Question

I have two tables:

doc_types:

doc_id | doc_name | expiry_days
-------|----------|-------------
  1    | Doc_1    |    365
  2    | Doc_2    |    90
  3    | Doc_3    |    30

docs_supplied:

evidence_id | doc_id  | client_id |  date_supplied
------------|---------|-----------|-----------------
       1    | 1       |    5432   |   13-05-2012
       2    | 1       |    3165   |   25-04-2011
       3    | 2       |    5432   |   23-10-2011

The output I would like to see is:

   doc_id | doc_name | expiry_days | client_id |  date_supplied
   -------|----------|-------------|-----------|----------------
     1    | Doc_1    |    365      |    5432   |  13-05-2012
     2    | Doc_2    |    90       |    5432   |  NULL    
     3    | Doc_3    |    30       |    5432   |  NULL
     1    | Doc_1    |    365      |    3165   |  25-04-2012
     2    | Doc_2    |    90       |    3165   |  NULL    
     3    | Doc_3    |    30       |    3165   |  NULL

Essentially for each client, I want to show the complete list of documents (14) but with the extra fields for date supplied etc populated, should there be an entry for that client and document in the DB. So for each client record I’d expect there to be 14 rows, one for each doc_type. Some of these rows will have client-specific data in them.

This is so that in my grid I can just filter against a client_id and see the list of all documents and their status for that person.

Is this possible?

Answer 1

SELECT   doc_types.*, client_id, date_supplied
FROM     doc_types
   CROSS JOIN (SELECT DISTINCT client_id FROM docs_supplied) AS clients
   LEFT  JOIN docs_supplied USING (client_id, doc_id)
ORDER BY client_id DESC, doc_id

See it on sqlfiddle.

If you have a table with a unique client_id column, you can simply use that instead of generating it from the SELECT DISTINCT subquery:

SELECT   doc_types.*, client_id, date_supplied
FROM     doc_types
   CROSS JOIN client_accounts
   LEFT  JOIN docs_supplied USING (client_id, doc_id)
ORDER BY client_id DESC, doc_id

Learn about SQL joins.