PostgreSQL pl/perl trigger, differentiate null vs empty

Question

I've been trying to speed up a generic auditing trigger that I had written in pl/pgsql some time back. On update, it generates a list of columns in the table being updated and inserts rows in the auditing table recording any changes (by table, column, data before, data after, etc). The same trigger function is used across multiple tables.

I'm toying with pl/perl as it seems to be much faster for the task at hand, but I seem to have run into a problem on differentiating between NULL and empty string ('') value in the database.

In my mind if a column goes from NULL to an empty string (or vice versa), that's a change that I need to record. However, using the available new/old column references ($_TD->{new/old}{$columnName}), I can't seem to differentiate between columns that are actually null and those that contain the empty string. Columns that I know are null are being caught by both the empty and the undef checks, as are columns that I know are empty.

    if($_TD->{new}{$column} eq '') {
        elog(NOTICE, "New value in column $column is empty");
    }
    if($_TD->{old}{$column} eq '') {
        elog(NOTICE, "Old value in column $column is empty");
    }
    if($_TD->{new}{$column} eq undef) {
        elog(NOTICE, "New value in column $column is not defined");
    }
    if($_TD->{old}{$column} eq undef) {
        elog(NOTICE, "Old value in column $column is not defined");
    }

I suspect that I'm doing something silly here, but maybe I'm trying to do something that I simply can't do. Any advice?

Edit - Using Postgres 8.4.4 for what it's worth

Edit - After looking at filiprem's post below (and lots more testing), I ended up with this, which seems to be working:

    my %newrow = %{$_TD->{new}};
    my %oldrow = %{$_TD->{old}};

    my $valChanged;
    while (($column,$value) = each %newrow) {

        $valChanged = 0;

        if($newrow{$column} ne $oldrow{$column}) {
            $valChanged = 1;
            elog(NOTICE, "Values in column $column differ.  New: $_TD->{new}{$column}, Old: $_TD->{old}{$column}");
        }   
        elsif(!defined($newrow{$column}) && defined($oldrow{$column})) {
            elog(NOTICE, "New row contains nulled out field");
            $valChanged = 1;
        }
        elsif(defined($newrow{$column}) && !defined($oldrow{$column})) {
            elog(NOTICE, "New row contains newly populated field");
            $valChanged = 1;
        }

        if($valChanged) {
            ### Update audit table
        }
    }

It catches the difference between empty strings and NULLs and logs them all appropriately in the audit table.

Answer 1

You did not specify PostgreSQL version. On 9.0.5 I have observed the same behavior (not sure if it's a bug, see comments below).

This is easy to workaround - you can first test for definedness to catch NULL, and if it passed, test for empty string.

if ( not defined $_TD->{ new }{ $column } ) {
    elog( NOTICE, "New value in column $column is not defined" );
}
elsif ( $_TD->{ new }{ $column } eq '' ) {
    elog( NOTICE, "New value in column $column is empty" );
}
if ( not defined $_TD->{ old }{ $column } ) {
    elog( NOTICE, "Old value in column $column is not defined" );
}
elsif ( $_TD->{ old }{ $column } eq '' ) {
    elog( NOTICE, "Old value in column $column is empty" );
}