When it's just a matter of giving perl a couple of file names, one for input and one for output, I really prefer having that stuff all on the command line, rather than requiring the perl script itself to prompt for an interactive response from the person running the script. There are a few ways to do this:

• expect input and output file names as command line args, which show up in @ARGV:

  #!/usr/bin/perl
  
  my $Usage = "$0  input_file.name  output_file.name\n";
  ( @ARGV == 2 and -f $ARGV[0] ) or die $Usage;
  
  open( IN, "<", $ARGV[0] ) or die "$ARGV[0]: $!";
  open( OUT, ">", $ARGV[1] ) or die "$ARGV[1]: $!";
  
  while (<IN>) {
      tr/a-zA-Z/A-Za-z/;  # invert case, just for fun
      print OUT;
  }
  [download]

• use output redirection in the shell command (and just an input file-name arg, or input redirection, or a pipe), which means the perl script does not need to explicitly open anything:

  #!/usr/bin/perl
  
  my $Usage = "Usage: $0 [<] input.file > output.file\n";
  die $Usage if (( @ARGV > 0 ) ^ ( -t STDIN ));
  
  while (<>) {
      tr/a-zA-Z/A-Za-z/; # invert case, just for fun
      print;
  }
  [download]

(In the second example, the "if" condition will cause the script to die with with usage summary if you run it with no args and STDIN is not coming from a pipe or redirection.)

As for your usage of "tr///", I wonder what you're trying to accomplish with  tr/a-zA-Z/0-310/ because that really means:

  a -> 0  \
  b -> 1   \ "0-3" defines a continuous character range
  c -> 2   / starting at "0", ending at "3", incl. "1" & "2"
  d -> 3  /
  e -> 1
  fg-zA-Z -> 0
[download]

When I see a dash used between two numerics like that with the tr/// operator, my first impression is that the intent is probably different from the actual result...