big command line programs...

[smckittr]

..such as ms dos command promt or unix, etc.
How do they figure out what to do with the command you gave them? for example if you type in exit how does it know to exit. Do they use a giant switch statement or some other method?
Just curious
Thanks
Scott

[Feis]

Kinda. First off please feel free to correct me as Im sure I will make a mistake or two in the finer details. In each commandline (@ least Unix or windows) you have a PATH variable. It tells the commandline where to look if you just enter a command. so when u enter exit, it will search locations in the PATh variable (in Windows its system32) for the command, and then execute it. That why if u want to add a command or shell script so that u can run it from anywhere, u can either dump it in a folder such as system32, or you could add its location to the path variable itself. Hope that explains your ? without too many blatant errors

[smckittr]

makes sense to me,
thanks
Scott

[teknomage1]

I'm pretty sure the way a shell's top loop looks is like this

//psuedo code
read_input_line();
tokenize();
if (expandable)  //for aliases and history substitutions
{
  rewrite_exapandable_tokens();
}

//Compare the first token to a list of built in commands
foreach ( builtin in builtin_list ) {
 if ( command_token == builtin )
 {
   execute->command_token( parameter_tokens );
   break;
 }
}

//If that didn't work try executing it as a filesystem command
try_execute_file_in_path( command_token );

//Finally just fail
signal_error();