Are there any solution for this? In my case, I use it for OAS discriminator
annotation. It looks like:
```kotlin
class Pet(val pet: Discriminator("petType") OneOf2<Dog, Cat>)
```
Without TYPE_USE annotation support, I can't get the discriminator
information.
Or are there any better choice to storage the information?
Thank you.
在2017年12月29日星期五 UTC+8 上午5:08:49<Tatu Saloranta> 写道：
> On Thu, Dec 28, 2017 at 2:16 AM, Bojan Tomic <[email protected]> wrote:
> > Yes, TYPE_USE annotations and AnnotatedType were added in Java 8.
> > As you noted, processing these can get quite hairy, but at least some
> > support in Jackson would be very welcome.
> > I am intimately familiar with the complexities involved, as I am the
> > developer behind GeAnTyRef, a fork of GeTyRef (a library similar to
> > ClassMate), that deeply supports working with AnnotatedTypes.
>
> Ah. Yes, your help here is appreciated and we could use your expertise.
> Especially since you have an actual use case -- it's much easier to
> add support for existing
> use case, as opposed to expected or potential one.
>
> > I'm of the opinion Jackson would greatly gain in versatility if this
> style
> > of annotations was supported (as it makes it possible to convey
> > (de)serialization hints on a very granular level).
> > As I've mentioned, I'd be happy to contribute in order to get this
> feature
> > faster, if you'd be interested.
>
> Yes, I think that would make sense.
>
> As I said, `master` branch would be the the way to go at this point.
> Code has been refactored quite a bit
> since 2.9 in some places, but not so much wrt functionality for this,
> I think. But code for 3.0 requires Java 8,
> and 2 out of 3 modules (parameter names, Optional and other new types)
> are part of `jackson-databind` now.
>
> -+ Tatu +-
>
>
> >
> >
> >
> > On Thursday, December 28, 2017 at 6:42:28 AM UTC+1, Tatu Saloranta wrote:
> >>
> >> On Wed, Dec 27, 2017 at 9:33 AM, Bojan Tomic <[email protected]> wrote:
> >> > Just for the sake of experimentation, I've cobbled together the
> >> > following
> >> > wrapper class:
> >> >
> >> > class AnnotatedJavaType extends JavaType implements AnnotatedElement {
> >> >
> >> > private final JavaType base;
> >> > private final Annotation[] annotations;
> >> >
> >> > AnnotatedJavaType(JavaType base, Annotation[] annotations) { ... }
> >> >
> >> > //all methods simply delegate to this.base
> >> > }
> >> >
> >> >
> >> >
> >> > And the corresponding TypeFactory wrapper:
> >> >
> >> > class AnnotatedTypeFactory {
> >> >
> >> > JavaType construct(AnnotatedType type, TypeFactory typeFactory) {
> >> > if (type instanceof AnnotatedParameterizedType) {
> >> > JavaType[] args = Arrays.stream(((AnnotatedParameterizedType)
> >> > type).getAnnotatedActualTypeArguments())
> >> > .map(arg -> fromJavaType(arg, typeFactory))
> >> > .toArray(JavaType[]::new);
> >> >
> >> > TypeBindings bindings =
> >> > TypeBindings.create(ClassUtils.getRawType(type.getType()), args);
> >> >
> >> > return new
> >> > AnnotatedJavaType(typeFactory.constructType(type.getType(),
> >> > bindings), type.getAnnotations());
> >> > }
> >> >
> >> > //handle wildcards, arrays, variables in the similar fashion
> >> >
> >> > return new
> >> > AnnotatedJavaType(typeFactory.constructType(type.getType()),
> >> > type.getAnnotations());
> >> > }
> >> > }
> >> >
> >> > Amusingly enough, it works! I'd never keep it in my code base, but it
> >> > was an
> >> > interesting experiment.
> >> >
> >> > If the functionality I'm after is impossible (and I really hope it is
> >> > not),
> >> > would you be open to a contribution? I'd of course need some
> >> > consultation/guidance on the approach, but would be happy to develop
> it.
> >>
> >> As per my other note, this would need to be in Jackson 3.0 I think,
> >> due to multiple reasons.
> >>
> >> I am bit hesitant to say much more since I really fear all complexity
> >> that would come from trying
> >> to create parallel annotation processing system. I can see some
> >> benefits from type annotations,
> >> but it wouldn't come for free, if we are talking about whole
> >> AnnotationIntrospector style system.
> >>
> >> However: if the idea would be to do something simpler -- I don't
> >> really see need for full AnnotationIntrospector
> >> in this case, to be honest -- for example, simply keep track of
> >> annotations for type, offer access... well,
> >> that seems quite doable and useful. If so, `JavaType` could just have
> >> `findAnnotation(Class<?> type)` method
> >> or something?
> >>
> >> -+ Tatu +-
> >>
> >>
> >>
> >> >
> >> >
> >> >
> >> >
> >> > On Wednesday, December 27, 2017 at 11:34:37 AM UTC+1, Bojan Tomic
> wrote:
> >> >>
> >> >> I have a rather exotic use-case that I have to support. Types
> annotated
> >> >> with @Id need custom deserialization logic, e.g. the value should be
> >> >> Base64
> >> >> decoded before deserializing it (not the only case).
> >> >>
> >> >> This means a type such as List<@Id Key> will be represented as a list
> >> >> of
> >> >> Base64 Strings that I deserialize using my custom logic.
> >> >>
> >> >> I originally have an AnnotatedType that I convert to JavaType using
> >> >> TypeFactory.constructType(annotatedType.getType()) prior to
> >> >> deserialization,
> >> >> which of course looses all the annotations.
> >> >> Later, in my custom Deserializers instance, in its various
> >> >> findXXXDeserializer methods, I have no way to find the correct
> >> >> deserializer
> >> >> as the annotations needed for the decision are gone...
> >> >>
> >> >> I was hoping the JDK8 module would have it's own TypeFactory-like
> >> >> mechanism to construct JavaType subclasses that preserve the
> >> >> annotations
> >> >> from AnnotatedTypes (e.g.
> >> >> AnnotatedTypeFactory.constructType(annotatedType)), but it doesn't
> seem
> >> >> to
> >> >> be the case.
> >> >>
> >> >> Is there any way at all to implement what I'm after?
> >> >
> >> > --
> >> > You received this message because you are subscribed to the Google
> >> > Groups
> >> > "jackson-user" group.
> >> > To unsubscribe from this group and stop receiving emails from it, send
> >> > an
> >> > email to [email protected].
> >> > To post to this group, send email to [email protected].
> >> > For more options, visit https://groups.google.com/d/optout.
> >
> > --
> > You received this message because you are subscribed to the Google Groups
> > "jackson-user" group.
> > To unsubscribe from this group and stop receiving emails from it, send an
> > email to [email protected].
> > To post to this group, send email to [email protected].
> > For more options, visit https://groups.google.com/d/optout.
>
--
You received this message because you are subscribed to the Google Groups
"jackson-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jackson-user/e5f815f4-fac5-495a-9201-ca4f2a97acf0n%40googlegroups.com.
