Identifies a has-many relationship in an Ember Data Model or Ember.Object.
This is used to create a validation collection of the has-many validations.
Note: Validations must exist on all models/objects
Ember Models
// model/users.js
const Validations = buildValidations({
friends: validator('has-many')
});
export default DS.Model.extend(Validations, {
friends: DS.hasMany('user')
});
Ember Objects
// model/users.js
const Validations = buildValidations({
friends: validator('has-many')
});
export default Ember.Object.extend(Validations, {
friends: null
});
From our user model, we can now check validation properties on the friends attribute.
get(model, 'validations.attrs.friends.isValid')
get(model, 'validations.attrs.friends.messages')
-
options -
defaultOptions -
globalOptions
Build options hook. Merges default options into options object. This method gets called on init and is the ideal place to normalize your options. The presence validator is a good example to checkout
Returns:
-
type -
value -
options
Used by all pre-defined validators to build an error message that is present
in validators/message or declared in your i18n solution.
If we extended our default messages to include uniqueUsername: '{username} already exists',
we can use this method to generate our error message.
validate(value, options) {
const exists = false;
// check with server if username exists...
if(exists) {
// The username key on the options object will be used to create the error message
options.username = value;
return this.createErrorMessage('uniqueUsername', value, options);
}
return true;
}
If we input johndoe and that username already exists, the returned message would be 'johndoe already exists'.
Parameters:
Returns:
The generated message
Wrapper method to value that passes the necessary parameters
Returns:
value
-
type -
args
Easily compose complicated validations by using this method to validate against other validators.
validate(value, options, ...args) {
let result = this.test('presence', value, { presence: true }, ...args);
if (!result.isValid) {
return result.message;
}
// You can even test against your own custom validators
result = this.test('my-validator', value, { foo: 'bar' }, ...args);
if (!result.isValid) {
return result.message;
}
result = this.test('number', value, { integer: true }, ...args);
// You can easily override the error message by returning your own.
if (!result.isValid) {
return 'This value must be an integer!';
}
// Add custom logic...
return true;
}
Parameters:
-
typeStringThe validator type (e.x. 'presence', 'length', etc.) The following types are unsupported: 'alias', 'belongs-to', 'dependent', 'has-many'
-
args...argsThe arguments to pass through to the validator
Returns:
The test result object which will contain isValid
and message. If the validator is async, then the
return value will be a promise.
-
value -
options -
model -
attribute
The validate method is where all of your logic should go. It will get passed in the current value of the attribute this validator is attached to. Within the validator object, you will have access to the following properties:
Parameters:
Returns:
One of the following types:
Boolean:trueif the current value passed the validationString: The error messagePromise: A promise that will either resolve or reject, and will finally return eithertrueor the final error message string.