View source for Using the union methods in database queries

==A simple example== <!--T:7-->
</translate>
<translate>
<!--T:8-->
Suppose you have want to send out a mailshot to a group of people but the database is such that the names and email addresses you want to send to are not all in the same table.  To invent an arbitrary example, suppose that you want to send the mail to all customers and all suppliers and that that the names and email addresses are stored in tables called, unsurprisingly, ''customers'' and ''suppliers'', respectively.</translate>

<translate>
<!--T:9-->
This query will retrieve all the customer information that we need for the mailshot:
</translate>

<syntaxhighlight lang="php">
$query
    ->select('name, email')
    ->from('customers')
    ;
$mailshot = $db->setQuery($query)->loadObjectList();
</syntaxhighlight>

<translate>
<!--T:10-->
while this query will do the same for all suppliers:
</translate>

<syntaxhighlight lang="php">
$query
    ->select('name, email')
    ->from('suppliers')
    ;
$mailshot = $db->setQuery($query)->loadObjectList();
</syntaxhighlight>

<translate>
<!--T:11-->
Then you can combine the results into a single query like this:
</translate>

<syntaxhighlight lang="php">
$query
    ->select('name, email')
    ->from('customers')
    ->union($q2->select('name , email')->from('suppliers'))
    ;
$mailshot = $db->setQuery($query)->loadObjectList();
</syntaxhighlight>

<translate>
<!--T:12-->
The result set obtained from the union query will actually be a little different from running the individual queries separately because the union query will automatically eliminate duplicates. If you don't mind that the result set may contain duplicates (which mathematically speaking means it isn't a set), then using <tt>unionAll</tt> instead of <tt>union</tt> will improve performance.
</translate>

<translate>