Bah! Why use regexes when there's a faster function designed for the purpose? =)

#assuming $n contains the number ...

my $first_n = substr($string, 0, $n);

#to grab $n off the end

my $last_n = substr($string, -1, $n);
[download]

Use substr when possible, it can be many times faster then a simple regex:

my $a = "abcdefghijklmnopqrstuvwxyz";

use Benchmark;
timethese(1000000, {
  's1' => sub { s1($a, 5) },
  's2' => sub { s2($a, 5) },
});

sub s1 {
  my $a = $_[0];
  $a =~ s/^.{$_[1]}//;
  return $a;
}

sub s2 { 
  substr($_[0], $_[1]);
}

Results:
        s1:  8 wallclock secs ( 7.95 usr +  0.09 sys =  8.04 CPU)
        s2:  2 wallclock secs ( 2.54 usr +  0.01 sys =  2.55 CPU)
[download]

The other examples all handed off the first $n characters. But they didn't remove those first $n characters from the original $string, as the question required.

As always, there is more than one way to do it. For each of the examples below, assume the following setup:

my $string = "1234567890abcdefghijABCDEFGHIJK";
my $chars;
my $n = 2;
[download]

With substr:

$chars = substr( $string, 0, $n );
substr( $string, 0, $n ) = "";
[download]

Another with substr:

( $chars, $string ) = ( substr($string,0,$n), substr($string,$n) );
[download]

And probaby the best substr solution:

$chars = substr ($string, 0, $n, "");
[download]

With a substitution s/// regexp:

$string = s/^(.{$n})(.*)$/$2/s;
$chars = $1;
[download]

With split:

($chars, $string) = split /^(.{$n})/, $string, 1;
[download]

With a pattern match (m//):

($chars, $string) = $string =~ /^(.{$n})(.*)$/s;
[download]

With unpack (Not for the faint of heart):

( $chars, $string ) = unpack "a$n a@{[length($string)-$n]}", $string;
[download]

my ($x,$y) = ($string =~ /^(.{n})(.{n})/);
[download]

my ($x, $y, $rest) = ($string =~ /^(.{n})(.{n})(.*)/);
[download]