Overview
Important: Please read Relations first.
Note:
This relation best works with databases that support foreign key constraints (SQL). Using this relation with NoSQL databases will result in unexpected behavior, such as the ability to create a relation with a model that does not exist. We are working on a solution to better handle this. It is fine to use this relation with NoSQL databases for purposes such as navigating related models, where the referential integrity is not critical.
A belongsTo
relation denotes a many-to-one connection of a model to another
model through referential integrity. The referential integrity is enforced by a
foreign key constraint on the source model which usually references a primary
key on the target model. This relation indicates that each instance of the
declaring or source model belongs to exactly one instance of the target model.
For example, in an application with customers and orders, an order always
belongs to exactly one customer as illustrated in the diagram below.
The diagram shows the declaring (source) model Order has property customerId as the foreign key to reference the target model Customer’s primary key id.
To add a belongsTo
relation to your LoopBack application and expose its
related routes, you need to perform the following steps:
- Add a property to your source model to define the foreign key.
- Modify the source model repository class to provide an accessor function for obtaining the target model instance.
- Call the accessor function in your controller methods.
Defining a belongsTo Relation
This section describes how to define a belongsTo
relation at the model level
using the @belongsTo
decorator to define the constraining foreign key.
LB4 also provides an CLI tool lb4 relation
to generate belongsTo
relation
for you. Before you check out the
Relation Generator
page, read on to learn how you can define relations to meet your requirements.
Relation Metadata
LB4 uses three keyFrom
, keyTo
and name
fields in the belongsTo
relation
metadata to configure relations. The relation metadata has its own default
values for these three fields:
Field Name | Description | Default Value | Example |
---|---|---|---|
keyFrom |
the foreign key of the source model | the target model name appended with `Id` in camel case | Order.customerId |
keyTo |
the source key of the target model | the primary key in the target model | Customer.id |
name |
the name of the relation | @belongsTo decorated property name without Id |
customer |
We recommend to use default values. If you’d like to customize the foreign key name or the relation name, you’ll need to specify some fields through the relation decorator.
The standard naming convention for the foreign key property in the source model
is target name
+ Id
(for example, Order.customerId).
import {belongsTo, Entity, model, property} from '@loopback/repository';
import {Customer} from './customer.model';
@model()
export class Order extends Entity {
@property({
type: 'number',
id: true,
})
id: number;
@belongsTo(() => Customer)
customerId: number; // relation name will default to `customer`
@property({type: 'number'})
quantity: number;
constructor(data: Partial<Order>) {
super(data);
}
}
export interface OrderRelations {
// describe navigational properties here
}
export type OrderWithRelations = Order & OrderRelations;
If the foreign key property name in the source model has to be customized
(customer_id
instead of customerId
for example), the relation name has to be
explicitly specified in the name
attribute of the relation definition.
Otherwise the default relation name generates by LB4 (customer_id
in this
case) will be the same as the customized foreign key name, which is invalid.
Warning: Make sure that you have different names for the foreign key and the relation name in BelongsTo relations.
// import statements
@model()
export class Order extends Entity {
@property({
type: 'number',
id: true,
})
id: number;
@belongsTo(() => Customer, {name: 'customer'}) // specify the relation name if fk is customized
customer_id: number; // customized fk
// other properties, constructor, etc.
}
In addition, if you have a corresponding hasMany
or hasOne
relation in the
target model (for example, a Customer has many Orders), the keyTo
attribute of
that corresponding relation needs to be stated explicitly:
// import statements
@model()
export class Order extends Entity {
// constructor, properties, etc.
@belongsTo(() => Customer, {name: 'customer'})
customer_id: number; // customized foreign key name
}
@model()
export class Customer extends Entity {
// constructor, properties, etc.
@hasMany(() => Order, {keyTo: 'customer_id'})
orders: Order[];
}
Check the relation metadata in hasMany and hasOne for more details.
If you need to use different names for models and database columns, to use
customer_id
as db column name other than customerId
for example, passing the
column name in the third argument to the belongsTo
decorator would allow you
to do so:
class Order extends Entity {
// constructor, properties, etc.
@belongsTo(() => Customer, {keyFrom: 'customerId'}, {name: 'customer_id'})
customerId: number;
}
If you need to use another attribute other than the id property to be the source
key of the target model (joining two tables on non-primary attribute), the
keyTo
attribute in the relation definition has to be stated explicitly.
class Order extends Entity {
// constructor, properties, etc.
@belongsTo(() => Customer, { keyTo: 'customized_target_property' }),
customerId: number;
}
export interface OrderRelations {
customer?: CustomerWithRelations;
}
Important: LB4 doesn’t support composite keys for now. e.g joining two tables with more than one source key. Related GitHub issue: Composite primary/foreign keys
Configuring a belongsTo relation
The configuration and resolution of a belongsTo
relation takes place at the
repository level. Once belongsTo
relation is defined on the source model, then
there are a couple of steps involved to configure it and use it. On the source
repository, the following are required:
- In the constructor of your source repository class, use Dependency Injection to receive a getter function for obtaining an instance of the target repository. Note: We need a getter function, accepting a string repository name instead of a repository constructor, or a repository instance, in order to break a cyclic dependency between a repository with a belongsTo relation and a repository with the matching hasMany relation.
- Declare a property with the factory function type
BelongsToAccessor<targetModel, typeof sourceModel.prototype.id>
on the source repository class. - call the
createBelongsToAccessorFor
function in the constructor of the source repository class with the relation name (decorated relation property on the source model) and target repository instance and assign it the property mentioned above.
The following code snippet shows how it would look like:
import {Getter, inject} from '@loopback/core';
import {
BelongsToAccessor,
DefaultCrudRepository,
juggler,
repository,
} from '@loopback/repository';
import {Customer, Order, OrderRelations} from '../models';
import {CustomerRepository} from '../repositories';
export class OrderRepository extends DefaultCrudRepository<
Order,
typeof Order.prototype.id,
OrderRelations
> {
public readonly customer: BelongsToAccessor<
Customer,
typeof Order.prototype.id
>;
constructor(
@inject('datasources.db') protected db: juggler.DataSource,
@repository.getter('CustomerRepository')
customerRepositoryGetter: Getter<CustomerRepository>,
) {
super(Order, db);
this.customer = this.createBelongsToAccessorFor(
'customer',
customerRepositoryGetter,
);
}
}
BelongsToAccessor
is a function accepting the primary key (id) of a source
model instance (e.g. order.id
) and returning back the related target model
instance (e.g. a Customer
the order belongs to). See also
API Docs
Note:
Notice that OrderRepository.create()
expects an Order
model only, navigational properties are not expected to be included in the target data. For instance, the following request will be rejected:
orderRepository.create({
` id: 1,
customerId: 1
customer:{id: 1, name: ‘rejected’}
})`
Using BelongsToAccessor in a controller
The same pattern used for ordinary repositories to expose their CRUD APIs via
controller methods is employed for belongsTo
relation too. Once the belongsTo
relation has been defined and configured, a new controller method can expose the
accessor API as a new endpoint.
import {repository} from '@loopback/repository';
import {get} from '@loopback/rest';
import {Customer, Order} from '../models/';
import {OrderRepository} from '../repositories/';
export class OrderController {
constructor(
@repository(OrderRepository) protected orderRepository: OrderRepository,
) {}
// (skipping regular CRUD methods for Order)
@get('/orders/{id}/customer')
async getCustomer(
@param.path.number('id') orderId: typeof Order.prototype.id,
): Promise<Customer> {
return this.orderRepository.customer(orderId);
}
}
In LoopBack 3, the REST APIs for relations were exposed using static methods
with the name following the pattern __{methodName}__{relationName}__
(e.g.
Order.__get__customer
). While we recommend to create a new controller for each
hasMany relation in LoopBack 4, we also think it’s best to use the main CRUD
controller as the place where to expose belongsTo
API.
Handling recursive relations
Given an e-commerce system has many Category
, each Category
may have several
sub-categories, and may belong to 1 parent-category.
export class Category extends Entity {
@property({
type: 'number',
id: true,
generated: true,
})
id?: number;
@hasMany(() => Category, {keyTo: 'parentId'})
categories?: Category[];
@belongsTo(() => Category)
parentId?: number;
constructor(data?: Partial<Category>) {
super(data);
}
}
export interface CategoryRelations {
categories?: CategoryWithRelations[];
parent?: CategoryWithRelations;
}
export type CategoryWithRelations = Category & CategoryRelations;
The CategoryRepository
must be declared like below
export class CategoryRepository extends DefaultCrudRepository<
Category,
typeof Category.prototype.id,
CategoryRelations
> {
public readonly parent: BelongsToAccessor<
Category,
typeof Category.prototype.id
>;
public readonly categories: HasManyRepositoryFactory<
Category,
typeof Category.prototype.id
>;
constructor(@inject('datasources.db') dataSource: DbDataSource) {
super(Category, dataSource);
this.categories = this.createHasManyRepositoryFactoryFor(
'categories',
Getter.fromValue(this),
);
this.parent = this.createBelongsToAccessorFor(
'parent',
Getter.fromValue(this),
); // for recursive relationship
}
}
DO NOT declare
@repository.getter(CategoryRepository) protected categoryRepositoryGetter: Getter<CategoryRepository>
on constructor to avoid “Circular dependency” error (see
issue #2118)
Querying related models
Different from LB3, LB4 creates a different inclusion resolver for each relation
type to query related models. Each relation has its own inclusion resolver
inclusionResolver
. And each repository has a built-in property
inclusionResolvers
as a registry for its inclusionResolvers. Here is a diagram
to show the idea:
A belongsTo
relation has an inclusionResolver
function as a property. It
fetches target models for the given list of source model instances.
Use the relation between Customer
and Order
we show above, an Order
belongs to a Customer
.
After setting up the relation in the repository class, the inclusion resolver allows users to retrieve all orders along with their related customers through the following code at the repository level:
orderRepo.find({include: ['customer']});
or use APIs with controllers:
GET http://localhost:3000/orders?filter[include][]=customer
Enable/disable the inclusion resolvers
- Base repository classes have a public property
inclusionResolvers
, which maintains a map containing inclusion resolvers for each relation. - The
inclusionResolver
of a certain relation is built when the source repository class calls thecreateBelongsToAccessorFor
function in the constructor with the relation name. - Call
registerInclusionResolver
to add the resolver of that relation to theinclusionResolvers
map. (As we realized in LB3, not all relations are allowed to be traversed. Users can decide to which resolvers can be added.) The first parameter is the name of the relation.
The following code snippet shows how to register the inclusion resolver for the belongsTo relation ‘customer’:
export class OrderRepository extends DefaultCrudRepository {
customer: BelongsToAccessor<Customer, typeof Order.prototype.id>;
constructor(
dataSource: juggler.DataSource,
customerRepositoryGetter: Getter<CustomerRepository>,
) {
super(Order, dataSource);
// we already have this line to create a BelongsToRepository factory
this.customer = this.createBelongsToAccessorFor(
'customer',
customerRepositoryGetter,
);
// add this line to register inclusion resolver.
this.registerInclusionResolver('customer', this.customer.inclusionResolver);
}
}
-
We can simply include the relation in queries via
find()
,findOne()
, andfindById()
methods. For example, these queries return all orders with theirCustomer
:if you process data at the repository level:
orderRepository.find({include: ['customer']});
this is the same as the url:
GET http://localhost:3000/orders?filter[include][]=customer
which returns:
[ { id: 1, description: 'Mjolnir', customerId: 1, customer: { id: 12, name: 'Thor', }, }, { id: 2, description: 'Shield', customer: { id: 10, name: 'Captain', }, }, { id: 3, description: 'Rocket Raccoon', customerId: 1, customer: { id: 12, name: 'Thor', }, }, ];
Here is a diagram to make this more intuitive:
- You can delete a relation from
inclusionResolvers
to disable the inclusion for a certain relation. e.gorderRepository.inclusionResolvers.delete('customer')
Query multiple relations
It is possible to query several relations or nested include relations with custom scope once you have the inclusion resolver of each relation set up. CheckHasMany - Query multiple relations for the usage and examples.