I wonder how you tell the folders and fields apart, why you would want a CSV file with a variable number of values per line, and what you are going to do with the %BIGLIST when you've created it. I think your datastructure might be overly complex.

But lets say that you REALLY want a tree structure for this kind of thing:

sub add_line {
   my ($tree,@line) = @_;
   my $first_element = shift @line;
   if (is_a_folder_title($first_element)) { # TODO: implement is_a_fol
+der_title()
     $tree->{$first_element} ||= {}; # create empty hashref if none ex
+ists, yet
     add_line($tree->{$first_element},@line);
   }
   else {
     $tree = [ $first_element, @line ];
   }
}
[download]

To get the values out again:

sub get_elements {
   my ($tree,@folder_names) = @_;
   return $tree unless @folder_names;
   my $subtree  = $tree->{shift(@folder_names)};
   return get_elements($subtree,@folder_names);
}
[download]

If you have a seperate list of fields and folders, it's a lot easier to use a "simple" hash of arrays instead:

$directories{ join("/",@folder_names } = \@fields;

my @fields = @{ $directory{ join("/",@folder_names) } };
[download]