[Rust-Axum][Breaking Change] Improve the oneOf model generator

[Victoria-Casasampere-BeeTheData]

This PR aims to replace the current oneOf and discriminator implementation for the rust-axum generator to one that is usable.

The idea is to generate an enum for oneOf models, with a variant for each specified model. It uses untagged when the discriminator is not specified, and tag = "discriminator" when it is.
This implementation however is not the default present in serde, as it lacks serialization for the discriminator field, a field that has to be specified in all oneOf variants per OAS, and will fail to serialize when present in the generated structure. The custom implementation allows for the field to be present, and still be able to serialize to the adequate model.
Another benefit of the custom serialization is that the variant structure can be serialized directly without the need to convert to the original enum (an option which is still available), which has its benefits when the models have further use outside the API (such as WS messages), and facilitates the use of the frunk crate.

TODO:

• Figure out why ./bin/generate-samples.sh bin/configs/manual/rust-axum-* freezes, but manually generating each model still works.
• Fix the validator implementation of primitive type enum variants.
• Move model modifications from postProcessOperationsWithModels to its own method.
• Support discriminators specified in non oneOf objects...?
• Add a test case for tagged discriminators in oneOf objects.

Fixes #20101

PR checklist

• Read the contribution guidelines.
• Pull Request title clearly describes the work in the pull request and Pull Request description provides details about how to validate the work. Missing information here may result in delayed response from the community.
• Run the following to build the project and update samples:

  ./mvnw clean package || exit
  ./bin/generate-samples.sh ./bin/configs/*.yaml || exit
  ./bin/utils/export_docs_generators.sh || exit
  

  (For Windows users, please run the script in Git BASH)
  Commit all changed files.
  This is important, as CI jobs will verify all generator outputs of your HEAD commit as it would merge with master.
  These must match the expectations made by your contribution.
  You may regenerate an individual generator by passing the relevant config(s) as an argument to the script, for example ./bin/generate-samples.sh bin/configs/java*.
  IMPORTANT: Do NOT purge/delete any folders/files (e.g. tests) when regenerating the samples as manually written tests may be removed.
• File the PR against the correct branch: master (upcoming 7.x.0 minor release - breaking changes with fallbacks), 8.0.x (breaking changes without fallbacks)
• If your PR is targeting a particular programming language, @mention the technical committee members, so they are more likely to review the pull request.

[Victoria-Casasampere-BeeTheData]

This was annoying to me, it broke string formatting, which made debugging slightly harder.

[wing328]

👌

thanks for the fix.

cc @OpenAPITools/generator-core-team

[Victoria-Casasampere-BeeTheData]

@linxGnu @wing328
Ready for review

[wing328]

thanks for the PR

would it be possible to add some oneOf model tests (serialization, deserialization) similar to what we've done in the java client, e.g. https://github.com/OpenAPITools/openapi-generator/blob/master/samples/client/petstore/java/okhttp-gson/src/test/java/org/openapitools/client/JSONTest.java#L410 ?

[Victoria-Casasampere-BeeTheData]

I will have a look at implementing a test case

[Victoria-Casasampere-BeeTheData]

@wing328 Tests for tagged and untagged enums have been implemented and are passing on the CI. Ready for review.

[wing328]

@Victoria-Casasampere-BeeTheData thanks for adding the tests.

@linxGnu can you please review when you've time?

[linxGnu]

@Victoria-Casasampere-BeeTheData thank you for your contribution and happy new year.

Please wait for me, about 1 or 2 days to review.

[linxGnu]

Hi @Victoria-Casasampere-BeeTheData

1. Please find my comments below
2. Could you please describe the scope of the PR? Is it for Response Model only?

[linxGnu]

Could you please explain why we need __ as prefix?

[linxGnu]

Please add tests in another PR

[Victoria-Casasampere-BeeTheData]

2. Could you please describe the scope of the PR? Is it for `Response` Model only?

The models are needed for both incoming (request body) and outgoing (response) requests, which the added test sample includes. An example where this is useful is, for example, a WebSocket service, where you receive different models through the same endpoint, and outgoing models vary as well. This change goes a step further to allow the individual models as a possible response, without the need to upcast to the enum of the endpoint, which is a grateful QOL improvement.

[linxGnu]

LGTM.

I have reviewed everything. It's a wonderful PR from @Victoria-Casasampere-BeeTheData which leverages the power of serde enum internal tag

Kindly ping @wing328 for further actions

[wing328]

@Victoria-Casasampere-BeeTheData thanks for the PR

when you've time, can you please PM me via Slack when you've time?

https://join.slack.com/t/openapi-generator/shared_invite/zt-2wmkn4s8g-n19PJ99Y6Vei74WMUIehQA

@linxGnu thanks for reviewing the change

If anyone prefers to previous/old implementation somehow, we can create an option to fallback to the old template but users will need to maintain it moving forward