Kaval

This is a Kotlin multi-platform library to validate your model.

Why

The goal is to find the right balance between conciseness, expressiveness, and composability to validate POJO.

Or you can see it as a validation DSL.

Example

We want to validate this model

data class User(
    val firstName: String,
    val lastName: String,
    val address: Address?
)

data class Address(
    val line1: String,
    val line2: String,
    val zipCode: Int,
    val city: String
)

We can validate the Address with these constraints:

• line1: not blank, and length <= 255
• line2: length <= 255
• zipCode: > 0
• city: not blank

val addressValidator: Validator<Address> =
    reflectValidator {
        Address::line1 { notBlank and maxLength(255) }
        Address::line2 { maxLength(255) }
        Address::zipCode { greaterThan(0) }
        Address::city { notBlank }
    }

And the User with these constraints:

• firstName: not blank, and length <= 128
• lastName: length <= 255
• address: see above

val userValidator: Validator<User> =
    reflectValidator {
        User::firstName { notBlank and maxLength(128) }
        User::lastName { maxLength(255) }
        User::address { nullOr { Address.validator } }
    }

Now we can use the validators:

val user = User(
   firstName = "",
   lastName = "x".repeat(500),
   address = Address(
       line1 = "",
       line2 = "",
       zipCode = -1,
       city = ""
   )
)

val result: ValidationResult = userValidator.validate(user)
println(result)
// Invalid:
//  - [firstName] requires to be not blank
//  - [lastName.length] requires to be lower or equals to 255, got 500
//  - [address.line1] requires to be not blank
//  - [address.zipCode] requires to be greater than 0, got -1
//  - [address.city] requires to be not blank should be Valid

Modules

• kaval-core: the core, start here
• kaval-reflect: add a validator using Kotlin reflection
• kaval-kotest: add custom matchers for kotest
• kaval-arrow: integration with the awesome Arrow library
• some samples