag[i++]) != '"')
			temp[i-1] = *a;
		tag = temp;
	}
}