@files = sort { my ($ad) = ( $a =~ /fwlog\.(\d+)\w+/ ); 
                my ($bd) = ( $b =~ /fwlog\.(\d+)\w+/ ); 
                $ad <=> $bd } @files;

##</code><code>##

foo1.tla
foo10.tla
foo100.tla
foo101.tla
foo102.tla
etc.

##</code><code>##

sub naive {
    sort {
        my ( $ma ) = $a =~ /foo(\d+)\.tla/;
        my ( $mb ) = $b =~ /foo(\d+)\.tla/;
        $ma <=> $mb
    } @_;
}

##</code><code>##

sub orcish {       
    my %c;
    sort {
        $c{$a} or ( $c{$a} ) = $a =~ /foo(\d+)\.tla/;
        $c{$b} or ( $c{$b} ) = $b =~ /foo(\d+)\.tla/;
        $c{$a} <=> $c{$b};
    } @_;
}

##</code><code>##

sub schwartzian {
    map { $_->[0] }
    sort { $a->[1] <=> $b->[1] }
    map { [ $_, /foo(\d+)\.tla/ ] } @_;
}

##</code><code>##

sub guttros {  
    map { substr( $_, 3 ) }
    sort
    map { sprintf( "%03d", /foo(\d+)\.tla/ ) . $_ } @_;
}

##</code><code>##

Benchmark: timing 5000 iterations of guttros, naive, orcish, 
schwartzian...
 guttros: 14 wallclock secs 
  (14.10 usr +  0.00 sys = 14.10 CPU) @ 354.61/s (n=5000)
 naive:  0 wallclock secs 
  ( 0.04 usr +  0.00 sys =  0.04 CPU) @ 125000.00/s (n=5000)
 orcish:  1 wallclock secs 
  ( 0.04 usr +  0.00 sys =  0.04 CPU) @ 125000.00/s (n=5000)
 schwartzian: 26 wallclock secs 
  (25.70 usr +  0.03 sys = 25.73 CPU) @ 194.33/s (n=5000)