Are wildcard media types supported in content?

[hkosova]

The 3.0 Specification contains examples where the request and response content includes wildcard media types, e.g. here:
https://github.com/OAI/OpenAPI-Specification/blob/OpenAPI.next/versions/3.0.md#request-body-examples
https://github.com/OAI/OpenAPI-Specification/blob/OpenAPI.next/versions/3.0.md#path-item-object-example

description: user to add to the system
content: 
  'application/json':
    schema:
      $ref: '#/components/schemas/User'
  ...
  '*/*':
    examples:
      user: 
        summary: User example in other format
        externalValue: 'http://foo.bar/examples/user-example.whatever'

Are these examples valid and wildcard types actually supported?

The description of content says just this (emphasis mine):

content
Map[string, Media Type Object]
REQUIRED. The content of the request body. The key is the media type and the value describes it.

For comparison, the Encoding Object explicitly states that wildcard types are allowed:

contentType
... The value can be a specific media type (e.g. application/json), a wildcard media type (e.g. image/*), or a comma-separated list of the two types.

If wildcard types are supported in content, how does this work? Do more specific types have preference over wildcard ones, e.g. image/png > image/* > */*?

[webron]

Perhaps we shouldn't have been explicit in the Encoding Object, and perhaps we should have been more explicit elsewhere. Wildcard media type is a standard valid value.

As for the processing order, what you wrote seems right. I don't know if these rules are provided by any other RFC or if they should be added to the spec (/cc @darrelmiller).

[darrelmiller]

For the encoding object, when the contentType property identifies more than one possible media type, it is effectively saying that any of the matching media types may be present in the multi-part document. It is expected that the content-type header field in the actual request or response will describe the exact media type being sent. I do not believe there is a need to describe any kind of precedence.

[webron]

Fixed by #1295.