read file start from different positions

Hi all,

I have a file which contains some specials chars to gather one for each line. For example if the special char is $ (dollar) the file may looks like:

Code:

---

hello world $ hello
$ hello world
hello world hello world $

---

I have to gather the $ and ignore the rest. I found a solution but not efficient. In fact once I found the $ is there a way to jump directly to the next row without keep reading ?

The second row in the example contains the $ at the beginning and it is useless to keep reading after the first char.

Is there any other solution?
Thanks in advance

Within a text file, you can only fseek() to the start of a line if you've already read the file at least once, and recorded all the ftell() positions.

What is your "inefficient" method?
I would have thought fgets() + strchr() would do the job just fine.

Hi,

the method I use is a :

Code:

---

while((c=getc(stream)) != '\n'){

---

so I read always the entire row

thanks for reply

Quote:

---

Originally Posted by Dedalus

Hi all,

I have a file which contains some specials chars to gather one for each line. For example if the special char is $ (dollar) the file may looks like:

Code:

---

hello world $ hello
$ hello world
hello world hello world $

---

---

To what end? So you end up with 3 $ signs... what good is that, except to count them...

Code:

---

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

int main (void)
 { int count;
    int fs;
    char *fb;
    char *bp;
    FILE *fp;
 
  fp = fopen("filename.txt")
    if (!fp)
    { printf("Fopen Error\n");
        exit (1); }
   
    fseek(fp,0,SEEK_END);
    fs = ftell(fp)
    fb = calloc(fs, sizeof(char));
    fseek(fp,0,SEEK_SET);
    if (!fb)
    { printf("Calloc error\n");
        exit (1); }

  fs -= fread(fb,sizeof(vhar),fs,fp);
  if (fs)
    { fprintf("Fread Error\n");
      exit (1); }
  fclose(fp);

  bp = fb;
  while (bp = strchr(bp,'$'))
      count++;

  free(fb);
  printf("I found %d\n", count);
  return 0; }

---

That should be pretty close.

Hello

thanks for sharing your solution.
I use it to parse some text files, usually I use bash but wanted to try C too.

Cheers

Quote:

---

Originally Posted by Dedalus

I have to gather the $ and ignore the rest. I found a solution but not efficient. In fact once I found the $ is there a way to jump directly to the next row without keep reading ?

---

Underlying assumption is that there's only one $ per line?
What if there were more than one $ per line; then what?

Quote:

---

Originally Posted by Dedalus

The second row in the example contains the $ at the beginning and it is useless to keep reading after the first char.

Is there any other solution?

---

Depends if you can predict where the next $ would be :wink:

Quote:

---

Originally Posted by Salem

I would have thought fgets() + strchr() would do the job just fine.

---

Quote:

---

Originally Posted by Dedalus

Hi,

the method I use is a :

Code:

---

while((c=getc(stream)) != '\n'){

---

so I read always the entire row

thanks for reply

---

The easiest and quickest method is what Salem suggested. Just use fgets to grab each line of the file and dump it into a buffer then use strchr on the buffer to get all the occurences of your token. This is just about as sleak as you are going to get. An example of strchr() can be found here.

Quote:

---

Originally Posted by AndrewHunter

The easiest and quickest method is what Salem suggested. Just use fgets to grab each line of the file and dump it into a buffer then use strchr on the buffer to get all the occurences of your token. This is just about as sleak as you are going to get. An example of strchr() can be found here.

---

On the thought that $ is a comment indicator like C's // ... he can do fgets() to dump it into a buffer then use strchr() to find his symbol and replace it with a \0 to truncate the string. That's pretty efficient parsing...