Thread: Some direction please

I need to create a program that will perform a diff on all files in two different directories. Basically it's going to check for any changes between these 2 directories.

The main "base standard" directory will change as will the comparison directory.

I read the man on ftw and I believe that is the way I want to go for traversing the directory. However, I need to use the "diff" command for this processing so I need to perform an exec(X) of some sort.

Here's some sample code I tried and it works well for one file at a time:

Code:

#include <stdio.h>

main() {
  FILE *fpipe;
  char *command="diff /base/standard.txt /comp/standard.txt";
  char line[256];

  if (!(fpipe = (FILE*)popen(command, "r"))) {
    perror("Problem with pipe");
    exit(1);
  }
  while (fget(line, sizeof line, fpipe)) {
    printf("%s", line);
  }
  pclose(fpipe);
}

My question is this. Should I use the popen for each file and store all the output from the diff command in a variable? In which case, I guess I'll have to pclose each time as well. Or should I be looking at pipe/fork/exec? At times there will be directory trees that go 5-6 levels deep and could be hundreds of files. What is the best way to make sure this scales up well?

Will sending the output of each diff command to a file incur too much time delay? This process needs to be extremely fast.

Thank you in advance for your guidance.

Why not just use a shell script?

If you need fast thing, never do multiprocessing.

new proccess == new resource == more load time == slow

capturing another proccess output == ++slow

I suggest you to make your diff routine, then you may make multithreading program to gain more speed.

Or, if you have to use C you could use tmpfile Copy the content across, and just scan once in a while. Its not going to be fast, but it would be pretty easy.