Good day Monks. I have been using a little subroutine to do inserts on MySQL (via DBI and Pg) and get the inserted id:

sub insertsql {
    my ($dbh,$table,$data,$ignore) = @_;
    my @qm;
    my @keys;
    my @values;
    my $i = -1;
    foreach my $k (keys %$data) {
        if (defined($data->{$k})) {
            $i++;
            $keys[$i] = $k;
            $values[$i] = $data->{$k};
            $qm[$i] = '?';
        }
    }
    my $keylist = join(",",@keys);
    my $qlist = join(",",@qm);
    my $sqlstatement = "insert into $table ($keylist) values ($qlist)"
+;
    if ($ignore) {
        my $sqlstatement = "insert ignore into $table ($keylist) value
+s ($qlist)";
    }
    
    my $sth = $dbh->prepare($sqlstatement);
    #$sth->execute(@values) || die "putsql could not execute MySQL sta
+tement: $sqlstatement $sth->errstr";
    $sth->execute(@values) || die $sth->errstr;
    $sth->finish();
    return $dbh->{'mysql_insertid'};
}
[download]

I'm now working with Postgres and, alas, there is no 'mysql_insertid' for postgres, and it doesn't provide such an easy way to get the inserted id.

I found this thread on Stack Overflow where somebody suggests using a separate statement SELECT currval(pg_get_serial_sequence('[tablename]','id')) but that returns nothing. They also suggest adding returning id to the end of the insert statement. But I can't figure out how to make that work in the context of the subroutine.

Can anyone help?