Roman.pm has saved my bacon several times. The best was when I had to replace translation lists marked a, b, c, ... with i, ii, iii, ...

Roman did exactly what I needed. Trivial example:

use Roman;
while (<DATA>) {
    chomp;
    last if (/^$/);
    my ($key, $data) = split(/\s+/, $_, 2);
    $key =~ s/([a-z])/roman(1+ord($1)-ord('a'))/eg;
    print join("\t", $key, $data), "\n";
}
exit;
__DATA__
a   Johan 
b   Gambolputty
c   von Haukopft
d-y (much deleted)
z   of Ulm
[download]

which gives:

i      Johan 
ii     Gambolputty
iii    von Haukopft
iv-xxv (much deleted)
xxvi   of Ulm
[download]

Not bad for so litle code and effort, eh?

--
foreach(split('',"\3\3\3c>\0>c\177cc\0~c~``\0cc\177cc")) {$_=unpack('B8',$_);tr,01,\40#,;print$_,"\n";}##IYDKINT!