reactive_validator 1.0.0-dev.1

Type safe validation controller with base validators. Allows to manage validation state manually or validate when reactive value changes.

Reactive Validator

Reactive Validator is a type safe validation manager package for Streams and Value Notifiers.

It provides ability to define scoped validation rules for set of stream or value listenable objects. Also it allows to manage error state manually to render validation response from the Server or any other resource.

In addition to error management this package also provides sett of validators for some common validation cases. If need - any callable validator instance can be used to validate target value

Getting Started

Install and import package

import 'package:reactive_validator/reactive_validator.dart';

Create controller instance

/// For controller with stream
StreamValidationController<String> controller = SubjectStreamValidationController<String>();

/// OR for controller with [ValueNotifier]
ListenableValidationController<String> controller = ValueListenableValidationController<String>();

Define and attach value for notification

/// Connectors can be attached or with controller constructor, or with
/// [attachConnectors] method
controller.attachConnectors([
  StreamValidationConnector<String, String>(
    field: 'some_field', /// field name that will be used within the error bucket
    stream: someValueStream.distinct(), /// your [Stream] with field value
    validator: const EmailValidator(),
  ),
  ValueListenableValidationConnector<String, String>(
    field: 'some_other_field', /// field name that will be used within the error bucket
    valueListenable: someValueListenable, /// your [ValueListenable] with field value
    validator: const NotEmptyStringValidator(),
  )
]);

/// or with [attach] method
ValueListenableValidationConnector<String, String>(
  field: 'some_third_field', /// field name that will be used within the error bucket
  valueListenable: someValueListenable, /// your [ValueListenable] with field value
  validator: const NotEmptyStringValidator(),
).attach(controller);

/// or with extension method
someValueListenable.connectValidator(
  field: 'some_fifth_field', /// field name that will be used within the error bucket
  valueListenable: someValueListenable, /// your [ValueListenable] with field value
  validator: const NotEmptyStringValidator(),
).attach(controller);

When you're ready to validate, just validate it

controller.validate();

Or add error message manually

controller.addFieldError('field', 'error');

Validation controller

Controller is a manager of your validation state.

Package provides controller based on ValueNotifier object and controller based on BehaviorSubject stream controller.

Both gives you ability to get error messages or ErrorProvider in a sync way.

Also they're gives you sync access to the all errors Map and validation status.

Controller also gives you ability to add or remove any validation error massage from the error bucket. It can be particularly useful if you validate your values, submit data to the server (if it's valid) and receive validation error, which you want to show to the user.

Important Don't forget to dispose your controller when you no need it anymore.

Invoke validation

To invoke validation against all attached connectors, just call validate method. It returns Future as soon as validation is finished.

StreamValidationController

StreamValidationController, which is implemented by SubjectStreamValidationController aside of default access methods, also gives you access error streams. Like isValidStream, errorsStream, fieldErrorStream etc.

Also, this implementation provides you extended ErrorProvider that contains reference to the stream with field error.

Note Since SubjectStreamValidationController uses BehaviorSubject from the RxDart, StreamErrorProvider also ensures that provided stream won't emit value as soon as anyone subscribes on it.

Validation connectors

Package provides 2 types of validation connectors. Both define relation between value that needs to be validated and validation rules.

Each of them starts to track value changes as soon as it attached to the controller.

Package allows to use both connector types within the same controller instance as long as they using the same field type.

ValueListenableValidationConnector

This validation connector is needed to attach validation rules to ValueListenable object.

  ValueListenableValidationConnector<K,V>({
    K field, /// required, field name
    ValueListenable<V> valueListenable, /// required, value provider
    Validator<V> validator, /// required, validation rules that needs to be applied
    bool validateOnChange = false, /// if value should be validated as soon as it changes
    bool clearOnChange = true, /// if validation error should be cleared as soon as value changes
    bool validateOnAttach = false, /// if value should be validated as soon as connector attached to the controller
  });

To prevent collision, connector can't invoke validation on change and clear validation error on change. Which means that both validateOnChange and clearOnChange can't be true in the same time

StreamValidationConnector

This validation connector is needed to attach validation rules to a Stream object.

  StreamValidationConnector<K,V>({
    K field, /// required, field name
    Stream<V> stream, /// required, value provider
    Validator<V> validator, /// required, validation rules that needs to be applied
    bool validateOnChange = false, /// if value should be validated as soon as it changes
    bool clearOnChange = true, /// if validation error should be cleared as soon as value changes
    bool validateOnAttach = false, /// if value should be validated as soon as connector attached to the controller
  });

  StreamValidationConnector<K,V>.seeded({
    V initialValue, /// required, initial value
    K field, /// required, field name
    Stream<V> stream, /// required, value provider
    Validator<V> validator, /// required, validation rules that needs to be applied
    bool validateOnChange = false, /// if value should be validated as soon as it changes
    bool clearOnChange = true, /// if validation error should be cleared as soon as value changes
    bool validateOnAttach = false, /// if value should be validated as soon as connector attached to the controller
  });

To prevent collision, connector can't invoke validation on change and clear validation error on change. Which means that both validateOnChange and clearOnChange can't be true in the same time.

Stream validation specific

Take into account that this connector will validate value only after at least one event was emitted in the Stream. If you want to validate value even if no event was emitted use StreamValidationConnector.seeded constructor. Or use stream from BehaviorSubject from RxDart package so event could be fired as soon as connector subscribes to it during attach.

Create validation connector

Validation connector can be created in 2 ways. First one with just constructor

final valueConnector = ValueListenableValidationConnector(...);
/// Or
final streamConnector = StreamValidationConnector(...);

And second one with extensions, that package provides

final valueConnector = valueListenable.connectValidator(...);
/// Or
final streamConnector = streamListenable.connectValidator(...);

Validators

Each connector requires validator to be provided.

By default Validator is a callable instance, that returns error message when it's invoked with value.

All build in validators has 2 type of constructors.

• constructor with predefined error message with ability to override it
• constructor with fully custom error message

Some default validators can be found here .

Validation against multiple validators

If more than one validator is required for your field, you can use AndValidator or OrValidator.

Most commonly used is AndValidator, which ensures that value is validated against all provided validators. It stops as soon as any child validator returns error message.

Less commonly used is OrValidator. But it's useful if you have field that may have 2 valid states. For instance if you have not required field, that can be empty string or null or, if provided, contains not less than 6 symbols.

OrValidator([
  IsNullValidator(),
  EmptyStringValidator(),
  MinCharactersValidator(6),
])

Keep in mind that this validator will validate value against all child validators, until any child validator says that value is valid. Which means that validation error message (if the's any) will be returned from the last not valid validator. So ensure that your primary validator is a last one in the OrValidator group.

Custom validation

If you need any custom validation or you want to use some external package validators, you have 2 options.

1. Implement Validator interface

class SomeValidator<I> implements Validator<I> {
  String call(I value) {
    /// Your validation goes here.
    /// Ensure that this method returns error message if [value] is invalid.
    /// 
    /// If it returns [null] validation result will be treated as successful
  }
}

2. Extend from FieldValidator

class SomeValidator extends FieldValidator<String> {
  /// implement your constructor
  const SomeValidator({
    String fieldName,
    String errorMessage,
  })  : super(
    fieldName: fieldName,
    message: errorMessage ?? 'value should be a valid',
  );

  @override
  bool isValid(value) {
    /// you validation goes here
  }
}

3. Use CustomValidator

  final validator = CustomValidator<String>(
    fieldName: 'Field name',
    message: (String value) {
      /// return error message
      /// 
      /// Callback will be invoked only when [isValid] returns false
    },
    isValid: (String value) {
      /// validate [value]
    },
    ignoreNullable: true, /// if nullable value should be treated as valid value
  );
  
  /// Or
  final validator = CustomValidator<String>.withMessage(
    message: (String value) {
      /// return error message
      /// 
      /// Callback will be invoked only when [isValid] returns false
    },
    isValid: (String value) {
      /// validate [value]
    },
    ignoreNullable: true, /// if nullable value should be treated as valid value
  );

First constructor will concat fieldName and message from the message callback.

Second one - will use error message from the builder as it is. So ensure that error message is returned, otherwise ValidationConnector will think that validation is valid

External validator package usage

If needed, this package can use any external validator packages. For instance you can take a look at EmailValidator, which uses email_validator package to validate if value is a valid email value.

Changelog

Please see the Changelog page to know what's recently changed.

Bugs/Requests

If you encounter any problems feel free to open an issue. If you feel the library is missing a feature, please raise a ticket on Github and I'll look into it.

Pull request are also welcome.