In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED] (Mike Ni) writes: >While I was playing with perl's built-in debuger >(perl -d), I notice that debugger doesn't seem >care too much about whether a variable is declared? > >For example, I would get response of "empty hash" if >I do a "x \%dummy" regardless dummy hasn't been >delcared yet. > >The same response "empty hash" if I dump a declred >hash before it is initialize. > >It is as if there isn't too much difference between an >undeclared variable & unitialized variable? > >Is this the way it is ? Or I am not doing right?
That is the way it is. The debugger evaluates your commands in a lexical scope within which "use strict" has not been declared. You can't make it be strict without editing the debugger (perl5db.pl). This is just as well, the debugger is a sandbox. The same goes for use warnings, but if you use the -w or -W flags instead, they are non-lexical and will extend to the debugger. Try the p command on an undefined value some time with -w enabled and you'll see a quite verbose and annoying message. One reason to use warnings instead of -w is to avoid that spew when debugging. -- Peter Scott http://www.perldebugged.com/ *** NEW *** http://www.perlmedic.com/ -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] <http://learn.perl.org/> <http://learn.perl.org/first-response>
