[BUG][GO] Discriminator not rendering

Bug Report Checklist

• Have you provided a full/minimal spec to reproduce the issue?
• Have you validated the input using an OpenAPI validator (example)?
• Have you tested with the latest master to confirm the issue still exists?
• Have you searched for related issues/PRs?
• What's the actual output vs expected output?
• [Optional] Sponsorship to speed up the bug fix or feature request (example)

Description

I've added a discriminator to our openAPI spec, and I can't seem to get the generated client to unmarshal the response properly.

❯ go run main.go 
panic: unable to unmarshal data into slo.SloResponse: data matches more than one schema in oneOf(SloResponseIndicator)

goroutine 1 [running]:
main.main()
        /Users/nick/work/busted-slo-parsing/main.go:149 +0x148
exit status 2

I would expect some code in this file to check the type field and unmarshal it appropriately.

openapi-generator version

❯ docker run --rm  openapitools/openapi-generator-cli:latest  version
7.0.0-SNAPSHOT

OpenAPI declaration file content or url

https://github.com/wandergeek/busted-slo-parsing/blob/master/bundled.yaml#L1341-L1353

Generation Details

docker run --rm -v "${PWD}:/local" openapitools/openapi-generator-cli:latest generate \
		-i /local/bundled.yaml \
		--git-repo-id terraform-provider-elasticstack \
		--git-user-id elastic \
		-p isGoSubmodule=true \
		-p packageName=slo \
		-p generateInterfaces=true \
		-g go \
		-o /local/generated/slo

Steps to reproduce

git clone git@github.com:wandergeek/busted-slo-parsing.git
cd busted-slo-parsing
make generate-slo-client
go run main.go

Have I just specified this incorrectly? Any help much appreciated!

Reproduce the failure with the bundled.yaml spec using the documented Docker generation command, then run main.go. Start with generated/slo/model_slo_response_indicator.go and compare the discriminator declaration at bundled.yaml lines 1341-1353 with the generated unmarshalling behavior. Done means the generated Go client unmarshals the response without the multiple-schema panic.