You could slurp in said files (read everything, ignoring newlines/carriage returns), then split on whitespace:

# slurp file contents
my $contents;
{
  local $/ = undef;
  open(my $handle, '<', "filename") or die("error: $!");
  $contents = <$handle>;
  close $handle;
}
# split into array on consecutive whitespace
my @words = split /\s+/, $contents;
[download]

As far as removing the common words, you could use indexes from List::MoreUtils to get the indices of the common words to remove them, OR get the indices of the not common words to add them to another array. One of the more experienced monks may have a better solution, though.