Page Contents

A schema imposes restrictions on the model, to ensure (for example) that the model will save data that matches the corresponding database table.

A model can validate data before passing it on to a data store such as a database to ensure that it conforms to the backend schema.

Adding a schema to a model

One way to validate data is to create a model schema; LoopBack will then ensure that data conforms to that schema definition.

For example, suppose your app has a product model. The following code defines a schema and assigns it to the product model. The schema defines two properties: name, a required string property and price, an optional number property. 


var productSchema = {
  "name": { "type": "string", "required": true },
  "price": "number"
var Product = Model.extend('product', productSchema);

If a client tries to save a product with extra properties (for example, description), those properties are removed before the app saves the data in the model. Also, since name is a required value, the model will only be saved if the product contains a value for the name property.

Using validation methods

Every model attached to a persistent data source has validations methods mixed in from Validatable.

Method Description
validatesAbsenceOf Validate absence of one or more specified properties. A model should not include a property to be considered valid; fails when validated field not blank.
validatesExclusionOf Validate exclusion. Require a property value not be in the specified array.

Validate format. Require a model to include a property that matches the given format.

validatesInclusionOf Validate inclusion in set. Require a value for property to be in the specified array.

Validate length. Require a property length to be within a specified range. Three kinds of validations: "min," "max," and "is." Default error messages are:

  • min: too short
  • max: too long
  • is: length is wrong
validatesNumericalityOf Validate numericality. Requires a value for property to be either an integer or number.
validatesPresenceOf Validate presence of one or more specified properties. Requires a model to include a property to be considered valid; fails when validated field is blank.

Validate uniqueness. Ensure the value of the property is unique for the model. Not available for all connectors. Currently supported with these connectors:

  • In Memory
  • Oracle
  • MongoDB

Use these methods to perform specific data validation, as shown in the examples below.

Options object

Most of the validation methods accept an options argument whose properties depend on the specific method being used; however it does have two properties common to all the methods:

  • message - Error message to use instead of the default message, if validation fails.
  • allowNull - Whether null values are allowed.



module.exports = function(user) {
  user.validatesPresenceOf('name', 'email');
  user.validatesLengthOf('password', {min: 5, message: {min: 'Password is too short'}});
  user.validatesInclusionOf('gender', {in: ['male', 'female']});
  user.validatesExclusionOf('domain', {in: ['www', 'billing', 'admin']});
  user.validatesNumericalityOf('age', {int: true});
  user.validatesUniquenessOf('email', {message: 'email is not unique'});

To invoke the validation constraints explicitly, call isValid().

For example:

user.isValid(function (valid) {
  if (!valid) {
    user.errors // hash of errors {attr: [errmessage, errmessage, ...], attr: ...}

Another example of defining validation constraints, this time using a regular expresson:


var re = /^(([^<>()[\]\\.,;:\s@\"]-(\.[^<>()[\]\\.,;:\s@\"]-)*)|(\".-\"))@((\[[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\])|(([a-zA-Z\-0-9]-\.)-[a-zA-Z]{2,}))$/;

UserModel.validatesFormatOf('email', {with: re, message: 'Must provide a valid email'});
if (!(UserModel.settings.realmRequired || UserModel.settings.realmDelimiter)) {
  UserModel.validatesUniquenessOf('email', {message: 'Email already exists'});
  UserModel.validatesUniquenessOf('username', {message: 'User already exists'});

To add validation to model for creating a new model instance, you do not need to call isValid(). You can add validation by simply adding the validator calls:


module.exports = function(MyModel) {
  MyModel.validatesLengthOf('name', { min: 5, message: { min: 'Name should be 5- characters' } });

Use isValid() as an additional _ad-hoc _way to check validity. You can also call validate() or validateAsync() with custom validation functions.

Localizing validation messages

Rather than modifying the error responses returned by the server, you can localize the error message on the client. The validation error response contains error codes in, which enables clients to map errors to localized messages.

Here is an example error response:

  "name": "ValidationError",
  "status": 422,
  "message": "The Model instance is not valid. See error object `details` property for more info.",
  "statusCode": 422,
  "details": {
    "context": "user",
    "codes": {
      "password": [
      "email": [
    "messages": {
      "password": [
       "can't be blank"
      "email": [
        "Email already exists"