Ktor에 직렬화 연결

You could decode every response manually, but Ktor can wire serialization in so requests and replies become typed automatically.

The ContentNegotiation plugin teaches the HttpClient how to convert request and response bodies to and from your models.

First add the Ktor content-negotiation artifact plus the JSON serializer module to commonMain.

You also include the ktor-serialization-kotlinx-json module, which connects Ktor to kotlinx.serialization specifically.

Inside the HttpClient builder you install ContentNegotiation and register the json converter once.

The json block accepts a configured Json instance, so you can pass ignoreUnknownKeys and friends right here.

Now a GET can return your model directly. Ktor reads the body and hands you a typed object with no manual decode call.

The same plugin serializes outgoing data, so you can setBody with a Kotlin object and Ktor writes the JSON for you.

When sending a body, declare contentType as application/json so the server reads it correctly.

Because all of this lives in commonMain, Android and iOS share the identical typed networking with zero duplicate code.

Ktor plus serialization gives you a tidy pipeline: call an endpoint, get a model, send a model, all type-safe across platforms.

Which Ktor plugin enables automatic JSON decoding of responses?

Install ContentNegotiation with json() on your HttpClient so Ktor auto-encodes request bodies and decodes responses into typed models on both apps.