(Shrug...) What you've got now is clear and it works. What's not to luv?

Please give me code that is “clear,” any day, over code that is “clever.” I can look at your example and instantly comprehend how it works. I can also tweak it without too much fear of breaking it. That, to me, is a very compelling argument in favor of what you started with.

if($string = /^client\s+(\d+\.\d+\.\d+\.\d+)/) {
  $ipaddr = $1;
} else {
  die 'horribly';
}
[download]

You can specify multiple delimiters for 'split':

while (<FH>){
    print +(split /#|\s/)[1];
}
[download]

Pretty close. I think you want:

(split /#/)[0]
[download]

use Regexp::Common qw/net/;                                           
+                         
my $ip_regexp = qr($RE{net {IPv4});                                   
+                          
while(<DATA>)  {                                                      
+                                        
  if (/($ip_regexp)/)  {                                              
+                                               
   my $ip_address = $1;                                               
+                            
    print "$ip_address\n";                                            
+                          
 }                                                                    
+                         
}                                                                     
+                          
                                                                      
+                          
__DATA__                                                              
+                          
client 127.0.0.1#47560: query: host.example.com IN AAAA +
[download]

$ perl -wle 'print +(split /#/, "a#b")[1];'
b
[download]

You can also use undef where you don't want variable names:

my (undef, $a) = split ...
[download]

use strict;
use warnings;

while (<DATA>) {
    my $ipaddr = (split /\s+/, (split /#/, $_)[0])[1];
    print "ipaddr = $ipaddr\n";
}

__DATA__
client 127.0.0.1#47560: query: host.example.com IN AAAA +
[download]

prints:

ipaddr = 127.0.0.1
[download]

The first split is in a list context and there is no need to assign "junk". The 2nd split should also be put into a list context and then just take the 2nd thing via a "list slice", [1] subscript. See below..

#!/usr/bin/perl -w
use strict;

my $line = 'client 127.0.0.1#47560: query: host.example.com IN AAAA';

my ( $part1)   = split (/#/,$line);
my ( $ipaddr ) = (split(/\s+/, $part1))[1];

print "  part1:  $part1\n  ipaddr: $ipaddr\n";
__END__
Prints:
  part1:  client 127.0.0.1
  ipaddr: 127.0.0.1
[download]

(undef, my $ipaddr) =  split(/\s+/, $part1);
[download]

Update: Now of course, if you just want the IP address:

my $line = 'client 127.0.0.1#47560: query: host.example.com IN AAAA';

(my $ipaddr) = (split/[\s#]/,$line)[1];

print "$ipaddr\n";
__END__
prints:
127.0.0.1
[download]

You want the stuff between the first space and the first hash

$_ = "client 127.0.0.1#47560: query: host.example.com IN AAAA +";
(my $ip) = /\s(.*?)#/;
print $ip; #127.0.0.1
[download]

If you are sure about the format of your input data (order of fields, number of whitespaces, ...), you can try something like this:

my $line = 'client 127.0.0.1#47560: query: host.example.com IN AAAA +'
+;
my ( $ip, $host ) = ( split /[ #]/, $line, 6 )[1,4];

print "IP: $ip, HOST: $host\n";
[download]