Let the database do the work.

$sql = <<SQL;
SELECT DISTINCT
  student_id,
  student_name,
  student_address,
  student_gpa,
  college
FROM school
WHERE minor = 'CS'
ORDER BY college, student_id
SQL
[download]

Once you've got that, you could set up your output loop so that it compares the current value of college to the one in the previous record; if it's different, you print a header with the new information. Something like this (untested):

my $college_last ='';
while (@row = $sth_report->fetchrow_array) {
  # assign $student_name etc.
  print "<h1>$college</h1>" if $college ne $college_last;
  $college_last = $college;
  # print the actual record contents
}
[download]

Now, some handy short-cuts in Perl programming:

Instead of using the ternery operator twice on each column for each row, you could just handle empty columns in the first check:

$student_name = (defined $row[1]) ? $row[1] : ' ';
$student_blah = .... ad naseum
[download]

while (my @row = $sth_report->fetchrow_array) {
   @row = map { $_ || ' ' } @row;

   # .... rest of code
}
[download]

my %college;
while (my @row = $sth_report->fetchrow_array) {
   @row = map { $_ || ' ' } @row;
   my $college = pop @row;
  
   push $college{$college}, [@row];
}
[download]

foreach my $college (keys %college) {
    print "$college:\n";
    foreach my $row (@{$college{$college}}) {
        print "\t", join(', ', @$row),"\n";
    }
}
[download]

jeffa

L-LL-L--L-LL-L--L-LL-L--
-R--R-RR-R--R-RR-R--R-RR
B--B--B--B--B--B--B--B--
H---H---H---H---H---H---
(the triplet paradiddle with high-hat)

My favorite idiom for defaulting a set of values is:

# psuedocode
$_ = DEFAULT for grep !CONDITION, @VALUES;

# as it applies in this situation
$_ = ' ' for grep !$_, @row;
[download]

-Blake

The problem with !$_ (and jeffa's map example), is that it tests for truth, so you also end up substituting any values that match 0 instead of only those that are undef.

Perhaps you meant to use !defined instead?

    --k.

You could modify your SQL to use "GROUP BY college" and "ORDER BY student_id". Also, you could use Perl's sort function. This will negate the need for your loops and be much cleaner code.

One method i've used to do something similar is using the 1st column as a hash key, and the rest of the columns pushed on to an array.

#fetch the rows into an array reference
my %rethash;
while (my @row = $sth->fetchrow_array){  
    # for a multicolumn query, the 1st column becomes the hash key
    push @{$rethash{shift(@row)}}, @row;
}
[download]

I dont see why it couldnt be rearranged to do achieve what you need in order to provide a more direct data structure, ie use the college as the hash key then structure the rest of your data underneath it (as a hash or array).