Surprises From NULL in MySQL

`NULL` is one of the things in relational databases that will never disappoint you with the number of surprises it can bring.

Aleksei Kankov

Updated Sep. 07, 22 · Tutorial

Likes (3)

Comment

Save

5.6K Views

The `NULL` value can be surprising until you get used to it. And after you get used to it. And it will forever be surprising. `NULL` is one of the things in relational databases that will never disappoint you with the number of surprises it can bring. Most of them are not pleasant, but it's easy to understand them only thinking about NULL as an absence of value.

History of NULL

NULL was introduced to relational databases as a way to describe missing or inapplicable information. It was a way to say, "I don't know" or "I don't care" or "I don't have this information". And It is a very good way to do so. Let's take a look at a simple example.

     Plain Text 
   
 
 
   | id  | name | amount_of_money |
| --- | ---- | --------------- |
| 1   | John | 1000.00         |
| 2   | Jane | NULL            |
| 3   | Jack | 2000.00         |
| 4   | Jill | -1.00           | 
  

In this table, we have a list of people and their amount of money. We can see that Jane has `NULL` in the amount of money column. It does not mean that Jane has no money. We simply don't know how much money she has. Sometimes developers use -1 to indicate the absence of value, but it will not work for our case. A negative amount of money is a valid value, and it means a person has debt.

NULL on Data Storage Level

NULLs are often criticized for taking up space in the database. It is true that NULLs take up space, but it is not as bad as it seems. And, of course, they should take some space because information about the absence of information is information itself. Most relational databases add a single bit for each field that can be null. But, actually, it's not that simple. Different storage engines can solve this problem in different ways. For example, NDB storage engine reserves 4 bytes per row if the table definition contains any columns allowing NULL, up to 32 NULL columns. So in case, the table has from 1 to 32 nullable columns, it will take 4 bytes per row no matter what. Considering all of that, it's even possible to save some storage space by using NULLs.

NULL in Indexes

It is allowed to have multiple NULLs in the unique index. Let's take a look at the following example.

     MySQL 
   
 
 
   CREATE TABLE users (
    id              INT(11)      NOT NULL AUTO_INCREMENT,
    name            VARCHAR(255) NOT NULL,
    email           VARCHAR(255) DEFAULT NULL,
    amount_of_money INT(11)      DEFAULT NULL,
    PRIMARY KEY (id),
    UNIQUE KEY email (email)
)
    ENGINE = InnoDB
    DEFAULT CHARSET = utf8mb4
    COLLATE = utf8mb4_unicode_ci;


INSERT
INTO
    users (name, email, amount_of_money)
VALUES
    ('John', 'John@example.com', 10);


INSERT
INTO
    users (name)
VALUES
    ('Jane');


INSERT
INTO
    users (name, amount_of_money)
VALUES
    ('Ben', -5);

SELECT *
FROM
    users; 
  

     Plain Text 
   
 
 
   +----+------+------------------+-----------------+
| id | name | email            | amount_of_money |
+----+------+------------------+-----------------+
|  1 | John | John@example.com |              10 |
|  2 | Jane | NULL             |            NULL |
|  3 | Ben  | NULL             |              -5 |
+----+------+------------------+-----------------+
3 rows in set (0.00 sec) 
  

As you can see, we have two users with the same email. It is allowed because NULL is not equal to NULL. It is a special case. NULL is not equal to anything, including NULL. So, we can have multiple NULLs in the unique index. But, of course, we can't have multiple NULLs in the primary key. It is not allowed because the primary key is used to identify a row. And if we have two rows with the same primary key, we won't be able to identify them. So, we can't have multiple NULLs in the primary key. But we can have multiple NULLs in the unique index. It is allowed because the unique index is not used to identify a row. It is used to find a row by a specific value. And if we have two rows with the same value in the unique index, we can still find them by the primary key.

Compare With NULL

Let's make some comparisons with NULL.

     MySQL 
   
   SELECT NULL = NULL, NULL != NULL, NULL > NULL, NULL < NULL;

     Plain Text 
   
 
 
   +-------------+--------------+-------------+-------------+
| NULL = NULL | NULL != NULL | NULL > NULL | NULL < NULL |
+-------------+--------------+-------------+-------------+
|        NULL |         NULL |        NULL |        NULL |
+-------------+--------------+-------------+-------------+ 
  

     MySQL 
   
   SELECT NULL IS NULL, NULL IS NOT NULL;

     Plain Text 
   
 
 
   +--------------+------------------+
| NULL IS NULL | NULL IS NOT NULL |
+--------------+------------------+
|            1 |                0 |
+--------------+------------------+ 
  

NULL is not equal to anything, even to NULL. It's easy to understand, but it can bring some surprises. For example, let's select users from our users' table where email is not equal to.

     MySQL 
   
 
 
   SELECT *
FROM
    users
WHERE
    email <> 'John@example.com'; 
  

And the result is

     Plain Text 
   
   Empty set (0.00 sec)

Even though we have a couple of users with `NULL` emails, they are not selected. Usually, this is not what we want. But this is how MySQL works.

`select null <> 'John@example.com'` will return `null`.

To make it work as we want, we can use `IS NULL` operator or `<=>` NULL-safe equal to operator.

     MySQL 
   
 
 
   SELECT *
FROM
    users
WHERE
     email <> 'John@example.com'
  OR email IS NULL;

SELECT *
FROM
    users
WHERE
    NOT email <=> 'John@example.com'; 
  

     Plain Text 
   
 
 
   +----+------+-------+
| id | name | email |
+----+------+-------+
|  2 | Jane | NULL  |
|  3 | Ben  | NULL  |
+----+------+-------+ 
  

NULL and Arithmetic Operations

Math with NULLs is not interesting. No matter what you do, the result will be NULL.

Addition and subtraction.

     MySQL 
   
 
 
   SELECT
    NULL + 10,
    10 + NULL,
    NULL - 10,
    10 - NULL; 
  

     Plain Text 
   
 
 
   +-----------+-----------+-----------+-----------+
| NULL + 10 | 10 + NULL | NULL - 10 | 10 - NULL |
+-----------+-----------+-----------+-----------+
|      NULL |      NULL |      NULL |      NULL |
+-----------+-----------+-----------+-----------+ 
  

Division and multiplication.

     MySQL 
   
 
 
   SELECT
    10 * NULL,
    NULL * 10,
    NULL / 10,
    10 / NULL,
    0 / NULL,
    NULL / 0; 
  

     Plain Text 
   
   +-----------+-----------+-----------+-----------+----------+----------+
| 10 * NULL | NULL * 10 | NULL / 10 | 10 / NULL | 0 / NULL | NULL / 0 |
+-----------+-----------+-----------+-----------+----------+----------+
|      NULL |      NULL |      NULL |      NULL |     NULL |     NULL |
+-----------+-----------+-----------+-----------+----------+----------+

Module and power.

     MySQL 
   
 
 
   SELECT
    NULL % 10,
    10 % NULL,
    POWER(10, NULL),
    POWER(NULL, 10); 
  

     Plain Text 
   
 
 
   +-----------+-----------+-----------------+-----------------+
| NULL % 10 | 10 % NULL | POWER(10, NULL) | POWER(NULL, 10) |
+-----------+-----------+-----------------+-----------------+
|      NULL |      NULL |            NULL |            NULL |
+-----------+-----------+-----------------+-----------------+ 
  

Group By

Let's take a look at the following example.

     MySQL 
   
   SELECT email FROM users GROUP BY email;
SELECT DISTINCT email FROM users;

Both queries return the same result.

     Plain Text 
   
 
 
   +------------------+
| email            |
+------------------+
| NULL             |
| John@example.com |
+------------------+ 
  

And even though NULL is equal no nothing, here it is equal to itself. I'd say this is an exception from the rules and allows us to group by NULL.

NULL and Boolean Operators

With boolean operators, NULL behaves like an absence of value. So in every case with uncertainty, the result will be null.

     MySQL 
   
 
 
   SELECT
    NULL OR TRUE,
    TRUE OR NULL,
    NULL OR FALSE,
    FALSE OR NULL; 
  

     Plain Text 
   
 
 
   +--------------+--------------+---------------+---------------+
| NULL OR TRUE | TRUE OR NULL | NULL OR FALSE | FALSE OR NULL |
+--------------+--------------+---------------+---------------+
|            1 |            1 |          NULL |          NULL |
+--------------+--------------+---------------+---------------+ 
  

AND

     MySQL 
   
 
 
   SELECT
    NULL AND TRUE,
    TRUE AND NULL,
    NULL AND FALSE,
    FALSE AND NULL; 
  

     Plain Text 
   
 
 
   +---------------+---------------+----------------+----------------+
| NULL AND TRUE | TRUE AND NULL | NULL AND FALSE | FALSE AND NULL |
+---------------+---------------+----------------+----------------+
|          NULL |          NULL |              0 |              0 |
+---------------+---------------+----------------+----------------+ 
  

NOT

     MySQL 
   
   SELECT
    NOT NULL;

     Plain Text 
   
 
 
   +----------+
| NOT NULL |
+----------+
|     NULL |
+----------+ 
  

IS UNKNOWN

     MySQL 
   
   SELECT NULL IS UNKNOWN, NULL IS NOT UNKNOWN;

     Plain Text 
   
 
 
   +-----------------+---------------------+
| NULL IS UNKNOWN | NULL IS NOT UNKNOWN |
+-----------------+---------------------+
|               1 |                   0 |
+-----------------+---------------------+ 
  

IS TRUE, IS FALSE

     MySQL 
   
 
 
   SELECT
    NULL IS TRUE,
    NULL IS NOT TRUE,
    NULL IS FALSE,
    NULL IS NOT FALSE; 
  

     MySQL 
   
 
 
   +--------------+------------------+---------------+-------------------+
| NULL IS TRUE | NULL IS NOT TRUE | NULL IS FALSE | NULL IS NOT FALSE |
+--------------+------------------+---------------+-------------------+
|            0 |                1 |             0 |                 1 |
+--------------+------------------+---------------+-------------------+ 
  

NULL and String Functions

Let's take a look at some string functions.

Concatenation and group concatenation will behave differently with null values.

     MySQL 
   
   SELECT CONCAT('hello', NULL, 'world');

     Plain Text 
   
 
 
   +--------------------------------+
| concat('hello', null, 'world') |
+--------------------------------+
| NULL                           |
+--------------------------------+ 
  

     MySQL 
   
   SELECT
    GROUP_CONCAT(email)
FROM
    users;

     Plain Text 
   
 
 
   +---------------------+
| GROUP_CONCAT(email) |
+---------------------+
| John@example.com    |
+---------------------+ 
  

`CONCAT_WS` will simply ignore NULL values.

     MySQL 
   
   SELECT CONCAT_WS(',', 'Hello', NULL, ' world');

     Plain Text 
   
 
 
   +-----------------------------------------+
| CONCAT_WS(',', 'Hello', NULL, ' world') |
+-----------------------------------------+
| Hello, world                            |
+-----------------------------------------+ 
  

     MySQL 
   
   SELECT CONCAT_WS(NULL, 'Hello', NULL, ' world');

     Plain Text 
   
 
 
   +------------------------------------------+
| CONCAT_WS(NULL, 'Hello', NULL, ' world') |
+------------------------------------------+
| NULL                                     |
+------------------------------------------+ 
  

`ELT` will return the position of the null element.

     MySQL 
   
   SELECT ELT(3, null, 'Bb', null, 'Dd');

     Plain Text 
   
 
 
   +--------------------------------+
| ELT(3, null, 'Bb', null, 'Dd') |
+--------------------------------+
| NULL                           |
+--------------------------------+ 
  

`FIELD` will not find the null element.

     MySQL 
   
   SELECT FIELD(null, 'Aa', 'Bb', 'Cc', 'Dd', 'Ff', null);

     Plain Text 
   
 
 
   +-------------------------------------------------+
| FIELD(null, 'Aa', 'Bb', 'Cc', 'Dd', 'Ff', null) |
+-------------------------------------------------+
|                                               0 |
+-------------------------------------------------+ 
  

`INSERT` will not work with null.

     MySQL 
   
   SELECT INSERT('Quadratic', 3, 4, null);

     Plain Text 
   
 
 
   +---------------------------------+
| INSERT('Quadratic', 3, 4, null) |
+---------------------------------+
| NULL                            |
+---------------------------------+ 
  

`REPEAT` will not work with null.

     MySQL 
   
   SELECT REPEAT('MySQL', NULL), REPEAT(NULL, 3);

     Plain Text 
   
 
 
   +-----------------------+----------------------------------+
| REPEAT('MySQL', NULL) | REPEAT(NULL, 3)                  |
+-----------------------+----------------------------------+
| NULL                  | NULL                             |
+-----------------------+----------------------------------+ 
  

NULL and Aggregate Functions

`COUNT` will ignore null values.

     MySQL 
   
 
 
   SELECT
    COUNT(email),
    COUNT(*)
FROM
    users; 
  

     Plain Text 
   
 
 
   +--------------+----------+
| count(email) | count(*) |
+--------------+----------+
|            1 |        3 |
+--------------+----------+ 
  

`COUNT(DISTINCT)` will behave in the same way even though `SELECT DISTINCT` will return null values.

     MySQL 
   
   SELECT
    COUNT(DISTINCT email)
FROM
    users;

     Plain Text 
   
 
 
   +-----------------------+
| COUNT(DISTINCT email) |
+-----------------------+
|                     1 |
+-----------------------+ 
  

Same story with `AVG`. It will ignore null values.

     MySQL 
   
   SELECT
    AVG(amount_of_money)
FROM
    users;

     Plain Text 
   
 
 
   +----------------------+
| AVG(amount_of_money) |
+----------------------+
|               2.5000 |
+----------------------+ 
  

But depending on your needs, you can use `COALESCE` to replace null values with something else.

     MySQL 
   
   SELECT
    AVG(COALESCE(amount_of_money, 0))
FROM
    users;

     Plain Text 
   
 
 
   +-----------------------------------+
| AVG(COALESCE(amount_of_money, 0)) |
+-----------------------------------+
|                            1.6667 |
+-----------------------------------+ 
  

Conclusion

NULLs in the relational topic in relational databases. If you think you know a lot about it, be careful. That rabbit hole is deep. I hope this article will help you to understand NULLs better. Thanks for reading.

MySQL Plain text

Opinions expressed by DZone contributors are their own.

Related

Trending