I'm having a little trouble understanding your goals here, so correct me if the following code does not address your issue. I am reading this as you wish to locate all occurrences of [Sidenote: *] and extract the unique text, displaying it surrounded by <div> tags. A (re)read of perlretut might be useful for you, but the following code does what I describe above.

use strict;
use warnings;

my $text = "C:\\letters.txt";

my @letters;
open IN, '<', $text or die "Can't open $text";
@letters = <IN>;
close(IN);
chomp @letters;

foreach my $indiv_note (@letters) {
    if ($indiv_note =~ /\[Sidenote\:\s(.*?)\]/) {
        print "<div>$1</div>\n";
    }
}
[download]

Note that the way I have written this, sidenotes cannot cross line boundaries, though doing this would be fairly trivial. Also note that if you want to include [ and ] in your posts, you should use html entities, i.e. &#91; and &#93;