Creating JSON columns in PostgreSQL - Why You Should Know This

The Big Idea

What if your database could hold any kind of data without breaking a sweat?

The Scenario

Imagine you have a spreadsheet where each row needs to store different types of information, like names, addresses, phone numbers, and even some extra notes that don't fit neatly into columns.

You try to keep all this extra info in separate columns, but soon the sheet becomes huge and messy.

The Problem

Manually adding many columns for every little detail is slow and confusing.

It's easy to make mistakes, like forgetting to add a new column or mixing up data types.

Also, if the extra info changes often, you have to keep changing the table structure, which is painful and risky.

The Solution

Creating JSON columns lets you store flexible, structured data inside a single column.

This means you can keep all the extra details together without changing the table layout every time.

It's like having a mini-database inside one column that can hold different shapes of data easily.

Before vs After

✗ Before

ALTER TABLE contacts ADD COLUMN phone VARCHAR(20);
ALTER TABLE contacts ADD COLUMN notes TEXT;

✓ After

ALTER TABLE contacts ADD COLUMN details JSON;

What It Enables

You can now store and query complex, changing data easily without redesigning your database every time.

Real Life Example

A company storing customer profiles where each customer might have different preferences, multiple phone numbers, or special notes that don't fit a fixed format.

Using JSON columns, they keep all this info neatly in one place and can still search inside it.

Key Takeaways

Manual columns for every detail get messy and hard to maintain.

JSON columns store flexible data in one place.

This makes your database easier to update and query.

Practice

1. What is the main purpose of creating a JSON column in a PostgreSQL table?

easy

A. To enforce strict relational data types

B. To create indexes automatically

C. To speed up numeric calculations

D. To store flexible, structured data inside a single column

Solution

1. Step 1: Understand JSON column usage

   JSON columns allow storing data that can have varying structure, unlike fixed columns.

2. Step 2: Compare with other options

   Options A, C, and D do not describe JSON columns' purpose; they relate to other database features.

3. Final Answer:

   To store flexible, structured data inside a single column -> Option D

4. Quick Check:

   JSON columns = flexible data storage [OK]

Hint: JSON columns store flexible data, not fixed types [OK]

Common Mistakes:

• Thinking JSON columns enforce strict types
• Confusing JSON with indexing features
• Assuming JSON speeds numeric calculations

2. Which of the following is the correct syntax to create a table with a JSON column named data?

easy

A. CREATE TABLE mytable (id INT, data JSONB STRING);

B. CREATE TABLE mytable (id INT, data JSON);

C. CREATE TABLE mytable (id INT, data TEXT JSON);

D. CREATE TABLE mytable (id INT, data STRING);

Solution

1. Step 1: Recall JSON column syntax

   In PostgreSQL, JSON columns are declared with type JSON or JSONB directly.

2. Step 2: Check each option

   CREATE TABLE mytable (id INT, data JSON); uses correct syntax. Options A, B, and C use invalid or mixed types.

3. Final Answer:

   CREATE TABLE mytable (id INT, data JSON); -> Option B

4. Quick Check:

   JSON column syntax = CREATE TABLE mytable (id INT, data JSON); [OK]

Hint: Use JSON or JSONB as column type directly [OK]

Common Mistakes:

• Using STRING or TEXT with JSON incorrectly
• Combining JSON with other types in declaration
• Missing commas or parentheses

3. Given the table created as CREATE TABLE products (id SERIAL, info JSONB); and the insert statement:
INSERT INTO products (info) VALUES ('{"name": "Pen", "price": 1.5}');
What will be the result of the query:
SELECT info->>'name' FROM products WHERE id = 1;

medium

A. {"name": "Pen", "price": 1.5}

B. 1.5

C. Pen

D. NULL

Solution

1. Step 1: Understand JSONB insert and query

   The JSONB column stores the JSON object. The operator ->> extracts the text value of the key.

2. Step 2: Analyze the query

   The query selects the 'name' field as text from the JSONB column for id 1, which is 'Pen'.

3. Final Answer:

   Pen -> Option C

4. Quick Check:

   info->>'name' = 'Pen' [OK]

Hint: Use ->> to get JSON text value by key [OK]

Common Mistakes:

• Using -> instead of ->> returns JSON, not text
• Expecting full JSON instead of single value
• Not matching the correct id in WHERE clause

4. You try to create a table with a JSON column using:
CREATE TABLE users (id INT, profile JSONB DEFAULT {});
But get a syntax error. What is the problem?

medium

A. Missing quotes around the default JSON value

B. JSONB columns cannot have default values

C. Empty braces {} are not valid default for JSONB

D. INT type cannot be used with JSONB columns

Solution

1. Step 1: Check default value syntax for JSONB

   Default JSON values must be valid JSON strings, so '{}' must be in single quotes.

2. Step 2: Identify the error cause

   Using {} without quotes causes syntax error. Correct is DEFAULT '{}'.

3. Final Answer:

   Missing quotes around the default JSON value -> Option A

4. Quick Check:

   Default JSON must be quoted [OK]

Hint: Always quote JSON default values as strings [OK]

Common Mistakes:

• Using unquoted braces {} as default
• Assuming JSONB cannot have defaults
• Confusing data types in table definition

5. You want to create a table orders with a JSONB column details that stores order info. You also want to ensure each order has a customer_id and a details JSON with a total field as a number. Which approach correctly creates the table and enforces the total field presence?

hard

A. CREATE TABLE orders (id SERIAL PRIMARY KEY, customer_id INT NOT NULL, details JSONB NOT NULL CHECK (details->>'total' IS NOT NULL));

B. CREATE TABLE orders (id SERIAL PRIMARY KEY, customer_id INT, details JSONB DEFAULT '{}');

C. CREATE TABLE orders (id SERIAL PRIMARY KEY, customer_id INT NOT NULL, details JSONB NOT NULL);

D. CREATE TABLE orders (id SERIAL PRIMARY KEY, customer_id INT NOT NULL, details JSONB NOT NULL CHECK (details ? 'total'));

Solution

1. Step 1: Understand JSONB constraints

   To ensure 'total' key exists and is not null, use a CHECK with details->>'total' IS NOT NULL.

2. Step 2: Evaluate options

   CREATE TABLE orders (id SERIAL PRIMARY KEY, customer_id INT NOT NULL, details JSONB NOT NULL CHECK (details ? 'total')); checks only key presence (details ? 'total'), but does not check value is not null. CREATE TABLE orders (id SERIAL PRIMARY KEY, customer_id INT NOT NULL, details JSONB NOT NULL CHECK (details->>'total' IS NOT NULL)); checks both key and value presence correctly.

3. Final Answer:

   CREATE TABLE orders (id SERIAL PRIMARY KEY, customer_id INT NOT NULL, details JSONB NOT NULL CHECK (details->>'total' IS NOT NULL)); -> Option A

4. Quick Check:

   CHECK ensures 'total' key with value [OK]

Hint: Use CHECK with details->>'key' IS NOT NULL to enforce JSON key presence [OK]

Common Mistakes:

• Using ? operator without checking value
• Not adding NOT NULL constraints
• Setting default empty JSON without validation