Question

I have two columns from a table; fromid and toid, which contain the same user id on occasions.

I have also a dropdown select box which selects all the ids from both these columns.

My problem is that the dropdown selects all users, but also duplicates when one user is in both columns.

What I would like is to select only a user once and give one of the columns a priority or MAX(is that it?) if the user appears in both columns. Here is the code I have so far:

$sql="
                            SELECT
                                DISTINCT u.".($config->realnames ? "name" : "username")." AS displayname,
                                u.id 
                            FROM
                                (#__users AS u
                            INNER JOIN
                                #__uddeim AS um
                            ON
                                u.id=um.fromid)
                            WHERE
                                um.toid=".(int)$myself."
                            AND
                                um.totrash=0
                            AND
                                `um`.`delayed`=0".$filter."
                            UNION ALL
                            SELECT
                                DISTINCT u.".($config->realnames ? "name" : "username")." AS displayname,
                                u.id
                            FROM
                                (#__users AS u
                            INNER JOIN
                                #__uddeim AS um
                            ON
                                u.id=um.toid)
                            WHERE
                                um.fromid=".(int)$myself."
                            AND
                                um.totrashoutbox=0".$filter."
                            AND um.systemflag=0";

---

WHAT I’VE TRIED –

I tried using a count query but this dind’t work:

if (count ($row->displayname > 1))
    echo "duplicate rows";

Answer 1

Try changing UNION ALL in your query to just UNION. The former retains all rows from both datasets; the latter eliminates duplicates – it is a shortened form of UNION DISTINCT.