Hey! I wonder if there is any way of telling which json fields were missing or had incorrect type wh...

# http4k

[Raf Szałański]

Hey! I wonder if there is any way of telling which json fields were missing or had incorrect type when parsing request using lenses.

Let’s say my endpoint expects requests three fields: id: String, name: String and count: Int and the caller forgets to send the count or maybe sends a boolean there. When parsing such request, http4k will likely throw a lens failure, e.g:

body 'body' must be object
org.http4k.lens.LensFailure: body 'body' must be object

I found this piece of code recently https://github.com/http4k/http4k/blob/master/http4k-core/src/main/kotlin/org/http4k/lens/Validator.kt and thought it’d be fantastic to make it work with json parsing.

Unfortunately, I can only see it being used with web forms and not sure it works with json https://github.com/http4k/http4k/blob/master/http4k-core/src/test/kotlin/org/http4k/lens/ValidatorTest.kt

[Raf Szałański]

To be specific: I wanted to use Validator.Feedback for json parsing but maybe I’m approaching the problem the wrong way?

[dave]

Unfortunately we don't have the technology to provide this because we are reliant on the underlying JSON libraries for unmarshalling. You can extract the first error from a lens failure by interrogating the stack trace (where you get a path-like message of what went wrong) 🙂

[Raf Szałański]

Thanks! I noticed that Lens failures have got very dev-friendly structure and are easy to parse and that it’s the json library which obfuscates the error messages

[Raf Szałański]

I’m working on some solution for one json library that we use with lenses and will try sharing!

[Raf Szałański]

@dave I started building a handler like that which works with jackson and I’d be happy to share it with the community. would you be interested in that?

[dave]

it would be great to see what you've come up with - even if it's just a gist. A Gist with tests even better 😉