关于Fluent Validation是否支持在验证过程中自动修正模型属性值的技术问询

Question #

I'm using Fluent Validation to validate a specific model, and I've written the following validator code:

public class CarModelValidator : AbstractValidator<CarModel> 
{ 
    public CarModelValidator () 
    { 
        RuleFor(x => x.Name).NotEmpty(); 
        // RuleFor(x => x.NrOfDoors).... 
    } 
}

I want to ask: Is it possible to automatically set the NrOfDoors property to 2 if its value is less than 2 during validation? Or does Fluent Validation only handle model validation and feedback results to users, without supporting modifying model property values?

Answer #

Great question! Let's break this down clearly:

First, Fluent Validation's core design purpose is validation, not modifying model state. Its main job is to check if your model meets your specified rules and return validation errors—altering the underlying model data isn't part of its intended use. That said, there is a workaround to implement this behavior, but it's worth noting that this goes against the library's design principles.

If you still want to handle this within the validator, you can use a custom Must rule (or Custom rule) to both validate and adjust the property value. Here's how you could write it:

RuleFor(x => x.NrOfDoors)
    .Must((model, nrOfDoors) => 
    {
        if (nrOfDoors < 2)
        {
            model.NrOfDoors = 2; // Update the property value here
        }
        return true; // Return true to avoid triggering a validation error
    });

If you also want to notify the user that the value was corrected (while still enforcing the rule), you could adjust it like this:

RuleFor(x => x.NrOfDoors)
    .GreaterThanOrEqualTo(2)
    .WithMessage("Number of doors cannot be less than 2; it has been automatically set to 2.")
    .Must((model, nrOfDoors) => 
    {
        if (nrOfDoors < 2)
        {
            model.NrOfDoors = 2;
        }
        return true;
    });

That said, a cleaner and more maintainable approach (aligned with Fluent Validation's intended use) is to handle the property normalization before running validation. This separates your data cleanup logic from your validation logic, making your code easier to follow:

// Normalize the value first, before validation
if (carModel.NrOfDoors < 2)
{
    carModel.NrOfDoors = 2;
}

// Then perform validation as usual
var validator = new CarModelValidator();
var validationResult = validator.Validate(carModel);

内容的提问来源于stack exchange，提问作者user1765862