[REQ] Support for OpenAPI 3.0-style documented enums

[lemoinem]

Is your feature request related to a problem? Please describe.

With OpenAPI 3.0 there a pattern that allows to document integer-based enums without having to rely on x-enum-varnames and x-enum-descriptions: OAI/OpenAPI-Specification#681 (comment)

Describe the solution you'd like

It would be great to have this pattern supported. Currently, it doesn't generate anything.

Describe alternatives you've considered

We can keep using the old vendor extensions for now. But it would be great to support the new pattern as well.

Additional context

None that I can see.

[jmini]

This is a duplicate for #1693

Documentation for this feature can be found here: https://openapi-generator.tech/docs/templating#all-generators-core

Which version are you using and which generator? It works well for java with 4.0.0 (I am using the latest beta)

I did not read carefully the message.

openapi: 3.0.1
info:
  title: ping test
  version: '1.0'
servers:
  - url: 'https://localhost:9999/'
paths:
  /ping:
    get:
      operationId: pingGet
      responses:
        '201':
          description: OK
components:
  schemas:
    SomeObj:
      type: object
      properties:
        someProp:
          oneOf:
            - enum: [1]
              title: Sunny
              description: Blue sky
            - enum: [2]
              title: Cloudy
              description: Slightly overcast
            - enum: [3]
              title: Rainy
              description: Take an umbrella with you

We could support this, but I am not sure if this is better than x-enum-descriptions.
I am also not sure how much other OpenAPI tool support this (I just tried Swagger-UI and there it is also not the case.).

[lemoinem]

I understand the point. But if all the major tool implementors follow the same line of thoughts, it will never get supported.

Which is not necessarily an issue in itself. However, since it's the recommended way to describe enums by the folks designing JSON-schema and OpenAPI themselves. I believe it would still be great to at least keep it somewhere in the backlog.

Since it looks like something relatively easy to implement, maybe keep it tagged with "help wanted"?
If I can get some time to get around to it, I would be happy to implement it myself actually. Would you accept a PR for such support?

It wouldn't generate new data structure for the generators, just generate the same as what x-enum-names and x-enum-descriptions already do.

[grokify]

I am not sure if this is better than x-enum-descriptions.

It seems like this would solve the parallel arrays issue for using x-enum-descriptions and x-enum-names together.

Regarding the comparison with x-enum-descriptions, is this more comparable to x-enum-values described in this PR:

#1486