How to UPDATE SQLite column while using JOIN - sqlite

Question

For homework, I am currently trying to figure out how to update a column while using JOIN but I can't seem to get it right.
The first part is this:
The 'company' you work for has decided to hire a new bi-lingual support position. Your job is to locate all users who have purchased a Spanish language track so they can be assigned to a new support representative. Sales have begun to identify all the albums that are classified as Spanish language, so far they have found AlbumId's of 8, 21, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33, 34, 41, 42, 45, 47, 52, 53.
I solved this with this query:
SELECT * FROM customers
JOIN invoices USING (CustomerId)
JOIN invoice_items USING (InvoiceId)
JOIN tracks USING (TrackId)
WHERE tracks.AlbumId IN (8, 21, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33, 34, 41, 42, 45, 47, 52, 53)
GROUP BY invoices.InvoiceId;
Now the second part is this:
Help out the sales team by modifying your query from Part 1. Instead of just listing all the customers, it should update the customer's assigned support representative. The new support representative's id is 6.
I tried running this:
UPDATE customers
SET SupportRepId = 6
WHERE(SELECT * FROM customers
JOIN invoices USING (CustomerId)
JOIN invoice_items USING (InvoiceId)
JOIN tracks USING (TrackId)
WHERE tracks.AlbumId IN (8, 21, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33, 34, 41, 42, 45, 47, 52, 53)
GROUP BY invoices.InvoiceId);
But I am getting an error that says:
SQLITE_ERROR: sub-select returns 33 columns - expected 1
errno: 1
code: SQLITE_ERROR
name: Error

Answer 1

I got this working by using the following command:
UPDATE customers
SET SupportRepId = 6
WHERE CustomerId IN (
SELECT customers.CustomerId FROM customers
JOIN invoices USING (CustomerId)
JOIN invoice_items USING (InvoiceId)
JOIN tracks USING (TrackId)
WHERE AlbumId IN (8, 21, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33, 34, 41, 42, 45, 47, 52, 53)
);

Answer 2

The last error is the clue. WHERE checks for 1 value; 0 is false, positive is true. * is returning 33 values (columns).
To cater for multiple updates you would need to use a correlated subquery e.g. WHERE IN or WHERE EXISTS that will provide a result for each row processed by the update.
Using WHERE EXISTS the query could be :-
UPDATE customers
SET SupportRepId = 6
WHERE EXISTS (
SELECT 1 FROM customers AS B
JOIN invoices USING (CustomerId)
JOIN invoice_items USING (InvoiceId)
JOIN tracks USING (TrackId)
WHERE tracks.AlbumId IN (8, 21, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33, 34, 41, 42, 45, 47, 52, 53)
AND B.CustomerId = customers.CustomerId
);
Thus for each row processed by the UPDATE 1 will be returned if the customer has any invoice that has a track from one of the listed albums AND if the customerId from the correlatedsubquery (B.CustomerId) is the same as the CustomerId of the row being updated.
Using WHERE IN the query could be :-
UPDATE customers
SET SupportRepId = 6
WHERE CustomerId IN (
SELECT CustomerId FROM customers
JOIN invoices USING (CustomerId)
JOIN invoice_items USING (InvoiceId)
JOIN tracks USING (TrackId)
WHERE tracks.AlbumId IN (8, 21, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33, 34, 41, 42, 45, 47, 52, 53)
);