Program Input and Output

System and Exec

• Perl is a delightfull glue language, allowing you to tie together all sorts of operations in a single place, and to leverage all the resources of your operating system to perform your task.
• The most straightforward way to execute a particular command is via the system command.
• system takes a scalar string, and tells the shell to execute it.
• This process inherits the same STDIN, STDOUT, and STDERR as your perl process has, so it probably gets input from the users keyboard and reports results to the users screens (unless your perl program has told it to do otherwise).
• The command is executed by the shell, so you can use all the neat shell tricks and redirection symbols you wish. This is a huge advantage over the C system call, but can lead to security holes if you are not carefull, particularly in web applications.
• Normally, the system call will block until the invoked process completes. If you wish it to return immediately, add the normal shell background process indicator ("&") to the scalar string argument to system.
• The value returned by system is the value returned by the shell (usually zero for success, non zero for error).
• Example:
  

  # Get a long directory listing, grep it for files owned by user,
  # and write this list to a file named myfiles.txt
  $userName = "killbill";
  $sysCommand = "ls -l | grep $userName > myfiles.txt";
  if ( system($sysCommand) != 0 ){
      die "Unable to create file $!";
  }
  

• The complete shell environment is available to both your perl program and your system subshell. To get at the environmental variables within your Perl program. use the %ENV hash. To get at them in your subshell, just type them as normal (i.e. $PATH).