I am writing an AWS lambda in Java. Lambda provides two interfaces to implement the request handler.
- RequestHandler - Uses lambda’s serialization for converting input request to InputPojo and vice versa for output.
- RequestStreamHandler – Implementor writes serialization-deserialization logic on input/output stream.
I could not find documentation on serialization behavior of Lambda. I am not sure of questions like
- Will the serde ignore unknown properties?
- Will the serde ignore null fields?
- How to i specify names of json fields
- Can the above behavior change in future?
I wanted to use Jackson annotations to have more control on this. But the documentation specifically states “You shouldn't rely on any other features of serialization frameworks such as annotations..“ So I am planning to use the second interface.
Before proceeding, given all above, want to check the recommended interface to use? which to use when?
Thanks