First comment is that I'd suggest an array of hashes over an array of arrays because I find it way easier to read a hash than an array when it's a list of disparate objects. Your outer array is really a list of courses - lists are what arrays are for. Your inner arrays are all a single course with a bunch of attributes - attributes are what hashes are good at. However, that probably doesn't change how to sort in your mind, so it's not going to solve your problem.

To sort on multiple fields, you just need to set up your sorting on those fields. It should be as simple as:

  @courselist = sort {
    $a->[0] cmp $b->[0] or
    $a->[1] cmp $b->[1]
  } @courselist;
[download]

Hope that helps,

People vary on whether a hash style presentation would be visually clearer. For me it's far clearer to have just the data, lined up in columns. Having keys repeated on each line is, for me, visual clutter which I have to screen out, and it doesn't add information in a case like this where the type of each field is pretty much clear from the content.

Conceptually, it's a good rule to have like things in arrays, and unlike things in hashes, and I think that rule gets fairly wide agreement. Perceptual styles are different. What may be helpful markers for tanktaus are visual noise that needs to be screened out for slower readers like me. One man's sushi is another man's bait.

You already have a couple of suggestions for how to solve your problem, but since no one pointed you to resources where you can get more information about sorting, here are a few:

• perldsc
• perllol
• sort
• Complex sorting
• Resorting to Sorting

@courseindex = 
  sort { $a->[0] cmp $b->[0] || $a->[1] cmp $b->[1] } @courseindex;
[download]

I do commend Gnat53's attempt at the solution and also bobf's helpful resource reply.

Your description of what you wanted was pretty good:

I'd like to sort it by $Institution and within that sort by $CourseNumber (and possibly other values)

You just needed to break it down into pieces that perl can understand. "sort it by" you correctly translated as $$a[n] cmp $$b[n], but you left out the more difficult-to-translate "within that". It helps to ask yourself "what does this really mean, on a level that perl could understand?", the answer, in this case, being "if the institutions are different, sort by them, otherwise, sort by course number". Which leads pretty directly to the solution proposed to you. Extending it to handle the "and possibly other values" should indicate a loop to you, something like:

sort {
   my $compare;
   for (my $i = 0; !$compare && $i < @$a; ++$i) {
      $compare = $a->[$i] cmp $b->[$i];
   }
   return $compare;
}
[download]

My apologies then, and enjoy the excellent answers from Tanktalus et al. Good on you for building your own site - I hope you are able to get to where you want to be with the help of Perlmonks!

PS: I haven't been to MA for about 6 years, but I miss visiting Boston and the rest of New England - next time I come I'll try and pop in to your shop :)

use Data::Dumper;
my @array = ();
$array[0] = [1];
$array[2] = [2];
@array = sort { $$a[0] cmp $$b[0]} @array;
print Dumper \@array;
[download]

use strict;
use warnings;
use Data::Dumper;

my @courselist = ( 
    ["UM", "CS 34", ],
    ["AC", "PHIL 13", ],
    ["UM", 'BIO 567', ],
    ["UA", 'BIO 999', ],
    ["UM", 'BIO 1', ],
);  


@courselist = sort {$$a[0] cmp $$b[0] or $$a[1] cmp $$b[1]} @courselis
+t;
print Dumper(\@courselist);
[download]