You can try something like:

use strict;
use warnings;

my @replicas;
while (<DATA>){
  next if (! /^\d+/);
  /replica\s+"(.+)"/;
  push @replicas,$1;
}
__DATA__
2005-04-01     root        replica "ml_v_dialer"
2004-06-22     root        replica "pu_v_dialer"
2006-02-11     ccvob01     replica "rd_v_dialer"
  "v_dialer replica for Redmond"
2003-11-25     root        replica "v_dialer_drcalvin"
[download]

Update: Another possibility (if the file to process is not too big):

use strict;
use warnings;
{
  local $/;
  my @replica = <DATA> =~ /replica\s+"(.+)"/g;
}
[download]

citromatik

The third would return rd_v_dialer" "v_dialer replica for Redmond instead of the wanted rd_v_dialer. You need the non-greedy quantifier with "?" to match the nearest closing '"' (as little as possible). So it should be /replica\s+"(.+?)"/;.

---

Open source softwares? Share and enjoy. Make profit from them if you can. Yet, share and enjoy!

Hmmm, Are you sure? the dot "." matches everything but "\n", so it will not expand through newlines (see perlretut):

use strict;
use warnings;
use Data::Dumper;

{
  local $/;
  my @replica = <DATA> =~ /replica\s+"(.+)"/g;
  print Dumper \@replica;
}

__DATA__
2005-04-01     root        replica "ml_v_dialer"
2004-06-22     root        replica "pu_v_dialer"
2006-02-11     ccvob01     replica "rd_v_dialer"
  "v_dialer replica for Redmond"
2003-11-25     root        replica "v_dialer_drcalvin"
[download]

Outputs:

$VAR1 = [
          'ml_v_dialer',
          'pu_v_dialer',
          'rd_v_dialer',
          'v_dialer_drcalvin'
        ];
[download]

And:

use strict;
use warnings;
use Data::Dumper;

my @replica;
while (<DATA>){
  next if (! /^\d+/);
  /replica\s+"(.+)"/;
  push @replica,$1;
}
print Dumper \@replica;

__DATA__
2005-04-01     root        replica "ml_v_dialer"
2004-06-22     root        replica "pu_v_dialer"
2006-02-11     ccvob01     replica "rd_v_dialer"
  "v_dialer replica for Redmond"
2003-11-25     root        replica "v_dialer_drcalvin"
[download]

Prints the same:

$VAR1 = [
          'ml_v_dialer',
          'pu_v_dialer',
          'rd_v_dialer',
          'v_dialer_drcalvin'
        ];
[download]

citromatik

On the face of it you need simply match the first double quote character then capture all the following characters up to the next double quote. See perlretut and perlre for the regex documentation.

If you have trouble putting a regex together, write something (anything actually), then post your attempt in a follow up question and we will help more.

$ cat replica
my @dialers;
while (<DATA>) {
  chomp;
  push @dialers, $1
    if /replica\s+"([^"]+)/;
}

print $_, "\n" for @dialers;

__DATA__
2005-04-01     root        replica "ml_v_dialer"
2004-06-22     root        replica "pu_v_dialer"
2006-02-11     ccvob01     replica "rd_v_dialer" "v_dialer replica for
+ Redmond"
2003-11-25     root        replica "v_dialer_drcalvin"

$ perl replica
ml_v_dialer
pu_v_dialer
rd_v_dialer
v_dialer_drcalvin
[download]

  push @dialers, $1
    if /replica\s+"([^"]+)/;
[download]

  (my $dialer = (split)[3]) =~ s/"//g;
  push @dialers, $dialer;
[download]