using the 'rm' command in a c program using the system() function in stdlib.h

Question

i am trying to use the rm command in the main routine for deleting a file that i have taken a command line argument. The value is stored in argv[2] i have tried using

system("rm argv[2]");

system("rm ./argv[2]");

system("rm $HOME/argv[2]");

But it gives me an error saying

"cannot locate file argv[2]"

The filename is stored in argv[2] as I have checked it.

Someone please guide me!

Answer 1

The "argv[2]" in "rm ./argv[2]" is a literal string, if you want to use what is stored in argv[2], you need to use strcpy() or sprintf() to connect "rm ./" with the string stored in argv[2].

Answer 2

You need to construct the string first. Just putting argv[2] in quotes won't work.

Example:

char buf[256];

buf[255] = '\0';
if (snprintf(buf, 255, "rm %s", argv[2]) != -1) {
    system(buf);
}

Answer 3

Why not use the remove or unlink command instead of system("rm ...")?

remove(argv[2]); or unlink(argv[2]);

---

Update in the case that system("rm ...") must be used

If you must use system("rm ..."), ZelluX is correct in stating that you must retrieve the filename from argv[2]. You can retrieve the string in argv[2] using snprintf or strncpy. Using the variants of the functions that restrict the input size is a good idea as there is no guarantee as to the length of the string in argv[2].

Depending on your application, you may also want to call stat to verify that the string in argv[2] is indeed a file and possibly restrict the type of file.

Example:

This example calls stat to verify that argv[2] is a regular file and asprintf to dynamically allocate space for the buffer.

char *p;
struct stat st;

if (stat(argv[2], &st) == 0 && S_ISREG(st->st_mode))
{
    if (asprintf(&p, "rm %s", argv[2]) != -1)
    {
        system(p);
        free(p);
    }
}

Answer 4

argv[2] contains the string and it itself is not the string. Please use strcpy or other string manipulation mechanism to get the string from argv[2] into a buffer and then construct a command using rm and then pass it to system. That should work.

Answer 5

instead of calling system rm command, do it in C, eg

#include <stdio.h>
#include <dirent.h>
int  main()
{
        struct dirent *d;
        DIR *dir;
        char buf[256];
        dir = opendir("mydir");
        while(d = readdir(dir))
        {               
                sprintf(buf, "%s/%s", "mydir", d->d_name);
                remove(buf);
        }
        return 0;

}

Answer 6

See also unlink(2) (defined in unistd.h) if you want to remove files without calling system().

Answer 7

the problem is that argv[2] does not get expanded and is passed just like that, as there is no file named argv[2], there will be an error. A solution would be to produce a string using sprintf() and then pass it to system()

char str[100];
sprintf(str,"rm %s",argv[2]);
system(str);

Answer 8

Using system like this (in fact, using it at all) is a formula for disaster. Suppose argv[2] contains "-rf /". :-)