There is a more-or-less conventional way to do it, and then there are those that take advantage of perl's regular expression capabilities. Here's one that uses the 's' operator:

sub contains { # return true if $_[0] is contained in $_[1]
  my ($s, $t) = @_;
  for (split('', $s)) {
    return 0 unless ($t =~ s/$_//);
  }
  1;
}
print contains("dog", "good"), "\n"; # -> 1
print contains("food", "fodder"), "\n"; # -> 0
[download]

Update: The 'conventional' way would be to perform a count of each letter and compare counts:

sub contains {
  my ($s, $t) = @_;
  my (%sc, %tc);
  for (split('', $s)) { $sc{$_}++ };
  for (split('', $t)) { $tc($_}++ };
  for my $k (keys %sc) {
    return 0 unless $sc{$k} <= $tc{$k};
  }
  1;
}
[download]

While not the most efficient in this situation, this code exemplifies common types of data processing that is done with perl, so it's valuable to know about.