Summing a field while excluding certain fields in SQL

Question

I am trying to suma column in a table, while excluding certain records that have the paid field set to true.

I am doing so like this:

SELECT SUM( cost ) AS total
FROM sales
WHERE passport = 'xxxxx'
AND paid <>1

The table is full of data, and I can display costs by themselves, or the entire total. Just adding on AND paid <>1 Is what causes it to fail. The query does not fail as such, but NULL is returned which is quite useless.

This is the SQL for my table

CREATE TABLE sales (
  id int(10) unsigned NOT NULL AUTO_INCREMENT,
  uuid varchar(64) NOT NULL,
  firstname varchar(64) NOT NULL DEFAULT '',
  lastname varchar(64) NOT NULL DEFAULT '',
  passport varchar(64) DEFAULT NULL,
  product varchar(64) NOT NULL,
  quantity int(11) DEFAULT NULL,
  cost double DEFAULT NULL,
  paymenttype varchar(64) NOT NULL DEFAULT '',
  paid tinyint(1) DEFAULT NULL,
  tabno varchar(64) NOT NULL,
  createdby int(10) unsigned DEFAULT NULL,
  creationdate datetime DEFAULT NULL,
  modifiedby int(10) unsigned DEFAULT NULL,
  modifieddate timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
  PRIMARY KEY (id)
)

And the current data

INSERT INTO sales (id, uuid, firstname, lastname, passport, product, quantity, cost, paymenttype, paid, tabno, createdby, creationdate, modifiedby, modifieddate) VALUES
(20, ':8dcee958-d1ac-6791-6253-0a7344054295', 'Jason', 'Hoff', 'r454545', 'Nicaraguan nachoes', 4, 320, 'credit', 1, '23434', 2, '2010-07-06 04:10:18', 2, '2010-07-06 04:10:18'),
(19, ':3f03cda5-21bf-9d8c-5eaa-664eb2d4f5a6', 'Jason', 'Hoff', 'r454545', 'Nica Libre (doble 4 o 5 anos)', 1, 30, 'cash', NULL, '35', 2, '2010-07-06 03:35:35', 2, '2010-07-06 03:35:35'),
(18, ':f83da33b-2238-94b9-897c-debed0c3815e', 'Jason', 'Hoff', 'r454545', 'Helado con salsa de chocolate', 1, 40, 'cash', 1, '2', 2, '2010-07-05 21:30:58', 2, '2010-07-05 21:30:58');

Answer 1

This may mean there are no rows in the table for which paid <> 1. Or maybe there are but they cost of NULL.

Answer 2

Hi, I would not use "<>", but use "!=" instead;

SELECT SUM( cost ) AS total
FROM sales
WHERE passport = 'xxxxx'
AND paid != 1

If that doesn't work, can you post the table definition? And are there any records which do not have paid = 1?

Answer 3

SELECT SUM( cost ) AS total
FROM sales
WHERE passport = 'xxxxx'
AND paid = false or paid is null

EDITED

use

IS NOT 

instead of

!=

This should work

SELECT SUM( cost ) AS total
FROM sales
WHERE passport = 'xxxxx'
AND paid IS NOT true

Answer 4

The 'paid' value is NULL for that row. You would need to do

SELECT SUM( cost ) AS total
FROM test.sales
WHERE passport = 'r454545'
AND paid IS NULL or paid = 0 
     /*Or paid <> 1 as I see you are using tinyint datatype*/

Or, better would be to not allow NULLS in that column and have paid default to 0.

Answer 5

Your problem is that your condition does not match any rows.

The condition paid <> 1 does not match the row where paid is NULL.

Try this query: SELECT 1 <> NULL It will return NULL. A WHERE clause filters out rows in which the clause is either false or NULL.

Replace AND paid <> 1 with AND (paid IS NULL OR paid <> 1)

Answer 6

The book SQL Antipatterns describes this problem in detail, section Searching Nullable Columns. Strongly recommended book.