Thread: How to extract url from strings

Hi,
With the below function i can extract urls from strings. But it works with only "http://". So how can i change it to extract urls which begins with "www" also.

Thanks in advance.

Code owner: James K. Lawless

Code:

void print_urls(char *s) {
   char *p,*mark;
   _state=0;
   for(p=s;*p;p++) {
      switch(_state) {
         case 0:
            if(*p=='h') {
               _state=S_h;
               mark=p;
            } 
            break;
         case S_h:
            if(*p=='t') 
               _state=S_t1;
            else
               _state=0;
            break;
         case S_t1:
            if(*p=='t') 
               _state=S_t2;
            else
               _state=0;
            break;
         case S_t2:
            if(*p=='p') 
               _state=S_p;
            else
               _state=0;
            break;
         case S_p:
            if(*p==':') 
               _state=S_col;
            else
            if(*p=='s')
               _state=S_s;
            else
               _state=0;
            break;
         case S_s:
            if(*p==':')
               _state=S_col;
            else
               _state=0;
            break;
         case S_col:
            if(strchr(legal_chars,tolower(*p))==NULL) {
               while(mark<p) {
                  fputc(*mark,stdout);
                  mark++;
               }
               fputc('\n',stdout);
               _state=0;
               p--; // backtrack
            } 
      }
   }
   if(_state) {
      while(mark<p) {
         fputc(*mark,stdout);
         mark++;
      }
   }
}

Sorry for being such a douche and double post. Since i wrote something new, i wanted to show you.

Code:

#include "string.h"
#include "stdio.h"
#define BAD(x) (!(x) || (*(x) == '\0'))

static char *extract_link(char string[])
{
        static char url[256];
        int length;
        char *st;
        char *rl;
        url[0] = '\0';
        rl = strstr(string, "http://");
        if (BAD(rl))
                goto www;
        if (!BAD(rl) && strchr(rl, ' '))
        {
                st = strstr(rl, " ");
                length = strlen(rl) - strlen(st);
                strncpy(url, rl, length);
                return url;
        }
        else
                return rl;
        www:
                rl = strstr(string, "www.");
                if (BAD(rl))
                        return NULL;
                if (!BAD(rl) && strchr(rl, ' '))
                {
                        st = strstr(rl, " ");
                        length = strlen(rl) - strlen(st);
                        strncpy(url, rl, length);
                        return url;
                }
                else
                        return rl;
}
int main()
{
        char string1[] = "We have a http://www.google.com here";
        char string2[] = "We have a www.google.com/blabla here whatsoever";
        char *x = extract_link(string2);
        char *y = extract_link(string2);
        printf("%s\n%s\n", x, y);
        return 0;
}

It works with just one string.But if i make an another call with new string to extract_link function (like above) the urls are being mixed and shown as the same.
How can i fix this problem?

Yes. Because url[256] is a local buffer in the function whose contents disappears once the function exits.
You should reserve space for returned string...
char x[256], y[256] before you call the function. Then call it
extract_link(string1, x)... The second parameter being used within the function to strcpy the result there.
Oh and you called extract_link with string2 twice.

Thank you. (using the string 'string2' twice was my mistake.)

I changed the codes as

Code:

#include "string.h"
#include "stdio.h"
#define BAD(x) (!(x) || (*(x) == '\0'))

static char *extract_link(char *string, char url[512])
{
        int length;
        char *st;
        char *rl;
        url[0] = '\0';
        rl = strstr(string, "http://");
        if (BAD(rl))
                goto www;
        if (!BAD(rl) && strchr(rl, ' '))
        {
                st = strstr(rl, " ");
                length = strlen(rl) - strlen(st);
                strncpy(url, rl, length);
                return url;
        }
        else
                return rl;
        www:
                rl = strstr(string, "www.");
                if (BAD(rl))
                        return NULL;
                if (!BAD(rl) && strchr(rl, ' '))
                {
                        st = strstr(rl, " ");
                        length = strlen(rl) - strlen(st);
                        strncpy(url, rl, length);
                        return url;
                }
                else
                        return rl;
}
int main()
{
        char x[256], y[256];
        char *string1 = "We have a http://www.google.com here and there";
        char *string2 = "We have a www.google.com/blabla here whatsoever";
        extract_link(string1, x);
        extract_link(string2, y);
        printf("%s\n%s\n", x, y);
        return 0;
}

Still no avail..

Output:

Code:

./x
http://www.google.comûõ·Xøõ·à·à
www.google.com/blabla

Because you are not allocating memory for your strings.

What happens to rl and url when your function returns?

Code:

char link[512];

// function
void ExtractUrl(char *in, char *out)
  { 
     // figure out the url from the in variable

     strncpy(out,url,511);   // copy and return
}

Woah, late to the party, CommonTater.