So I am trying to read from standard input and then get the input ready so that later on it can be used inside execvp().
What I am implementing here is basically a pipe for some terminal commands.
Here is how an example of my code goes.
input:
ls -s1
sort -n
output:
commands[0]="ls"
commands[1]="-s1"
commands2[0]="��""
commands2[1]="��""
sort: cannot read: t: No such file or directory
Here is my code
# include <stdlib.h>
# include <stdio.h>
# include <unistd.h>
# include <string.h>
# include <sys/wait.h>
# define BUF_SIZE 256
int main()
{
char buffer[BUF_SIZE];
char *commands[5];
char *commands2[5];
int argc = 0;
int argc2 = 0;
fgets(buffer, BUF_SIZE, stdin);
for ( commands[argc] = strtok(buffer, " \t\n");
commands[argc] != NULL;
commands[++argc] = strtok(NULL, " \t\n") ) {
printf("commands[%d]=\"%s\"\n", argc, commands[argc]);
}
commands[argc] = NULL;
fgets(buffer, BUF_SIZE, stdin);
for ( commands2[argc2] = strtok(buffer, " \t\n");
commands2[argc2] != NULL;
commands2[++argc2] = strtok(NULL, " \t\n") ) {
printf("commands2[%d]=\"%s\"\n", argc2, commands2[argc]);
}
commands2[argc2] = NULL;
int my_pipe[2];
if (pipe(my_pipe) == -1)
{
perror("cannot create pipe\n");
}
pid_t my_pid;
my_pid = fork();
if (my_pid < 0)
{
perror("Failed fork\n");
}
if (my_pid > 0)
{
close(my_pipe[1]);
dup2(my_pipe[0], 0);
close(my_pipe[0]);
wait(NULL);
execvp(commands2[0],commands2);
}
else
{
close(my_pipe[0]);
dup2(my_pipe[1], 1);
close(my_pipe[1]);
execvp(commands[0],commands);
}
}
One major problem is that you read the second line over the first in buffer
, and the commands[]
array contains pointers into buffer
too. That's not a recipe for happiness. The simplest fix is to define char buffer2[BUF_SIZE];
and use that in the second fgets()
call and for
loop.
Using argc
in printf("commands2[%d]=\"%s\"\n", argc2, commands2[argc]);
is a copy'n'paste bug — it should reference argc2
twice. This helped hide the previous problem.
Note that perror()
does not exit; your code blunders on if pipe()
fails, or if fork()
fails.
The wait()
in if (my_pid > 0)
is bad; remove it.
If execvp()
fails, you should report an error and exit with a non-zero status.
Putting those changes together yields code such as:
#include <stdlib.h>
#include <stdio.h>
#include <unistd.h>
#include <string.h>
#include <sys/wait.h>
#define BUF_SIZE 256
int main(void)
{
char buffer[BUF_SIZE];
char buffer2[BUF_SIZE];
char *commands[5];
char *commands2[5];
int argc = 0;
int argc2 = 0;
fgets(buffer, BUF_SIZE, stdin);
for (commands[argc] = strtok(buffer, " \t\n");
commands[argc] != NULL;
commands[++argc] = strtok(NULL, " \t\n"))
{
printf("commands[%d]=\"%s\"\n", argc, commands[argc]);
}
commands[argc] = NULL;
fgets(buffer2, BUF_SIZE, stdin);
for (commands2[argc2] = strtok(buffer2, " \t\n");
commands2[argc2] != NULL;
commands2[++argc2] = strtok(NULL, " \t\n"))
{
printf("commands2[%d]=\"%s\"\n", argc2, commands2[argc2]);
}
commands2[argc2] = NULL;
int my_pipe[2];
if (pipe(my_pipe) == -1)
{
perror("cannot create pipe\n");
exit(EXIT_FAILURE);
}
pid_t my_pid = fork();
if (my_pid < 0)
{
perror("Failed fork\n");
exit(EXIT_FAILURE);
}
if (my_pid > 0)
{
close(my_pipe[1]);
dup2(my_pipe[0], 0);
close(my_pipe[0]);
execvp(commands2[0], commands2);
perror(commands2[0]);
exit(EXIT_FAILURE);
}
else
{
close(my_pipe[0]);
dup2(my_pipe[1], 1);
close(my_pipe[1]);
execvp(commands[0], commands);
perror(commands[0]);
exit(EXIT_FAILURE);
}
return EXIT_FAILURE;
}
When I run the program, it produces the appropriate output. Note that the return
at the end of main()
is actually never reached.