Collections serialization

[nmorel]

Collections serialization

[nmorel]

By default, all the Iterable are serialized with IterableJsonSerializer.
For the deserialization, it uses BaseCollectionJsonDeserializer and each implementation just create the Collection implementation by implementing the method newCollection(). See ArrayListJsonDeserializer for an example.

[reinert]

Ok. Thanks.

[nmorel]

Why ? :)

[reinert]

Just to make sure I'm not using unecessary things in my generated serdes powered by gwtjackson.

https://github.com/reinert/requestor/blob/master/src/main/java/io/reinert/requestor/rebind/JsonGwtJacksonGenerator.java#L176 (the critical methods would be: serializeFromCollection and deserializeAsCollection)

[reinert]

BTW... I'm moving turbogwt-http to a separated project called Requestor.
I also created a separated project for the Pomises API called GDeferred which is used by Requestor.

[reinert]

FYI, requestor was officially released yesterday. It has an extension integration the HTTP Client API to gwt-jackson, making a really powerful union.

[nmorel]

Yep, I saw your post on G+.
I read a bit the documentation and read how to use gwt-jackson.
Do we need to annotate every single POJO or only the root one ?

[nmorel]

Also, I may add a simple example using requestor as I did for gwtp. This way, I can see if I break anything.

[reinert]

The way you can tell Requestor to generate and automatically register serializer and deserializer for a type is annotating it with @Json. This was the easiest way I found to provide auto serialization for custom types. Without any information in the types, or about them anywhere else, it was hard to automatically provide serialization for them. So, yes, it's necessary to annotate each POJO you want to have auto serialization. (The same procedure works for the AutoBeans integration, but instead of POJOs one should annotate only Interfaces)

I thank you for exposing requestor here.