Cerberus

About Cerberus

Lightweight and extensible data validation library supporting complex data structures with customizable validation rules. Highly flexible for various validation needs.

Key Features

• 1Schema definition as plain Python dicts — no class subclassing needed
• 2Supports nested documents, lists, and dynamic schema composition
• 340+ built-in rules: type, min, max, regex, allowed values, and more
• 4Custom validators via simple Python functions or method overrides
• 5Lightweight with zero required dependencies

How Python Data Engineers Use Cerberus

Engineers use Cerberus to validate incoming records in ETL pipelines — defining a schema dict that specifies expected types and constraints, then calling `v.validate(record)` for each row. Invalid records are logged or quarantined based on `v.errors`, keeping bad data out of the warehouse while processing continues.