@Schema(defaultValue) inside @Parameter on @BeanParam field loses type and enum

[Pepo48]

When a @BeanParam field has @Parameter(schema = @Schema(defaultValue = "ASC")) on an enum-typed field, SmallRye produces a schema with only the default value, type and enum are lost.

I would expect SmallRye to pick up the type and enum values from the Java field type and combine them with the explicitly provided default value.

Example:

public static class QueryOptions {
    @QueryParam("sortOrder")
    @Parameter(description = "Sort direction", schema = @Schema(defaultValue = "ASC"))
    SortOrder sortOrder;  // enum with ASC, DESC
}

# Expected:
schema:
  type: string
  enum: [ASC, DESC]
  default: ASC

# Actual:
schema:
  default: ASC

[hiSandog]

This looks like a merge/target-location bug between @Parameter and the nested @Schema. A regression should put @Schema(defaultValue=..., enumeration=...) on a @BeanParam field and assert the generated parameter schema keeps the Java type and enum values, not just the default string. It would also be worth adding the same annotation directly on a method parameter as a control case.

[MikeEdgar]

@hiSandog I agree. The linked PR includes a test case to verify the parameter's schema overrides the values on the type's schema and also several samples for method argument @Parameters, not just via a @BeanParam.