bravado
bravado copied to clipboard
Yelp
Any plans to support OpenAPI 3.0
OpenAPI 3.0 was released: https://github.com/OAI/OpenAPI-Specification
Are there any plans to add support for this new version?
There are no plans currently to work on supporting version 3. That said, if you're interested please go ahead and create a pull request!
Here's the best of the packages that I've collected while wandering around the jsonschema and openapi landscape:
- connexion
- python-jsonschema-objects
- fastapi -- uses python classes and typing to infer openapi directly (also see molten)
- postgrest (yes, spelled correctly) -- awesome schema-to-rest API translation, introspected from db schema, with impressive support for pushing authn/z to db, schema versioning, and more
Perhaps one of these will work for your goals.
@advance512 https://github.com/swagger-api/swagger-codegen is probably the most popular tool for generating client code from OpenAPI specifications. It can generate a python client based on OpenAPI 3.
It looks like openapi3 might be a viable replacement, as it can be used as a client. You have to manually download the schema and load it as JSON/YAML, but once loaded in the library can make API calls as defined by the spec.
What's the status here? Is bravado deprecated due to state of the art being OpenAPI 3?
@HWiese1980 the fact that there is an OpenAPI 3.0 spec doesn't make this project deprecated. It just means you'll have to use something else to interact with OpenAPI 3 specs.
What's keeping Bravado from becoming OpenAPI 3 compatible?
If you or your organization, or any third-party organizations don't have the resources or interest in making this library openAPI 3 compatible, it never will be. Yelp won't have the resources to do it, unless someone takes over the codebase entirely and works on their own fork, this upgrade is too big to happen here.
I want to thank Yelp and its tremendous contributions to this project, but it is time to move away from Bravado. It is almost 2020, and no new projects should be writing specs in v2, it might end up similar to be the python 2->3 dilemma. Will bravado and swagger (v2) be relevant for the next while? Absolutely. Will it die? Yes.
