Working with Custom Data Types

Learn to extend Pydantic's functionality by defining and using custom data types for more granular control over your data validation.

By Kevin McAleer,    2 Minutes

Extending Pydantic with Custom Data Types

Pydantic’s powerful validation system can be extended with custom data types, enabling you to define bespoke validation logic and serialization/deserialization rules for specific data structures.

Defining Custom Data Types

A custom data type in Pydantic is usually defined by extending existing base types and adding custom validation or transformation logic. Here’s how you can create a custom type for handling email addresses:

from pydantic import BaseModel, EmailStr

class Email(EmailStr):
    @classmethod
    def __get_validators__(cls):
        yield cls.validate

    @classmethod
    def validate(cls, v):
        # Place additional custom validation logic here
        if not "@" in v:
            raise ValueError("Invalid email address")
        return v.lower()

This example shows a custom Email type that extends EmailStr, a built-in Pydantic type for email validation. It adds additional logic to ensure all emails are converted to lowercase.

Using Custom Data Types in Models

Once you’ve defined a custom data type, you can use it in your models just like any built-in type.

class User(BaseModel):
    name: str
    email: Email  # Using the custom Email type

In this model, the email field uses the custom Email data type, applying both the built-in validation from EmailStr and the additional custom logic.

Advantages of Custom Data Types

• Flexibility: Tailor validation logic to the specific needs of your application.
• Reusability: Define once and use across multiple models, ensuring consistency.
• Clarity: Improve code readability by using semantically meaningful types.

Lesson Assignment

Define a custom data type for PhoneNumber that:

• Validates that the input is a string formatted as a phone number (e.g., starts with “+” followed by country code and number).
• Normalizes the phone number by removing spaces and dashes.

Use this custom type in a Contact model that includes fields for name and phone_number. Test your model with various phone number formats to ensure your validation and normalization logic works as expected.

Additional Resources

• Pydantic Custom Types Documentation
• Python Regular Expressions for Pattern Matching