parsing real CSV data is an art, best not done with regular expressions. In your case, if it is that simple, use split:

The last field now gobbles up the rest of the line. When you'd use a real parser, it might be something like:

my $csv = Text::CSV_XS->new ({ binary => 1 });
foreach my $line (@out) {
    my @f = $csv->parse ($line);
    @f > 10 and $f[9] = join ",", splice @f, 9, $#f;
[download]

But above approach will loose quotation