I agree with your suggestion here. However, let me tell you a little about the genealogy of this bit of code, because there's another stage to come and I'd like a suggestion about how to complete this. For links I used to have something like this:

This worked fine for a long time. However, I would change occasionally the name of a script or move it to a different directory. For example, musing.cgi used to be called web_log.cgi. Then I'd have to go through my scripts and other places where script names were contained and change it. Last year I came up with the idea of entering the names of all of my scripts in a table in MySQL I named site_scripts. Now if I want to change the name of a script or location of a script, I change it in my site_scripts table in MySQL. I just use the same key name (e.g., musing) to retrieve it. This has worked well for me so far.

Now what I want to do is to add the particular ID for the script. For instance, for a musing I generally have musing_id as the CGI variable name. For it, I'd like to change '?musing_id=' to something less hard-coded to something like this:

To tighten this up more, and to reduce the interpolation you mention, I could add the ? and the = to the hash containing the id label. Of course, once I have this problem worked out, I can have a sub-routine for my hyperlinks: hyperlink('musing',$musing_id,$musing_heading)

That would be so much nicer, wouldn't it? It would be simple to set up another column in my site_scripts table in MySQL to hold the ID label. However, for some scripts I need more than one CGI variable: /cgi-bin/musing.cgi?musing_id=$musing_id&ref_id=$ref_id. That might also through off my hyperlink() idea. So, any suggestions on how I might organize multiple ID possibilities in MySQL or how to sort through them in Perl when deployed with a function?

You could put all the link particulars in a module so that changes can be made in one place. Here is a Links.pm to give you ideas. It would be better to go a step further and have the module read a configuration file, but I didn't want to spoil all your fun.

If you only have a handful of links to maintain, a database may be overkill. By putting this in a module, you should be able to make the change from a config file to database with little, if any, change in the calling scripts.

#!/usr/bin/perl

  package Links;

  use strict;
  use CGI qw(:standard);

  my $Links = {

    musings => {
      text => 'Musings',
      link => '/cgi-bin/musing.cgi',
      id_list => [
        'musing_id',
      ],
    },

    rants => {
      text => 'Rants',
      link => '/cgi-bin/rants.cgi',
      id_list => [
        'rant_id',
        'ref_id',
      ],
    },

    stuff => {
      text => 'Other Stuff and Junk',
      link => '/cgi-bin/misc.cgi',
      id_list => [
      ],
    },
  };

#-----------------------------------------------------------
sub make {

  my $key  = shift;
  my $vals = shift;

  (defined($Links->{$key})) or return;

  my $info = $Links->{$key};

  my @id_list;

  for my $id (@{$info->{id_list}}) {
    if (defined($vals->{$id})) {
      push @id_list, "$id=$vals->{$id}";
    }
    else {
      # what if?
    }
  }

  my $url = $info->{link};

  if (@id_list) {
    my $id_str = join('&', @id_list);
    $url = "$url?$id_str";
  }

  return a({-href=>$url}, $info->{text});
}

1;
[download]

#!/usr/bin/perl

  use strict;
  use Links;

  my $vals = {
    musing_id => 42,
    rant_id   => 1,
    ref_id    => 2,
    foo_id    => 3,
  };

  my $link = Links::make('musings', $vals) or die;
  print "$link\n";

  my $link   = Links::make('rants', $vals) or die;
  print "$link\n";

  my $link  = Links::make('stuff', $vals) or die;
  print "$link\n";
[download]