JSON 

show you how to use JSON data type to store JSON documents. (1)

Summary: in this tutorial, you will learn how to use MySQL JSON data type to store JSON documents in the database.

What is JSON

JSON stands for “JavaScript Object Notation”. JSON is a lightweight data-interchange format that is easy for humans to read and write and simple for computers to parse and generate.

JSON is built on two data structures: objects and arrays.

Objects

An object is an unordered collection of key-value pairs enclosed in curly braces {}. Each key is a string, followed by a colon (:), and then the associated value. For example:

{
    "name": "John Doe",
    "age": 22
}Code language: JSON / JSON with Comments (json)

Note that you must enclose the key in a JSON document with double quotes (").

Arrays

An array is an ordered list of values closed in square brackets []. An array may contain values of any valid JSON data type, including objects and other arrays. For example:

["John Doe", 22]Code language: JSON / JSON with Comments (json)

In this example, we have a JSON array of two values that represent the name and age of a person.

JSON supports several data types, including:

• String: “JSON”
• Number: 10, 12.25
• Boolean: true and false.
• Null: null

In practice, you use JSON in web development for:

• Configuration files
• Data exchange between a client and a server

Introduction to MySQL JSON data type

MySQL supports the native JSON data type defined by RFC 7159  starting in version 5.7.8. The native JSON data type offers the following advantages:

• MySQL validates the JSON documents stored in the JSON column and issues an error if they are invalid.
• MySQL stores the JSON documents in a binary format optimized for quick searches.

MySQL uses roughly the same storage for JSON documents as for LONGBLOG or LONGTEXT data.

The following shows how to define a table column with the JSON data type:

column_name JSONCode language: SQL (Structured Query Language) (sql)

The JSON column can store NULL values. Also, starting from MySQL 8.0.13, a JSON column can accept a default value.

It is not possible to directly index a JSON column. Instead, you can use a functional index to create an index on values extracted from the JSON column.

When querying data from a JSON column, the MySQL optimizer will search for functional indexes to match the JSON expressions.

MySQL JSON data type example

We’ll take an example of using the JSON data type.

1) Creating a table with a JSON column

The following statement creates a new table called products with a JSON column:

CREATE TABLE products(
   id INT AUTO_INCREMENT PRIMARY KEY,
   name VARCHAR(255) NOT NULL,
   price DECIMAL(10,2) NOT NULL,
   properties JSON
);Code language: SQL (Structured Query Language) (sql)

The products table has four columns:

• id: This is an auto-increment primary key column that uniquely identifies a product.
• name: This column stores the name of the product.
• price: This column stores the price of the product.
• properties: This column stores the properties of the product such as colors and sizes.

2) Inserting JSON data into the table

The following statement inserts a new row with JSON data into the products table:

INSERT INTO products(name, price, properties)
VALUES('T-Shirt', 25.99, '{"sizes":["S","M","L","XL"], "colors": ["white","black"]}');Code language: SQL (Structured Query Language) (sql)

In this example, we inserted a new row into the products table. The value of the propertiescolumn is a JSON document with the following format:

{"sizes":["S","M","L","XL"],"colors":["white","black"]}Code language: JSON / JSON with Comments (json)

And we place the JSON document inside a single quote (‘).

When the statement executes, MySQL performs the following steps:

• First, validate the JSON document for validity.
• Second, convert the JSON string into binary format and store it in the JSON column.

3) Querying JSON data from the table

The following query retrieves data from the products table:

SELECT name, properties
FROM products;Code language: SQL (Structured Query Language) (sql)

Output:

+---------+----------------------------------------------------------------+
| name    | properties                                                     |
+---------+----------------------------------------------------------------+
| T-Shirt | {"sizes": ["S", "M", "L", "XL"], "colors": ["white", "black"]} |
+---------+----------------------------------------------------------------+
1 row in set (0.00 sec)Code language: plaintext (plaintext)

If the output is difficult to read, you can use the JSON_PRETTY() function to format it.

SELECT JSON_PRETTY(properties)
FROM products;Code language: SQL (Structured Query Language) (sql)

Output:

+------------------------------------------------------------------------------------------------------+
| JSON_PRETTY(properties)                                                                              |
+------------------------------------------------------------------------------------------------------+
| {
  "sizes": [
    "S",
    "M",
    "L",
    "XL"
  ],
  "colors": [
    "white",
    "black"
  ]
} |
+------------------------------------------------------------------------------------------------------+
1 row in set (0.00 sec)Code language: plaintext (plaintext)

4) Getting the keys of a JSON document

The following uses the JSON_KEYS() function to list all the keys of the JSON document:

SELECT JSON_KEYS(properties)
FROM products;Code language: SQL (Structured Query Language) (sql)

Output:

+-----------------------+
| JSON_KEYS(properties) |
+-----------------------+
| ["sizes", "colors"]   |
+-----------------------+
1 row in set (0.01 sec)
Code language: plaintext (plaintext)

In this example, we use the JSON_KEYS() function to get all the keys of the JSON document stored in the properties column.

The function returns a JSON array containing all the keys of the JSON document.

5) Extracting data from a JSON document

To specify a location within a JSON document, you use a JSON path expression. A path expression allows you to navigate through the structure of a JSON document to access specific data.

• Use the dollar sign ($) to represent the current document.
• Use a key or array index to specify the exact location.

For example, to get the first colors of the product, you use the following path:

$.colors[0]Code language: SQL (Structured Query Language) (sql)

To extract the first color from the JSON document with the above path, you use the JSON_EXTRACT() function. For example:

SELECT JSON_EXTRACT(properties, "$.colors[0]")
FROM products;Code language: SQL (Structured Query Language) (sql)

Output:

+-----------------------------------------+
| JSON_EXTRACT(properties, "$.colors[0]") |
+-----------------------------------------+
| "white"                                 |
+-----------------------------------------+
1 row in set (0.00 sec)Code language: plaintext (plaintext)

Summary

• MySQL supports native JSON data types starting in version 5.7.8.
• MySQL stores JSON in a binary format optimized for quick searches.
• Use the JSON_PRETTY() function to format the JSON documents.
• Use the JSON_KEY() function to get all keys of a JSON document.
• Use the JSON_EXTRACT() function to extract data from a JSON document.