Question

I am trying to replicate what would be a left join in MySql in Mongo. I have a collection named Clients and another collection name Orders.

In the clients collection is have:

client_PK, FirstName, LastName, Company

In the orders collection I have:

order_PK, client_fk, OrderDate, OrderAmount,  

So i know that can use embedded documents but for the sake of this question i am looking to use the reference model.

My question is, using these two collections how would I construct a table or object similar to a left join in mysql? I know this is a document db not a relational db but im using sql language just to give you an idea of what im trying to accomplish. In MySql it would look like this:

SELECT * FROM orders LEFT JOIN clients ON clients.client_PK = orders.client_fk

with this i could now construct a table that looked like:

FirstName | LastName | Company | OrderDate | OrderAmount

then i could repeat the rows using a while loop to display all orders and display the clients name with the order. Again i know mongo isn’t a relational db but i am assuming there is a way simulate a table using two collections.

Thank you.

Answer 1

You almost certainly want to be storing these data all in the same MongoCollection (even in just a denormalization collection).

If you absolutely can’t do that, though, and if your set is small, you can do something similar to this (since you asked about PHP):

<?php
// gather orders
$orders = iterator_to_array($mongodb->orders->find());
$joinedOrders = array();
// gather clients
foreach ($db->clients->find() as $client) {
    // iterate orders (like a left join)
    foreach ($orders as $order) {
        // make a "joinedOrders" record for each join match
        if ($order['client_fk'] == $client['client_PK']) {
            $joinedOrders[] = array_merge($order, $client);
        }
    }
}
// result is now in $joinedOrders

This is, however, almost always a bad idea. (-: You really should be denormalizing your data, or using a relational database to store/query relational data.