Get the DB to tell you the maximum length of the field. Assuming a well-trimmed varchar field called 'reason':

select max( len( reason ) ) from table;
[download]

If the field is fixed length or can having space padding, then you'd need to trim them up. The functions for that tend to vary from DB to DB.

Actually even len/length/char_length varies from DB to DB. Of the two I've worked with (SQLite and MySQL), one returns the number of bytes for length where the other returns the number of characters. Neither provide len. MySQL provides char_length in addition to length to return the character count.

True laziness is hard work

...
$sth->execute;
my $all = $sth->fetchall_arrayref;
my @max = (0) x 3;
foreach my $row (@$all) {
    for (my $i = 0; $i < @$row; $i++) {
        $max[$i] = length $$row[$i] if length $$row[$i] > $max[$i];
    }
}
[download]

my $query = <<'SQL';
SELECT ticket_id, abstract, priority, maxlen_abstract
FROM
  (SELECT MAX(LENGTH(abstract)) AS maxlen_abstract FROM t1) AS s1
  CROSS JOIN t1;
SQL
[download]

In the real world I might get lazy and make the arguement that the high/medium/low info should be right after the ticket number - because it is important. Lining that up is easy.

In the world of doing it the way someone says to (also part of the real world I guess), check out the documentation for sprintf (perldoc -f sprintf), it'll help you pad stuff.

#!/usr/bin/perl -w
use strict;

my ($ticket,$text,$priority) = 
   ("Ticket567", "Critical server unpingable", "High");

printf "%-9s %-30s %s", $ticket,$text,$priority;
# Ticket567 Critical server unpingable     High
[download]

Invariably somebody is going to type in the "Gettysburg Address" instead of a simple problem abstract. If every line uses that column's huge width, then the whole report is going to look screwy. If you use fixed widths that work "almost all the time", then when this happens just one line of the printout doesn't line up nicely - rather than having huge amounts of white space in all of the other lines to accommodate this one weirdo line.

Another thing that happens, is that often folks will want to cut and paste lines from different report days together into some new document. If the report column widths are consistent day to day, this is easier.

It is not my place to say that your requirement is "wrong" or that you don't need it. But you asked: "Any help is greatly appreciated.", I'm just presenting a plausible, practical alternative from the "real world".

Update: In printf "%-9s %-30s %s", this could have been written as printf "%-10s%-31s%s", but it is usually a mistake to rely upon the field width to account for the blank space between columns. By putting an explicit space in between fields in the format spec, when a column "overflows" because it is too wide to fit into the minimum space, this guarantees that there still be at least one space between the columns. If we were say printing integers and had "%3i %4i", that space ensures that we won't windup with something like "12345123456789" instead of "12345 123456789". This is a fine point, but understanding this will save you grief on some reporting project. Putting that explicit space in the format spec also makes it easier to read, so its a "win-win".

@array1 = ("a", "a sdfasdf value" , "medium");
@array2 = ("b", "a  asdfa     asdfasdfsadfs value" , "medium");
@array3 = ("c", "a value" , "high");

use Text::ASCIITable;
$t = Text::ASCIITable->new();

$t->setCols('Id', 'Name' ,' Pri');

$t->addRow(@array1);
$t->addRow(@array2);
$t->addRow(@array3);

print $t;
[download]

.------------------------------------------------.
| Id | Name                             |  Pri   |
+----+----------------------------------+--------+
| a  | a sdfasdf value                  | medium |
| b  | a  asdfa     asdfasdfsadfs value | medium |
| c  | a value                          | high   |
'----+----------------------------------+--------'
[download]