AutoValidator's aim is to create a simple, fluent and intiutive framework to validate data and models in your dotnet applications. Simply put; an application would validate some data and get a result object. That object would state sucess or failure and detail the failing data.
There are two basic ways to validate data:
- Simply call validate on each item of data as you need to.
- Define a validation schema for a model, then call validate on an instance of that model. (preferred approach)
the validation schema process is heavily guided by AutoMapper.
If you just want to validate individual variables, simply create an instance of Validator
.
var validator = new Validator();
If you wish to use schema validation; first create an instance of AutoValidation
, define the ValidatorConfigurationExpression
then create a factory to create instance of validators.
Action<IValidatorConfigurationExpression> configure = cfg =>
{
cfg.AddProfile<ModelToBeValidatedProfile>();
};
var validator = new AutoValidation(configure);
var factory = mapper.CreateFactory();
var validator = factory.Create<ModelToBeValidated>();
You only need to create one instance of the AutoValidation, its configuration and the factory. Then use the factory instance to create new instances of the validator as required.
For the basic Validator, validations occur as that line of code executes. For Generic Validators the expressions are stored and only executed when .Validate()
is called.
within the initial configuration setup, you have access to the settings. These allow you to setup the way the application will run, such as:
- UseCamelCase - The system will camelCase the property names (default is false) - See BasicUseCaseExamples project, HomeController for an example.
For further information about configuration and setup
There are two types of validators. Generic and non generic.
- Generic validators use the schema validation process set out in configuration and expression setup
- Non Generic validators are for validating individual variables.
var validator = factory.Create<ModelToBeValidated>();
var model = new ModelToBeValidated();
var result = validator.Validate(model);
var validator = new Validator();
var result = validator.IsEmailAddress(someVariable).Validate();
validations can be used in a fluent fashion.
var validator = new Validator();
var result = validator
.isMinValue(someInt, 18)
.IsEmailAddress(someVariable)
.Validate();
- include all properties in validaiton result regardless of error or not
- object validation does not always deal with custom expressions for the format string.
- an extension method that turns the result into an extension of the original object
- validating lists of simple types
- add option to pass in a validationSpec object (TV) as well as the object to be validated. this can mean that the user can define dynamic expressions in a profile (i.e. the params used can change dynamically during execution)
- allow a dynamic object to be b used for validationSpec instead of TV, only apply that rule if a property has a value.
- test error message override in classValidator, regular string validator and regular fluent validator
- think of way to pass in values to mapping profiles, think it could be a validationParameters object (VT), would mean it can be used in any of the expressions.
- add use mapping expression to the iClassValidator so that it will know to check for other mapping expressions to validate that object
- validating single child property (add mappingexpression collection object, that will recursively get all child mappings)
- validating child property lists
- regular validator could make prop name optional. if not given could just return general errors, rename errors to prop errors
- have a way of setting some values in the validator initial setup
- add more IValidatorExpression functions (have a look at https://github.com/gnpretorius/simple-validator)