Roosterschema

Roosterschema is a declarative schema language and runtime designed to describe and validate structured data across software components. It provides a formal grammar for defining data shapes, constraints, and metadata that can be shared among APIs, configurations, and storage systems. The goal is to establish clear data contracts and improve interoperability between languages and platforms.

Design and origins: Roosterschema arose in the open-source community as an effort to unify elements of existing schema systems with features for modular composition, versioned schemas, and polymorphic data shapes. It is intended to facilitate contract-first design, enabling services to agree on data structure and validation rules before implementation.

Key concepts: The language supports primitive types (string, number, boolean), objects with named fields, arrays, and enumerations. Constraints cover requirements such as required fields, length and range checks, and pattern validation. It enables referencing and importing external schemas, composition through inheritance or overrides, and polymorphism via discriminators. Schemas can include metadata fields like descriptions, examples, and deprecation notes, as well as default values and simple transformation directives.

Syntax and tooling: Roosterschema uses a human-readable, indentation-based syntax similar to YAML, with a JSON-compatible representation available for interchange. Tooling typically includes validators, code generators, and language bindings to streamline integration with popular programming environments.

Applications: Common use cases include API data contracts, configuration validation, data interchange between services, and supporting database migrations. The framework aims to provide a single source of truth for data models, enabling automated client generation and consistent runtime validation.