Instead, you should declare the variables with use vars: use vars qw($opt_h $opt_s); (In perl5.6, you could declare them with our instead.)

Alternatively, you could switch to Getopt::Long, which allows you to specify whatever variables you want to get the values of the options. It has other useful features as well.

Thank you. Yes, that's exactly what I was doing. I've incorporated the change into the posted code.
___________________
Kurt

my @theseWords = undef; is nonesense. Arrays are not undef. Only scalars are undef. It's being converted to (), which is what my @theseWords; by itself would do anyway.

Now for the hash you are using (): my %Count = (); again, that's unnecessary since it's created empty in the first place.

Now look at my ( $line, $word, $count, $LineNum ) = undef; what do you think that does? Since the undef is converted to an empty list (someone correct me if I'm wrong, please), then you have a list of 4 lvalues being assigned an empty list. That is treated like (undef, undef, undef, undef) which again is what my would have done anyway.

More subtle, little-known issue: The $VERSION should be a number of a v-string, not a string. See Everything you wanted to know about Module Version Numbers and Checking for the issues and compatability things.

open ( IN, $file ); ... or die;?

Re: $word = $opt_s; Word(); That is nasty!. The variable $word is being used for loop control in a couple places, and the same variable is used for passing a parameter to this function. Pass a parameter instead of using a global! The only occurance of $word other than the ones which are loop control (set within the loop) is one use within Word(). Say: Word($opt_s) if $opt_s; and use the parameter in Word with my $word= shift;.

Meanwhile take $word out of the globals at the top and add it to each foreach loop instead. With strict turned on, you will find these.

undef @theseWords;
    @theseWords = split( / /, $line );
[download]

s/\s\s*?//g I'll let japhy tell you about that one. I think you want \s+?. But, if you split on whitespace (see docs on split) you don't need to do that at all! Extra blanks are no trouble.

       $Line{$word} =~ m/(\d*?)$/;
        if ( $1 == $count ) {
                next;
            } else {
                $Line{$word} .= ", $count";
[download]

push @{$Line{$word}}, $count
   unless exists $Line{$word} && $Line{$word}[-1] == $count;
[download]

Keep on coding...
—John </code>

I re-posted the code using most of the suggestions. With regard to the version thing, I primarily include a version number for my own use, I wasn't trying to adhere to a standard. I didn't even know that there was a standard, to be honest. The sort block was a remnant from when I was going to try to keep the case of the words, rather than converting everything to lower case. I was going to do sort { lc($a) cmp lc($b) }, but ended up dropping the idea cuz it was getting too messy. (Dang, I had a remnant thing in my Fibonacci generator, too. Sorry 'bout that.)

I tried to use your code for creating the list of line numbers, but I kept getting a trailing comma with the output. Any suggestions on that?
___________________
Kurt

I tried to use your code for creating the list of line numbers, but I kept getting a trailing comma with the output. Any suggestions on that?

Easy: given $x is a list ref (stored as a hash element in the real program), say print join (',',@$x);

re $VERSION, yea, it's a standard. You found the node on it? Short version for those just tuning in: you can specify a required version number when loading a module and it dies if the module found is too old.

re preserving case but sorting case insensitive: look up the Swartzian Transform (spelling something like that).

I'll go over your code again later tonight, perhaps.

Keep it up!

—John

Here's mine, from a discussion on OO in January 1996.

Notice that I used a totally different mechanism to prevent duplicate entries for one line.

sub input
  {
  while (<>) {
     # parse the line into words
     my %temp;
     while (/\b\w+\b/g) {
        ++$temp{$&};
        }
     # add the found words to my global results list
     my $key;
     while (($key)=each(%temp)) {
        push (@{$concord{$key}}, $.);
        }
     }
  }


sub output
  {
  my @words= sort keys %concord;
  my $key;
  foreach $key (@words) {
#     print "$key @{$concord{$key}}\n"
     print ("$key ", join (', ', @{$concord{$key}}), "\n");

     }
  }

input;
output;
[download]