I'm going to have to agree with holli on this and say Why not just use the hash?
If you must do something like this, know that it is in violation of some of the tenets of strict. Here's a version with limited un-strict-ness that does not use an eval:
#!/usr/bin/perl -l
use strict;
use warnings;
my %foo = (
_foo => 'FOO',
_bar => 'BAR',
_baz => 'BAZ',
zum => 'ZUM', # do not make this one a "real" variable
);
{ # limit the scope of the following 'no strict'
no strict 'refs';
# There be magic here! This will get the keys from %foo
# then attempt to remove the leading '_'. If it works,
# the *modified* key gets passed along; otherwise, the
# key is stripped from your processing list. We then
# dynamically create a variable with the name of the
# modified key and assign it the value of the unmodified
# key in the hash.
$$_ = $foo{"_$_"} for (grep { s/^_(.*)/$1/ } keys %foo);
}
# this is so we can speak the variable's name without Perl throwing a
+fit
no strict 'vars';
print for ($foo, $bar, $baz, $zum); # should produce one 'unitialized
+' warning
I don't agree that this is good practice. However, do what you must to get the task accomplished. I know too well what it is to maintain ugly code, even when the original author was me.
Ivan Heffner
Sr. Software Engineer, DAS Lead
WhitePages.com, Inc.
|