Question

I have this query:

        SELECT DISTINCT 1 as table_id, users.id, users.username, a.titles_name title1, b.titles_name as title2, contacts.accepted
        FROM users
        LEFT JOIN contacts ON users.id = contacts.contact_id
        LEFT JOIN titles as a ON a.id = users.title_1
        LEFT JOIN titles as b ON b.id = users.title_2
        WHERE contacts.request_id = ' + $this->session->userdata('user_id') . '
        UNION DISTINCT
        SELECT DISTINCT 2 as table_id, users.id, users.username, a.titles_name title1, b.titles_name as title2, contacts.accepted
        FROM users
        LEFT JOIN contacts ON users.id = contacts.request_id
        LEFT JOIN titles as a ON a.id = users.title_1
        LEFT JOIN titles as b ON b.id = users.title_2
        WHERE contact_id = ' . $this->session->userdata('user_id')

Would it be possible to do something similar in mongodb?

Answer 1

Lets break down the components of the query:

SELECT DISTINCT 1 as table_id, users.id, users.username, a.titles_name title1, b.titles_name as title2, contacts.accepted
FROM users

Yes, this is just a distinct query asking for a subset of fields. Mongo does that.

LEFT JOIN contacts ON users.id = contacts.contact_id
LEFT JOIN titles as a ON a.id = users.title_1
LEFT JOIN titles as b ON b.id = users.title_2
WHERE contacts.request_id = ' + $this->session->userdata('user_id') . '

This is a join, which you cannot do in mongo. However, you can de-normalize so it becomes unnecessary or do similar joining of data on the client side.

UNION DISTINCT
SELECT DISTINCT 2 as table_id, users.id, users.username, a.titles_name title1, b.titles_name as title2, contacts.accepted
FROM users
LEFT JOIN contacts ON users.id = contacts.request_id
LEFT JOIN titles as a ON a.id = users.title_1
LEFT JOIN titles as b ON b.id = users.title_2
WHERE contact_id = ' . $this->session->userdata('user_id')

This is a union, which you could do on the client side and for which the query contains the same components i’ve described above.

TLDNR: No, you can’t do this exact query in mongodb, but you can achieve the same result.