PostgreSQL hstore

Summary: in this tutorial, we will show you how to work with PostgreSQL hstore data type.

The hstore module implements hstore data type for storing key-value pairs in a single value. The hstore data type is very useful in many cases, such as semi-structured data, or rows with many attributes that are rarely queried. Notice that keys and values are just text strings only.

Enable PostgreSQL hstore extension

Before working with the hstore data type, you need to enable the hstore extension, which loads the contrib module to your PostgreSQL instance.

Create a table with hstore data type

We create a table named books that has three columns:

•  id is the primary key that identifies the book.
•  title is 0the title of the products
•  attr stores attributes of the book such as isbn, weight, paper-back, etc. The data type of the attr column is hstore.

We use the CREATE TABLE statement to create the books table as follows:

Insert data into PostgreSQL hstore column

We use the INSERT statement to insert data into the hstore column as follows:

The data that we insert into the hstore column is a list of comma separated key =>value pairs. Both keys and values are quoted using double quotes (“”).

Let’s insert one more row.

Query data from a hstore column

Querying data from a hstore column is similar to querying data from column with native data type using SELECT statement as follows:

Query value for a specific key

Postgresql hstore provides the -> operator to query value of a specific key from a hstore column. For example, if we want to know ISBN-13 of all available books in the books table, we can use the -> operator as follows:

Use value in the WHERE clause

You can use the -> operator in the WHERE clause to filter the rows whose values of the hstore column match the input value. For example, the following  query retrieves the title and weight of a book that has ISBN-13 value matches 978-1449370000:

Add key-value pairs to existing rows

With hstore column, we can easily add a new key-value pair to existing rows e.g., we can add free shipping key to the attr column of the books table as follows:

Now, we can check to see if the “freeshipping” => “yes”  pair has been added successfully.

Update existing key-value pair

You can update existing key-value pair using the UPDATE statement. The following statement updates the value of the "freeshipping" key to "no".

Remove existing key-value pair

PostgreSQL allows you to remove existing key-value pair from an hstore column. For example, the following statement removes the "freeshipping"=>"no" key-value pair in the attr column.

Check for a specific key in hstore column

You can check for a specific key in an hstore column using the ? operator in the WHERE clause. For example, the following statement returns all rows with attr contains publisher key.

Check for a key-value pair

You can query based on hstore key-value pair using the @> operator. The following statement retrieves all rows which attr column contains a key-value pair that matches "weight"=>"11.2 ounces".

Query rows that contain multiple specified keys

You can query the rows whose hstore column contain multiple keys using ?& operator. For example, you can get books where attr column contains both language and weight keys.

To check if a row whose hstore column contains any key you use the ?| operator instead of the ?& operator.

Get all keys from a hstore column

To get all keys from an hstore column, you use the akeys() function as follows:

Or you can use the  skey() function if you want PostgreSQL to return the result as a set.

Get all values from an hstore column

Like keys, you can get all values from an hstore column using the  avals() function in the form of arrays.

Or you can use the  svals() function if you want to get the result as a set.

Convert hstore data to JSON

PostgreSQL provides the hstore_to_json() function to convert hstore data to JSON. See the following statement:

Convert hstore data to sets

To convert hstore data to sets, you use the  each() function as follows:

In this tutorial, we have shown you how to work with PostgreSQL hstore data type and introduced you to the most useful operations that you can perform against hstore data type.