Ah-ha.
I'm not entirely crazy, just extremely forgetful.
I knew I had done a script like this a couple years ago
and I knew it used backticks and I knew it had a timeout feature.
The part I forgot was it used Tk::ExecuteCommand to
manage the process. It also shows the output of the
simulation. and it provides the user with an interactive
way to kill the simulation in the middle.
This script seems to timeout properly and kill the underlying
process so as to not leave any zombies around:
#!/usr/bin/perl
use warnings;
use strict;
use Tk;
use Tk::ExecuteCommand;
my $oscmd="echo \"hello\"; sleep 30000; echo \"world\";";
my $top=new MainWindow;
my $ec = $top->ExecuteCommand()->grid(-row=>1,-column=>1);
$SIG{ALRM} = sub { $ec->kill_command; };
alarm(10);
$ec->configure(-command => $oscmd);
$ec->execute_command;
$ec->update;
I need to add gui support anyway, so I'll just tweak
my script to tee its output during backtick commands
and use Tk::ExecuteCommand to handle the OS command.
There's still the problem of buffering, but I'll deal
with that later.
And apparently I need to read up on pipes in case
this bites me again and I can't use a gui.
Greg
> Don't do that.
>
> In general, you're going to run into a whole mess of problems with the
> kind
> of thing you want to do, and your best is to control everything and avoid
> the shell. Pipe opens, as Uri suggested, are the way to go, unless you
> need
> bi-/tri-directional support in which case there are
> IPC::Open[23]/IPC::Run.
>
> But often you can get away, from perlipc:
>
> On most Unix platforms, the "CHLD" (sometimes also known as "CLD")
> sig-
> nal has special behavior with respect to a value of 'IGNORE'.
> Setting
> $SIG{CHLD} to 'IGNORE' on such a platform has the effect of not
> creat-
> ing zombie processes when the parent process fails to "wait()" on
> its
> child processes (i.e. child processes are automatically reaped).
> Call-
> ing "wait()" with $SIG{CHLD} set to 'IGNORE' usually returns "-1"
> on
> such platforms.
>
--
_______________________________________________
Boston-pm mailing list
[email protected]
http://mail.pm.org/mailman/listinfo/boston-pm
