Basketry 0.2 is here, and it’s all about making life better for builders. It’s better for the developers using Basketry to generate code, but also for the engineers creating the parsers, rules, and generators that power the ecosystem.
Basketry 0.2 is now available as a Release Candidate, marking a major evolution in how Basketry defines and works with its Intermediate Representation (IR). This release formalizes the IR in JSON Schema, providing a machine-readable contract that enables consistent tooling across languages and environments. By taking a schema-first approach, Basketry can now use its own pipelines to generate TypeScript definitions, documentation, and language-agnostic representations directly from the IR specification—an important step toward a truly polyglot ecosystem.
Basketry recently improved how Data Transfer Objects (DTOs) are handled across both Express and HTTP client generators. This upgrade simplifies your project by centralizing DTO definitions and enhancing data validation through Zod.
This article will walk you through what changed and how to leverage these improvements in your projects.
Learn how to improve your code quality and type safety with discriminated unions, and use Basketry to streamline the process.
Discriminated unions, also known as tagged unions or algebraic data types, are a powerful feature in many modern programming languages that allow you to define a type that can take on several different but related forms.
Evolving an ecosystem with new features, fixes, and improvements.
Basketry is the pluggable, service-oriented code-generation pipeline for any language. It starts with a Service Definition (think Swagger/OpenAPI, API Blueprint, etc.) and generates code for your APIs and the applications that consume them.
The Basketry project is the 3rd or 4th iteration of a toolchain I’ve maintained for the better part of a decade for designing, building, and testing APIs. It started back in the Swagger 2.0 days as a simple tool for defining ExpressJS routes using a “Swagger doc.” The code and architecture are now almost unrecognizable from its long-lost ancestor and it now supports multiple versions of OpenAPI as well as multiple target languages. But nevertheless, it did indeed begin its life as a narrowly focused Swagger v2 tool. For this reason, until recently, it has lacked some features and data structures more closely associated with OpenAPI v3.
Basketry is the pluggable, service-oriented code-generation pipeline for any language. It starts with a Service Definition (think Swagger/OpenAPI, API Blueprint, etc.) and generates code for your APIs.
Let's unpack that.
Basketry isn't one single tool but rather a collection of small tools that work together as a pluggable pipeline. Individually, each component follows the mantra of "do one thing and do it well." Together, they comprise a powerful system for building service-oriented code.
A Basketry pipeline as a whole reads a Service Definition and outputs a set of code files. It does so by composing three types of components: Parsers, Rules, and Generators.
