public-data-catalogue

Build

This is a specification for the public-data-catalogue API.

  • specification/ This Open API Specification describes the endpoints, methods and messages exchanged by the API. Use it to generate interactive documentation; the contract between the API and its consumers.
  • sandbox/ This NodeJS application implements a mock implementation of the service. Use it as a back-end service to the interactive documentation to illustrate interactions and concepts. It is not intended to provide an exhaustive/faithful environment suitable for full development and testing.
  • scripts/ Utilities helpful to developers of this specification.
  • proxies/ Live (connecting to another service) and sandbox (using the sandbox container) Apigee API Proxy definitions.

Consumers of the API will find developer documentation on the NHS Digital Developer Hub.

Contributing

Contributions to this project are welcome from anyone, providing that they conform to the guidelines for contribution and the community code of conduct.

Licensing

This code is dual licensed under the MIT license and the OGL (Open Government License). Any new work added to this repository must conform to the conditions of these licenses. In particular this means that this project may not depend on GPL-licensed or AGPL-licensed libraries, as these would violate the terms of those libraries' licenses.

The contents of this repository are protected by Crown Copyright (C).

Development

Requirements

  • make
  • nodejs + npm/yarn
  • poetry
  • Java 8+

Install

$ make install

Updating hooks

You can install some pre-commit hooks to ensure you can't commit invalid spec changes by accident. These are also run in CI, but it's useful to run them locally too.

$ make install-hooks

Environment Variables

Various scripts and commands rely on environment variables being set. These are documented with the commands.

💡 Consider using direnv to manage your environment variables during development and maintaining your own .envrc file - the values of these variables will be specific to you and/or sensitive.

Make commands

There are make commands that alias some of this functionality:

  • lint -- Lints the spec and code
  • publish -- Outputs the specification as a single file into the build/ directory
  • serve -- Serves a preview of the specification in human-readable format

Testing

Each API and team is unique. We encourage you to use a test/ folder in the root of the project, and use whatever testing frameworks or apps your team feels comfortable with. It is important that the URL your test points to be configurable. We have included some stubs in the Makefile for running tests.

VS Code Plugins

Emacs Plugins

Speccy

Speccy A handy toolkit for OpenAPI, with a linter to enforce quality rules, documentation rendering, and resolution.

Speccy does the lifting for the following npm scripts:

  • test -- Lints the definition
  • publish -- Outputs the specification as a single file into the build/ directory
  • serve -- Serves a preview of the specification in human-readable format

(Workflow detailed in a post on the developerjack blog.)

💡 The publish command is useful when uploading to Apigee which requires the spec as a single file.

Caveats

Swagger UI

Swagger UI unfortunately doesn't correctly render $refs in examples, so use speccy serve instead.

Apigee Portal

The Apigee portal will not automatically pull examples from schemas, you must specify them manually.

Platform setup

As currently defined in your proxies folder, your proxies do pretty much nothing. Telling Apigee how to connect to your backend requires a Target Server, which you should call named public-data-catalogue-target. Our Target Servers defined in the api-management-infrasture repository.

💡 For Sandbox-running environments (test) these need to be present for successful deployment but can be set to empty/dummy values.