Martin Rademacher

peej - is this considered solved or not? From the above comments I would guess it should work, but it still seems to be broken?

Sounds like a good plan. I've tried hacking phpdoctor a few times (with varying success) and having a defined output to work backwards from should make things a bit easier...

This is a limitation in the type-hints in [swagger-php](https://github.com/zircote/swagger-php). Seems with the advance of type-hints we are getting more of these situations where properties are 'extended' with new value types...

> We may advise to use `attachables` instead of `content` for using `@Model`, I believe this works just as fine although maybe less syntactically pleasant. You are right, of course....

@GuilhemN sure. Always easier to discuss something concrete. Personally I'd prefer an interface so you are not limited by having to extend from yet another class but we'll see.

There is no need to wrap the nested properties in another `Schema`. Try just: ``` * @OA\Property(type="object", property="type", * @OA\Property(type="int", property="id", example="24"), * @OA\Property(type="string", property="name", example="Apartment") * ) ```

Interesting that your example works at all without a `Schema` annotation on the class itself! Using swagger-php directly on the command line with the following file I get the same...

Hmm, yes, that is possible. Its a pain to get both behaving the same and mixing attributes/annotations from different libraries doesn't make it any easier.

You could try to isolate the issue by reducing the code to avoid using bundle annotations while keeping the error. Then you could compare the `openapi` output with the bundle,...

Actually, after creating the issue it occurred to me that I could use a writer... I would like to do the post processing in the same process, as that processing...