I posted this solution in your previous thread, but maybe I was late and you missed it. It does the job while skipping phone numbers. If you don't understand any of it, just ask.

#!/usr/bin/env perl
use 5.010; use strict; use warnings;

my $text = <<END;
Apt.302. 123-4567. 4021
207-555-1531 Apt. 987
END

$text = " $text "; # pad it so my regex will work if a number 
                   # appears at the beginning or end of line
$text =~ s|(       # start capturing in $1
            [^\d-] # any single character other than a digit or dash
           )       # end capture of $1
           (       # start capturing in $2
            \d\d\d # three digits in a row
            \d?    # zero or one digits (for a total of 3-4)
           )       # end capture of $2
           (       # start capture of $3
            [^\d-] # any single character other than a digit or dash
           )       # end capture of $3
         |$1<a href="\?unit=$2"><b>$2</b></a>$3|gx; # Replace captured
                                                    # sections with HT
+ML added 
$text = substr($text, 1,-1);  # trim off padding
say $text;
[download]