Domain types for validation in PostgreSQL

Introduction

Domain types help you make sure data follows rules before it goes into your database. This keeps your data clean and correct.

When you want to check that a phone number always has 10 digits.

When you want to ensure an email address has a proper format.

When you want to limit a column to only positive numbers.

When you want to reuse the same validation rules in many tables.

When you want to avoid mistakes by catching bad data early.

Syntax

PostgreSQL

CREATE DOMAIN domain_name AS base_data_type
  [ DEFAULT default_expression ]
  [ CONSTRAINT constraint_name CHECK (expression) ];

A domain is like a custom data type with rules.

You use CREATE DOMAIN once, then use the domain as a column type.

Examples

This domain ensures the phone number is exactly 10 digits and only numbers.

PostgreSQL

CREATE DOMAIN phone_number AS TEXT
  CHECK (LENGTH(VALUE) = 10 AND VALUE ~ '^[0-9]+$');

This domain allows only positive integers.

PostgreSQL

CREATE DOMAIN positive_int AS INTEGER
  CHECK (VALUE > 0);

This domain checks if the text looks like an email address (simple pattern).

PostgreSQL

CREATE DOMAIN email_address AS TEXT
  CHECK (VALUE ~* '^[A-Z0-9._%+-]+@[A-Z0-9.-]+\.[A-Z]{2,}$');

Sample Program

This example creates a domain for age that must be between 0 and 120. Then it creates a table using that domain. The first insert works, the second fails because age is negative.

PostgreSQL

CREATE DOMAIN age_domain AS INTEGER
  CHECK (VALUE >= 0 AND VALUE <= 120);

CREATE TABLE people (
  id SERIAL PRIMARY KEY,
  name TEXT NOT NULL,
  age age_domain
);

INSERT INTO people (name, age) VALUES ('Alice', 30);
INSERT INTO people (name, age) VALUES ('Bob', -5);

Important Notes

Use the keyword VALUE inside CHECK to refer to the data being checked.

Domains help keep your data rules in one place, making your database easier to manage.

When inserting or updating, if data breaks domain rules, PostgreSQL will stop the action with an error.

Summary

Domains are custom data types with built-in validation rules.

They help keep data clean by enforcing rules automatically.

You create a domain once and use it like any other data type in tables.

Practice

1. What is the main purpose of using domain types in PostgreSQL?

easy

A. To speed up query execution by indexing

B. To create custom data types with automatic validation rules

C. To store large binary data efficiently

D. To create temporary tables for session use

Solution

1. Step 1: Understand what domain types are

   Domain types are user-defined data types that include validation rules to ensure data quality.

2. Step 2: Identify the main purpose

   The main purpose is to enforce rules automatically when data is inserted or updated, keeping data clean.

3. Final Answer:

   To create custom data types with automatic validation rules -> Option B

4. Quick Check:

   Domain types = custom types with validation [OK]

Hint: Domains add rules to types, not speed or storage [OK]

Common Mistakes:

• Confusing domains with indexing
• Thinking domains store large files
• Mixing domains with temporary tables

2. Which of the following is the correct syntax to create a domain named positive_int that only allows positive integers?

easy

A. CREATE DOMAIN positive_int AS integer CHECK (VALUE > 0);

B. CREATE DOMAIN positive_int TYPE integer WHERE VALUE > 0;

C. CREATE DOMAIN positive_int AS integer VALIDATE (VALUE > 0);

D. CREATE DOMAIN positive_int AS integer IF VALUE > 0;

Solution

1. Step 1: Recall the correct syntax for domain creation

   The syntax is: CREATE DOMAIN name AS base_type CHECK (condition);

2. Step 2: Match the syntax with options

   CREATE DOMAIN positive_int AS integer CHECK (VALUE > 0); matches the correct syntax with CHECK and VALUE keyword.

3. Final Answer:

   CREATE DOMAIN positive_int AS integer CHECK (VALUE > 0); -> Option A

4. Quick Check:

   CREATE DOMAIN ... AS ... CHECK(...) [OK]

Hint: Use CHECK with VALUE keyword in domain creation [OK]

Common Mistakes:

• Using TYPE instead of AS
• Using WHERE or VALIDATE instead of CHECK
• Missing parentheses around condition

3. Given the domain creation:

CREATE DOMAIN us_zipcode AS varchar(5) CHECK (VALUE ~ '^[0-9]{5}$');

What will happen if you try to insert '1234a' into a table column of type us_zipcode?

medium

A. The insert will succeed because it's a varchar

B. The insert will cause a syntax error

C. The insert will succeed but store NULL instead

D. The insert will fail due to the CHECK constraint violation

Solution

1. Step 1: Understand the domain's CHECK constraint

   The domain requires the value to match exactly 5 digits using a regular expression.

2. Step 2: Check the value '1234a' against the regex

   '1234a' contains a letter, so it does not match the pattern of 5 digits.

3. Final Answer:

   The insert will fail due to the CHECK constraint violation -> Option D

4. Quick Check:

   Regex check fails = insert rejected [OK]

Hint: Regex in CHECK rejects invalid patterns [OK]

Common Mistakes:

• Assuming varchar allows any string without checks
• Thinking invalid data stores as NULL automatically
• Confusing constraint violation with syntax error

4. You have this domain:

CREATE DOMAIN non_empty_text AS text CHECK (LENGTH(VALUE) > 0);

Which of the following INSERT statements will cause an error when inserting into a column of type non_empty_text?

medium

A. INSERT INTO table_name (col) VALUES ('');

B. INSERT INTO table_name (col) VALUES ('world');

C. INSERT INTO table_name (col) VALUES (' ');

D. INSERT INTO table_name (col) VALUES (' hello ');

Solution

1. Step 1: Understand the CHECK condition

   The domain requires text length to be greater than 0, so empty strings fail. Whitespace-only strings pass as they have length > 0.

2. Step 2: Evaluate each insert value

   INSERT INTO table_name (col) VALUES (''); inserts an empty string '', length 0, violating the check.

3. Final Answer:

   INSERT with empty string '' causes error -> Option A

4. Quick Check:

   Empty string fails length > 0 check [OK]

Hint: Empty strings fail length > 0 check [OK]

Common Mistakes:

• Thinking whitespace strings fail (they pass)
• Assuming empty string is allowed
• Confusing syntax errors with constraint errors

5. You want to create a domain rating that stores integers from 1 to 5 inclusive. Which domain definition correctly enforces this range and can be used in multiple tables?

hard

A. CREATE DOMAIN rating AS integer WHERE (VALUE BETWEEN 1 AND 5);

B. CREATE DOMAIN rating AS integer CHECK (VALUE > 1 AND VALUE < 5);

C. CREATE DOMAIN rating AS integer CHECK (VALUE >= 1 AND VALUE <= 5);

D. CREATE DOMAIN rating AS integer CHECK (VALUE IN (1, 2, 3, 4));

Solution

1. Step 1: Understand the required range

   The rating must include 1 and 5, so boundaries are inclusive.

2. Step 2: Analyze each CHECK condition

   The option using WHERE instead of CHECK has invalid syntax.
   CREATE DOMAIN rating AS integer CHECK (VALUE > 1 AND VALUE < 5); excludes 1 and 5.
   CREATE DOMAIN rating AS integer CHECK (VALUE >= 1 AND VALUE <= 5); uses >= 1 AND <= 5, correctly including boundaries.
   CREATE DOMAIN rating AS integer CHECK (VALUE IN (1, 2, 3, 4)); uses IN list which misses 5.

3. Step 3: Choose the best option for clarity and standard usage

   CREATE DOMAIN rating AS integer CHECK (VALUE >= 1 AND VALUE <= 5); is clear, standard, and commonly used for range checks.

4. Final Answer:

   CREATE DOMAIN rating AS integer CHECK (VALUE >= 1 AND VALUE <= 5); -> Option C

5. Quick Check:

   Inclusive range uses >= and <= [OK]

Hint: Use >= and <= for inclusive numeric domain checks [OK]

Common Mistakes:

• Using > and < excludes boundary values
• Using WHERE instead of CHECK
• Incomplete IN list misses values