#!/usr/bin/perl -w
use strict;
use warnings;
use RTF::Writer;
use Unicode::Subscript ':all';
use Text::CSV;
#use Unicode::Subscript qw(subscript superscript);

if(@ARGV != 2) {
    print STDERR "Usage: authorship_APP.pl input(.csv) output(.rtf)\n"
+;
    exit(0);
}

my ($inf, $outf)=@ARGV;
my $all_affi;
my %hash;
my @all_name;

my $count=1;

my @rows;
my $csv = Text::CSV->new ( { binary => 1 } )  # should set binary attr
+ibute.
                 or die "Cannot use CSV: ".Text::CSV->error_diag ();
 
open my $fh, "<", $inf or die "$inf: $!";
while ( my $row = $csv->getline( $fh ) ) {
    if($row->[0]=~/^\d/){
        my @a_num;
        for(my $i=3; $i<=7; $i++){
            if($row->[$i]=~/\w/){
                my $tmp=$row->[$i];
                if(! exists $hash{$tmp}){
                    $hash{$tmp}=$count;
                    push(@a_num, $hash{$tmp});
                    $all_affi.=$count.$tmp."\. ";
                    $count++;
                }
            }else{
                last;
            }
        }
        my $a_num=join"\+", @a_num;
        if($row->[2]=~/\w/){
            push(@all_name, "$row->[1] $row->[2]".superscript($a_num))
+;
        }elsif($row->[1]=~/\w/){
            push(@all_name, "$row->[1]".superscript($a_num));
        }else{
            print STDERR "Error! Wrong first name for this row:\n$row\
+n"; exit;
        }
    }
}
$csv->eof or $csv->error_diag();
close $fh;

my $all_name=join", ", @all_name;

print STDERR "There are ". scalar(@all_name)." authors and ".scalar(ke
+ys %hash)." affiliations in this list.\n";

my $final= "$all_name\n\n$all_affi\n";
#my $test='This algorithm is O(n' . superscript(3) . ')';
my $rtf = RTF::Writer->new_to_file($outf);
$rtf->prolog( 'title' => "Greetings, Naomi" );
$rtf->number_pages;
$rtf->paragraph(
    $final
);
$rtf->close;
[download]

I believe those characters you are thinking of as superscripts in the pdf document are actually just regular characters being displayed in a different font and different baseline. See this short discussion from stackoverflow. So you won't be able to get all the information into the plain text, instead you'll have to somehow encode the proper font formatting to send along with the text. Presumably the 'font' method in RTF::Writer might help do the job. But other monks may have better ideas.

I should mention that there is no need to rely on ".rtf" file. If there are any ways to generate the authorship list (including the superscrtipt comma) and could let me copy the context to microsoft word (.doc) file, it would be fine... Whatever format the output is... Thanks.