Finch框架中如何为jsonBody应用ValidationRule校验规则？

Got it, let's walk through how to adapt your existing validation logic for param to work with jsonBody[User] in Finch. The core idea is to either validate the entire parsed User instance or target specific fields after (or during) JSON deserialization.

Option 1: Validate the Entire User Instance #

First, you can reuse your existing mailRule and wrap it into a ValidationRule[User] that checks the email field of the parsed User. Then apply this rule directly to jsonBody[User] using the should method:

case class User(name: String, age: Int, email: String)

// Your existing email validation rule
val mailRule = ValidationRule[String]("mail should be a valid email") { email =>
  """(\w+)@([\w\.]+)""".r.unapplySeq(email).isDefined
}

// Create a validation rule for the full User instance
val userValidationRule = ValidationRule[User]("invalid user: email is not valid") { user =>
  mailRule(user.email).isValid
}

// Apply the rule to jsonBody
val userJson: Endpoint[User] = jsonBody[User].should(userValidationRule)

// Example endpoint using the validated jsonBody
val createUserEndpoint: Endpoint[User] = post("users" :: userJson) { user: User =>
  Ok(user) // Respond with the valid user if all checks pass
}

Option 2: Granular Field Validation with andThen #

If you prefer to handle field validation more explicitly, you can use andThen to chain the JSON parsing with field-level validation. This approach lets you return specific errors for each field:

val userJson: Endpoint[User] = jsonBody[User].andThen { user =>
  mailRule(user.email) // Validate the email field
    .map(_ => user) // Return the valid User if validation passes
}

Adding Multiple Validation Rules #

If you need to validate multiple fields (like age being positive), you can combine rules using a for-comprehension for cleaner code:

// Add an age validation rule
val ageRule = ValidationRule[Int]("age must be a positive number")(_ > 0)

// Combine email and age rules into a single User validation
val combinedUserRule = for {
  _ <- mailRule.contramap[User](_.email) // Map User to its email field for validation
  _ <- ageRule.contramap[User](_.age)    // Map User to its age field
} yield ()

val userJson: Endpoint[User] = jsonBody[User].should(combinedUserRule)

How It Works #

When a client sends an invalid JSON payload (e.g., an email without @), Finch will automatically return a FinchError with your custom error message, just like it does with invalid param values. This keeps your validation behavior consistent across query params and JSON bodies.

内容的提问来源于stack exchange，提问作者hyogo0212