Need a better way to count input lines

Theo has asked for the wisdom of the Perl Monks concerning the following question:

This is the second Perl script I've done from scratch and I'm havin' a ball! I'd like some constructive criticism over all and in two specific areas.

The objective is to convert an MS Word document of the company phone list to an MS Excel spreadsheet for sorting and conversion to several web pages (of differing sort orders).

The Word file is saved to a text file where each person's info is spread over 3 lines. The script pulls everything into one line per person with tabs separating each item. Having been hand entered, the data is somewhat "dirty". Most of the regexp work is cleaning linefeeds, carriage returns and the random non word characters around the valid data.

Question 1: Is there a more perlish way to count the three lines per person? (What I'm doing looks so much like C.)

Question 2: If you look at the input data for "Bo Jo Le Much" you'll see there is no space after the comma separator and there is a CR (^M) after the first name in the output. If I insert a space after the comma in the input data, the CR in the output goes away. Why?

#!/opt/bin/perl
#perl 5.8.3 built for sun4-solaris-64int-ld
 
use warnings;
use strict;
 
my $T = "\t";
my $fname = "U1";
my $lname = "U2";
my $phone = "U3";
my $room  = "U4";
my $bldg  = "U5";
my $email = "U6";
my $temp  = "Ut";
my $count = 0;
 
print "\nLast   First   Phone   Bldg    Room    email\n";
 
 open FH => "<testdata.txt" or die "can't find the data file: $!\n";
#open FH => "<phonedata.txt" or die "can't find the data file: $!\n";
 
while (<FH>) {
   chomp;
   $count++;
   if ($count == 1){
        ($lname, $fname) = split(",");
        $fname =~ s/^\s+(.+?)\W*$/$1/;
   }elsif ($count == 2){
        ($phone, $temp) = split("/");           #line 28
        if ( ($temp =~ m[^\d.*]) || ($temp =~ m[--- --]) ){
            ($room, $bldg) = split(" ", $temp);
        } else {
            $bldg = $temp;
            $bldg =~ s/^(.*?)\W*$/$1/;      #$bldg includes a 'cr'
            $room = "";             #no room number, just bldg name
        }
   }elsif ($count == 3){
        $email = $_;
        $email =~ s/^(.*?)\W*$/$1/;
        $count = 0;
        pA();
 
   }else{
        print "Got an overflow - $count $_\n";
   }
}
print "\n" if ($count == 0);
close FH;

sub pA{print "$lname$T$fname$T$phone$T$bldg$T$room$T$email\n";}
-------------------------------------------------------
__Input__
Alanon, Bart
5590/EL
---- 
O'Lewis, John.
----/--- --
john 
Le Much,Bo Jo
3406/165 NS 
ed@a.nl 
Abe-Jen, Mar-Jo
3421/164D NS
cbest

__Output__
Last    First   Phone   Bldg    Room    email
Alanon  Bart    5590    EL
O'Lewis John    ----    --      ---     john
Le Much Bo Jo^M 3406    NS      165     ed@a.nl
Abe-Jen Mar-Jo  3421    NS      164D    cbest
[download]

-Theo-
(so many nodes and so little time ... )

Back to Seekers of Perl Wisdom