Regular expression matching (~ operator) in PostgreSQL

Introduction

Regular expression matching helps find text patterns inside data, like searching for words or formats in a list.

You want to find all emails in a list that have a specific domain.

You need to check if phone numbers follow a certain pattern.

You want to filter names that start with a certain letter or letters.

You want to find rows where a text column contains digits.

You want to validate if a string matches a pattern before processing it.

Syntax

PostgreSQL

expression ~ pattern

The ~ operator checks if the expression matches the regular expression pattern.

It returns true if there is a match, false otherwise.

Examples

Checks if 'hello123' contains any digit.

PostgreSQL

SELECT 'hello123' ~ '[0-9]';

Checks if 'abc' starts with the letter 'a'.

PostgreSQL

SELECT 'abc' ~ '^a';

Checks if the string ends with '@example.com'.

PostgreSQL

SELECT '[e-mail]' ~ '@example\.com$';

Sample Program

This creates a temporary table with some names. Then it selects only the names that contain any digit.

PostgreSQL

CREATE TEMP TABLE users(name TEXT);
INSERT INTO users VALUES
('Alice'),
('Bob123'),
('Charlie_99'),
('David');

SELECT name FROM users WHERE name ~ '[0-9]';

Important Notes

Regular expressions are case sensitive by default. Use ~* for case-insensitive matching.

Escape special characters in patterns with double backslash (\\) if needed.

Use ^ and $ to match the start and end of the string respectively.

Summary

The ~ operator checks if a string matches a regular expression pattern.

It returns true or false, helping filter data based on text patterns.

Use it to find or validate text formats inside your database.

Practice

1. What does the PostgreSQL operator ~ do when used in a query like column_name ~ 'pattern'?

easy

A. Returns the length of the column value

B. Replaces the column value with the pattern

C. Checks if the column value matches the regular expression pattern

D. Counts the number of times the pattern appears in the column

Solution

1. Step 1: Understand the ~ operator

   The ~ operator in PostgreSQL is used to check if a string matches a regular expression pattern.

2. Step 2: Apply to the query context

   When used as column_name ~ 'pattern', it returns true if the column's value matches the pattern, otherwise false.

3. Final Answer:

   Checks if the column value matches the regular expression pattern -> Option C

4. Quick Check:

   ~ means match pattern [OK]

Hint: Remember: ~ means 'matches pattern' in PostgreSQL [OK]

Common Mistakes:

• Thinking ~ replaces or modifies data
• Confusing ~ with counting occurrences
• Assuming ~ returns string length

2. Which of the following is the correct syntax to find rows where the column email contains the domain example.com using the ~ operator?

easy

A. SELECT * FROM users WHERE email ~ 'example.com';

B. SELECT * FROM users WHERE email ~ '@example\.com$';

C. SELECT * FROM users WHERE email LIKE '%example.com%';

D. SELECT * FROM users WHERE email = '~example.com';

Solution

1. Step 1: Understand the pattern for matching domain

   To match emails ending with '@example.com', the regex should be @example\.com$ where \. escapes the literal dot and $ means end of string.

2. Step 2: Check the syntax of the ~ operator usage

   The correct syntax is email ~ '@example\.com$' to find emails ending with that domain.

3. Final Answer:

   SELECT * FROM users WHERE email ~ '@example\.com$'; -> Option B

4. Quick Check:

   Regex match with ~ and end anchor [OK]

Hint: Use $ to anchor regex at string end with ~ operator [OK]

Common Mistakes:

• Using LIKE instead of ~ for regex matching
• Missing anchors like $ for exact domain match
• Using = operator with regex pattern

3. Given the table products with a column code containing values: 'A123', 'B234', 'A345', 'C456', what will the query SELECT code FROM products WHERE code ~ '^A'; return?

medium

A. A123, A345

B. B234, C456

C. All codes

D. No rows

Solution

1. Step 1: Understand the regex pattern ^A

   The caret ^ means the string must start with 'A'.

2. Step 2: Apply pattern to each code

   Codes starting with 'A' are 'A123' and 'A345'.

3. Final Answer:

   A123, A345 -> Option A

4. Quick Check:

   ^ means start with 'A' [OK]

Hint: ^ anchors regex to start of string [OK]

Common Mistakes:

• Confusing ^ with end anchor $
• Expecting all rows to match
• Ignoring regex anchors

4. You wrote the query SELECT * FROM users WHERE username ~ '*'; but it returns an error. What is the most likely cause?

medium

A. The regex pattern is missing escape characters for special symbols

B. The ~ operator requires the pattern to be enclosed in double quotes

C. The regex pattern is invalid because it lacks anchors

D. The pattern is correct but the column contains NULL values causing error

Solution

1. Step 1: Check regex pattern syntax

   The pattern * is invalid because the * quantifier requires a preceding atom (e.g., a character or group).

2. Step 2: Identify common error with special characters

   Regex special symbols like *, +, ? need escape characters (e.g., \* in SQL string) when used literally.

3. Final Answer:

   The regex pattern is missing escape characters for special symbols -> Option A

4. Quick Check:

   Escape special regex chars in SQL strings [OK]

Hint: Escape special regex characters properly in SQL strings [OK]

Common Mistakes:

• Using single quotes incorrectly for regex
• Assuming ~ needs double quotes
• Ignoring NULL values causing errors

5. You want to select all rows from orders where the order_id starts with a letter followed by exactly three digits. Which query using the ~ operator is correct?

hard

A. SELECT * FROM orders WHERE order_id ~ '^\w{4}$';

B. SELECT * FROM orders WHERE order_id ~ '^[0-9]{3}[A-Za-z]$';

C. SELECT * FROM orders WHERE order_id ~ '[A-Za-z][0-9]{3}';

D. SELECT * FROM orders WHERE order_id ~ '^[A-Za-z][0-9]{3}$';

Solution

1. Step 1: Understand the pattern requirements

   The order_id must start with a letter [A-Za-z] followed by exactly three digits [0-9]{3}, and nothing else.

2. Step 2: Check regex anchors and pattern

   Using ^ and $ anchors ensures the entire string matches. '^[A-Za-z][0-9]{3}$' matches a letter then three digits exactly.

3. Final Answer:

   SELECT * FROM orders WHERE order_id ~ '^[A-Za-z][0-9]{3}$'; -> Option D

4. Quick Check:

   ^ and $ anchor full string match [OK]

Hint: Use ^ and $ to match entire string pattern [OK]

Common Mistakes:

• Swapping letter and digit order
• Missing anchors causing partial matches
• Using \w which includes digits and letters