Closed eed3si9n closed 7 years ago
I find Yoshida-san's encoding better (more obvious, more convenient, ..) Any particular reason you're avoiding it?
I think it's inconsistent with existing project
function.
Plus it's a bit too subtle for me that map
on JF will return only the reader. Doesn't that violate the left identity law? This is because JF inherits JsonReader.
With regards to the project function, maybe we should do something about it: https://github.com/eed3si9n/sjson-new/issues/41.
Given Argonaut does it exactly this way, I guess it's not unlawful? It's an interesting question, though.
http://argonaut.io/scaladocs/index.html#argonaut.CodecJson@map[B]((A)⇒B):DecodeJson[B]
Ref #49, #50, #51
/cc @xuwei-k