Perhaps I should have been more clear in my original comment when I said you could split on "DOMAIN\|".

use strict;
use warnings;

my @domain_names;

open ("FILE", "< ./file.csv")
  or die ("Can't open file $!");
while (my $line = <FILE>) {
  my @parts = split 'DOMAIN\|', $line;
  # throw away the stuff before DOMAIN|
  shift (@parts);
  # now get what's between the beginning
  # of the line and the | -- the domain, correct?
  my $part = shift (@parts);
  @parts = split '\|', $part;
  my $domain = shift (@parts);
  push @domain_names, $domain;
}
[download]

And, of course, if you only have one DOMAIN|test.com in every file you could add an if to check if there's one on the current line using a regular expression, and then a last to exit the loop once you found one.