This is the kind of place where aspect-oriented programming is awesome. If you're using Moose, Moo or another OO library that supports method modifiers, then it's probably just a matter of creating a role with a before execute method modifier and then applying it to all your command classes.

Otherwise, if you have a base class for all your command classes, you could try creating a method called before_execute and then manually modifying all your execute methods to call $_[0]->before_execute at the start.

Hi tobyink,

thank you for answering that fast. I have to admit that I hoped it. ;-) (a ++ for that)

The first suggestion is not an option for me right now. The second approach is which I'm doing right now. But I wanted to avoid the call of before_execute in every execute method. I really hoped that there is another callback which I haven't found.

Once again. Thank you.

Best regards
McA

"I really hoped that there is another callback which I haven't found."

The relevant code is App::Cmd::execute_command. As you can see, nothing happens between validate_args and execute that you could possibly hook onto.

One alternative approach could be to subclass App::Cmd, overriding that method with your own implementation that adds extra hooks between validate_args and execute:

{
  package My::App::Cmd;
  use base 'App::Cmd';
  sub execute_command {
    my ($self, $cmd, $opt, @args) = @_;
    local our $active_cmd = $cmd;
    $cmd->validate_args($opt, \@args);
    $cmd->before_execute($opt, \@args);
    $cmd->execute($opt, \@args);
  }
}
[download]

Then use My::App::Cmd instead of App::Cmd.

perl -E'sub Monkey::do{say$_,for@_,do{($monkey=[caller(0)]->[3])=~s{::}{ }and$monkey}}"Monkey say"->Monkey::do'