SQL Rules

SQL Data Quality Rules - Positive and Negative Conditions

With this new feature, users can now define both positive and negative SQL data quality conditions to validate their data more effectively.

Key Features:

1. Negative Conditions:

   • Users can write SQL queries and select an option from a drop-down indicating that the SQL should not produce any results (default option).
   • This is useful for ensuring that certain data conditions do not occur in the dataset (e.g., no records should have a value below 0 in a specific column).

   \

2. Positive Conditions:

   • Users can write SQL queries and select an option from a drop-down indicating that the SQL should produce results, with the following options:
     • The result count should be greater than 0.
     • The result count should be equal to a fixed value (e.g., exactly 10 results).
   • This allows users to confirm that certain data conditions are met (e.g., at least one record exists with a specific condition).

Example Use Cases:

• Negative Condition: Write SQL to ensure no invalid ages in a dataset, such as:

  SELECT * FROM users WHERE age < 0 OR age > 150;
  

• Positive Condition: Write SQL to ensure valid ages in a dataset, This should return results:

  SELECT * FROM users WHERE age >= 0 OR age < 150;
  

Write SQL rule or Plain English to generate SQL Rule

Write the test case in English in the AI Prompt section and click on Generate SQL.

Warning: Ensure that all the tables required for this test case are selected in the source section. This is necessary to generate SQL from English.

Below are examples to guide you in creating effective test cases for both single-table and multi-table test scripts.

Sample Data

1. Users Table

   | user_id | name | age | credit_limit | | -------- | ------- | --- | ------------- | | 1 | Alice | 25 | 5000 | | 2 | Bob | 34 | 15000 | | 3 | Charlie | 29 | 7000 |

1. Orders Table

   | order_id | user_id | order_amount | order_date | | --------- | -------- | ------------- | ----------- | | 101 | 1 | 300 | 2024-01-15 | | 102 | 2 | 2200 | 2024-01-18 | | 103 | 1 | 450 | 2024-02-05 |

English Inputs and Corresponding SQL

1. Validate that no users are underage

   • Rule: Ensure that all users are at least 18 years old in users table

   • SQL:

     SELECT * FROM Users WHERE age < 18;
     

**Result :** [<mark style="color:green;">**Pass**</mark>](#user-content-fn-1)[^1]

1. Validate that no users are beyond retirement age

   • Rule: Check that no user is older than 30 in Users table.

   • SQL:

     SELECT * FROM Users WHERE age > 30;
     

• Result: Fail

• Failed Record Count = 1

• Validate that all users have a positive credit limit

  • Rule: Confirm that every user has a credit limit greater than zero in users table.

  • SQL:

    SELECT * FROM Users WHERE credit_limit <= 0;
    

    Result: Pass

• Ensure total orders do not exceed credit limits

  • Rule: Validate that the sum of order amounts for each user in Orders table does not exceed their credit limit in Users table.

  • SQL:

    SELECT u.user_id, u.name, SUM(o.order_amount) AS total_orders, u.credit_limit
    FROM Users u
    JOIN Orders o ON u.user_id = o.user_id
    GROUP BY u.user_id
    HAVING SUM(o.order_amount) > u.credit_limit;
    

    Result: Pass

• Check for active users with no orders.

  • Rule: Identify users in users table who have not placed any orders.

  • SQL:

    SELECT u.user_id, u.name
    FROM Users u
    LEFT JOIN Orders o ON u.user_id = o.user_id
    WHERE o.order_id IS NULL;
    

These examples provide a clear demonstration of how English test case inputs can be translated into SQL queries to validate data across different scenarios in a database system.

Warning: Do not include a LIMIT clause in your SQL queries. The engine automatically applies a LIMIT operation to retrieve sample data. For failure counts, the full query is executed as a subquery to accurately calculate the number of failed records.

¹: