Another question - I'm porting code from ktor 1, and there's a unit test that tests that some code that uses ktor's HttpClient executes properly. Previously the request.body it would pass to the MockEngine would be of type

TextContent

, so the test could inspect it to make sure it had the right content. With ktor 2, however, the request.body is an

OutputStreamContent

- what's the best way for me to verify that its content is correct?

This is a request not a response.

Oh, sorry, missed it.

It looks like

suspend fun OutputStreamContent.asText(): String =
    String(ByteChannel().also { writeTo(it) }.toByteArray(), Charsets.UTF_8)

works, but is there a cleaner way?

Your could use

.body.toByteReadPacket().readText()

instead, but don't know, which is better 😄

That version doesn't make me presume that the type of

body

is

OutputStreamContent

, so it seems better. Thanks!

And it does not enforce utf-8

Indeed - presumably it picks up the right encoding from the request?

Yes, from the data

Great