Hey, guys. Does Ktor support punycode (googled a bit but found nothing)? Looking for some util methods to convert ascii <-> unicode

Bit of a dense question, but what's puny code? 😄

https://en.wikipedia.org/wiki/Punycode domains written with national characters

what kind support of punycode is required? Isn’t it designed to work using existing standards and do not require special support

So it does converting somewhere behind the scenes?

Ah, see, you just want to use util method

let’s say i want to get url content of `президент.рф`(eg. send GET request to that url) . So i guess this should be supported, right?

okay, it’s more complicated, it should be supported by HTTP client

from browser app textbox, let’s say i’m working on a browser app

than it depend on your http client

and ktor (or underlying http client) should convert it to utf, right?

it should convert it to Punycode ascii from utf

ok, thanks for the explanation. But are you aware of any kind of `expect`ed methods in

common

in ktor that any particular http client provides as

actual

?

I don’t know any MPP implementation

ok, thanks anyway

I mean Ktor probably does nothing about Punycode, it delegates to platform http client

yup, i got it

anyway, if you implementing Browser, than you need a way to encode and decode Punycode, for example browsers has own set of rules what should be decoded and what shouldn’t to avoid fishing etc

that’s exactly what i’m looking for (though i don’t work on browser itself)

just port some existing implementation to Kotlin

yeah, before doing that i just wanted to find smth existing. BTW jvm provides it in

java.net.IDN

so it should not be a problem to implement it for native (what i originally needed) i guess

not be a problem to implement it for native

How Java implementation related to native?

one can write mpp implementation using java’s code as an example

for Native I would just use some existing library probably

i’d love to find/adapt one or even mpp impl

one can write mpp implementation using java’s code as an example

Java delegates to native code. But there is also ICU4J implementation (which is used also for modern Android)

The rules are non trivial if you want to implement it properly. e.g. not mixing different charsets, outside the valid combinations. It's probably best to find good implementation on each platform.

You would be better off writing a wrapper MPP library. For instance to have MPP json parser i could write an expect class in common module and then use say Gson for JVM and JS JSON for js as actual implementations.