How do I pull n characters off the front of a string?

[Anonymous Monk]

If I have a string that I am parsing to put into a database and the same n characters are being pulled off a string, I want to pull those chars from the string and leave the rest of the string. I want the first n chars put into $x, then removed from the string. Then the first n chars in the new string put into $y.

Originally posted as a Categorized Question.

[arturo]

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);

[davido]

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;

With substr:


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

Another with substr:


( $chars, $string ) = ( substr($string,0,$n), substr($string,$n) );

And probaby the best substr solution:


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

With a substitution s/// regexp:


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

With split:


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

With a pattern match (m//):


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

With unpack (Not for the faint of heart):


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

[Aristotle]

unpack:

($chars, $string) = unpack "a$n a*", $string;

The four parameter substr is clearly the cleanest solution though.

[perlmonkey]

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)

[cbraga]

Or this way:

#assuming $n contains the number
$string =~ s/^(.{$n})//;
$first_n = $1;

Originally posted as a Categorized Answer.

[xtrem]

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

This will put the 1st n char into $x and the 2nd n chars into $y without touching $string... if you want the rest of the chars:

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

if you don't like regexp's you can allways try substr

[Masem]

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

Originally posted as a Categorized Answer.