FORM SCHEMA VALIDATION

1. Features
2. Installation
3. How to use
4. Constructor
5. Methods
6. Types
7. Example of custom validator
8. Example of additional validator
9. Example of Schema definition
10. Example of schema in schema
11. Schema keys description
12. Custom validation messages
13. Switch of keys validation

Features

• sync validation
• async validation (Promise)
• validate object structure
• validate object keys
• validate required fields
• validate optional fields
• validate by type
• validate by custom type
• validate by one of type
• validate field by custom validators
• validate fields relations by custom validators
• validate whole object tree by custom additional validators

Installation

$ npm install form-schema-validation --save

How to use

Schema give you posibility to validate object using schema validation. You can defined schema and use validate method to check object. Validate method allways returns errors object but if You don't have errors object is empty so You can check errors by

import Schema from 'form-schema-validation';

const schema = new Schema({
    companyName: {
        type: String
    }
});

const modelObject = {
    companyName: 'Test Company'
};

const errors = schema.validate(modelObject); // {}
const error = Object.keys(errors).length > 0; // false

Promises support

You can use validators that return Promise. If You return promis in validator then shema.validate(model) will return Promise.

import Schema from 'form-schema-validation';

const customValidator = {
    validator: (value) => {
        return new Promise((resolve) => {
            setTimeout(() => {
                resolve(value === 'test');
            }, 1000);
        });
    },
    errorMessage: 'async test error',
}

const schema = new Schema({
    companyName: {
        type: String,
        validators:[customValidator]
    }
});

const modelObject = {
    companyName: 'Test Company'
};

const results = schema.validate(modelObject); // Promise
results.then((errors) => {
    console.log(Object.keys(errors).length > 0); // true
});

Constructor

Name	Type	Description
schema	Object	Schema will be used when you validate object
errorMessages	Object	Errors messages that will be displayed on error
validateKeys	Boolean	This flag give you posibility to don't validate object keys not defined in schema

Methods

Name	Attributes	Description
validate	model: Object	Validate Object using defined schema
setError	key: String, message: String, index: Number	Set error on field
setModelError	path: String, message: String	Set error on model tree node
getDefaultValues		Get default values for model using defined schema
getField	name: String	Get field properties extended by parent schema instance (parentSchema)
getFields		Get all fields schemas
oneOfTypes	types: Array of types	Give posibility to validate one of type (Static method)
pick	fieldsToPick: [String]	Get fields from schema by keys
omit	fieldsToOmit: [String]	Get fields from schema and omit fieldsToOmit
extend	fieldsToExtend: [String]	Extend schema by new fields or overwrite them
extendFieldValidators	fieldName: String, validator: { validator: Function, errorMessage: String or Function }	Extend field validators
registerType	type: SchemaType	Register new schema type
isValidatorRegistred	validatorName: String	Check model validator exists in schema
addValidator	validator: Function(model: Object, schema: instance of Schema)	Add model validator
removeValidator	validator: Function	Remove model validator

Types

Name	Description
String	Simple String type
Number	Simple Number type
Object	Simple Object type this type give you posibility to black box
Boolean	Simple Boolean type
Date	This type check value is instance of Date
Array	This type check value is array of any value
new Schema	This type check value is instance of Schema and validate value by this schema
Schema.oneOfType([type1, type2, ...])	This type give you posibility check one of types it will return error if value don't match all types
Schema.optionalType(type)	This type will pass validation if value is null or undefined when field is not required
SchemaType	You can register new schema type that has name, validator, validator when field is required (requiredValidator) and getDefaultValue
[OneOfTypesAbove]	This type check value is array of type

Custom validator attributes

Name	Description
value	Field value
field	Field properties
model	Validated object
schema	Field parent schema instance

Example of custom validator

This validator will check two fields. You can validate one field on base another field.

const validateIfFieldTitleIsFilled = (minLength, message) => ({
    validator: (value, field, model, schema) => {
        if (model.title) {
            return !!value;
        }
        return true;
    },
    errorMessage: message
});

Example of additional validator

Additional validator can set error deep in the objects tree.

const fooSchema = new Schema({
    fooStart: {
        type: String,
    },
    fooEnd: {
        type: String,
    },
});
const modelSchema = new Schema({
    foo: {
        type: fooSchema,
        required: true,
    },
});
const dataModel = {
    foo: {
        fooStart: 'start',
        fooEnd: 'end',
    },
};

modelSchema.addValidator((model, schema) => {
    if(model.foo.fooStart === 'start') {
        schema.setModelError('foo.fooStart', 'my foo error message');
    }
});

modelSchema.validate(dataModel);

Example of dynamic error messages

There can be a need for error messages generated based on the validation outcome. In that case a string or array of strings can be returned from the validator function. Error messages returned from validator function have higher priority that the errorMessage property.

const MIN_AGE = 18;
const validateIfOfAge = () => ({
    validator: (value, fieldSchema, formData) => {
        const { age } = formData;
        if (age <= MIN_AGE) {
            return [`Given ${age} is lower than required age of ${MIN_AGE}`];
        }
    }
});

Example of Schema definition

If You want create new schema You must put object to constructor with information about object keys names and type of value on key.

import Schema from 'form-schema-validation';

const min = (minLength, message) => ({
    validator: (value) => {
        return value.length > minLength;
    },
    errorMessage: message
});

const schema = new Schema({
    companyName: {
        type: String,
        required: true,
        label: 'Company name',
        validators: [min(2, 'Company name should be longer then 2 chars')]
    },
    createdAt: {
        type: Schema.oneOfTypes([Date, String]),
        defaultValue: new Date(),
        label: 'When start'
    },
    workers: {
        type: Number,
        label: 'How many workers we have'
    }
});

Example of schema in schema

import Schema from 'form-schema-validation';

const userSchema = new Schema({
    name: {
        type: String,
        required: true
    },
    surname: {
        type: String,
        required: true
    },
    age: {
        type: Number
    }
});

const groupSchema = new Schema({
    name: {
        type: String,
        required: true,
        label: 'Group name'
    },
    createdAt: {
        type: Date,
        defaultValue: new Date(),
        label: 'Created at'
    },
    members: {
        type: [userSchema],
        label: 'Members'
    }
});

Example of use new schema type

import Schema, { SchemaType } from 'form-schema-validation';

const fooType = new SchemaType('Foo', {
    getDefaultValue() {
        return 'foo';
    },
    validator(value, key) {
        if (value.indexOf('foo') > -1 || value === '') {
            return true;
        }
        this.setError(key, 'foo error');
        return false;
    },
    requiredValidator(value, key) {
        if (value.indexOf('foo') > -1) {
            return true;
        }
        this.setError(key, 'foo required');
        return false;
    },
});

const schema = new Schema({
    foo: {
        type: fooType,
    },
    bar: {
        type: String,
    },
});

const modelWithErrors = {
    foo: 'test',
    bar: '',
};
const modelWithoutErros = {
    foo: '',
    bar: '',
};
const modelWithoutErros2 = {
    foo: 'foo',
    bar: '',
};

Schema keys description

When You defined schema You can use this keys:

Key	Allowed values	Description
companyName, createdAt, workers, ...	any name	this key defined object key name
type	String, Number, Object, Date, Boolean, Array, instance of Schema, [String] ...	this key tell as what type of value we should have on this key in model
required	true, false	this key tell as that field is required
defaultValue	Any	You can set default value for model
disableDefaultValue	Boolean	You can disable filed default value
options	Array of (String, Number, Object, Date, ...)	If you use schema for forms You can defined options for select field
label	Any instance of String	If you use schema for forms You can defined label for form field
validators	array of Functions	You can add custom validators for validate field for example min or max length of value.

Custom validation messages

import Schema from 'form-schema-validation';

const ErrorMessages = {
    notDefinedKey(key) { return `Key '${key}' is not defined in schema`; },
    modelIsUndefined() { return 'Validated model is undefined'; },
    validateRequired(key) { return `Field '${key}' is required`; },
    validateString(key) { return `Field '${key}' is not a String`; },
    validateNumber(key) { return `Field '${key}' is not a Number`; },
    validateObject(key) { return `Field '${key}' is not a Object`; },
    validateArray(key) { return `Field '${key}' is not a Array`; },
    validateBoolean(key) { return `Field '${key}' is not a Boolean`; },
    validateDate(key) { return `Field '${key}' is not a Date`; }
};

const groupSchema = new Schema({
    name: {
        type: String,
        required: true,
        label: 'Group name'
    },
    createdAt: {
        type: Date,
        defaultValue: new Date(),
        label: 'Created at'
    },
    members: {
        type: [userSchema],
        label: 'Members'
    }
}, ErrorMessages);

Switch of keys validation

import Schema from 'form-schema-validation';

const schema = new Schema({
    companyName: {
        type: String,
        required: true
    }
}, false, false);

const modelObject = {
    companyName: 'Test Company',
    _id: 'test1234567890',
};

const errors = schema.validate(modelObject);
console.log(Object.keys(errors).length > 0); // false