loopback-datasource-juggler/docs/datasource-connector.md

LoopBack DataSource and Connector Guide

Overview

LoopBack is centered around models, which represent data and behaviors. The concept of DataSource is introduced to encapsulate business logic to exchange data between models and various data sources. Data sources are typically databases that provide create, retrieve, update, and delete (CRUD) functions. LoopBack also generalize other backend services, such as REST APIs, SOAP Web Services, and Storage Services, as data sources.

Data sources are backed by connectors which implement the data exchange logic using database drivers or other client APIs. In general, connectors are not used directly by application code. The DataSource class provides APIs to configure the underlying connector and exposes functions via DataSource or model classes.

The diagram above illustrates the relationship between LoopBack Model, DataSource, and Connector.

1. Define the Model using LoopBack Definition Language (LDL). Now we have a model definition in plain JSON or JavaScript object.

2. Create an instance of ModelBuilder or DataSource. Please note that DataSource extends from ModelBuilder. ModelBuilder is responsible for compiling model definitions to JavaScript constructors representing model classes. DataSource inherits that function from ModelBuilder. In addition, DataSource adds behaviors to model classes by mixing in methods from the DataAccessObject into the model class.

3. Use ModelBuilder or DataSource to build a JavaScript constructor (i.e, the model class) from the model definition. Model classes built from ModelBuilder can be later attached to a DataSource to receive the mixin of data access functions.

4. As part of step 2, DataSource initializes the underlying Connector with a settings object which provides configurations to the connector instance. Connector collaborates with DataSource to define the functions as DataAccessObject to be mixed into the model class. The DataAccessObject consists of a list of static and prototype methods. It can be CRUD operations or other specific functions depending on the connector's capabilities.

LoopBack DataSource

DataSource is the unified interface for LoopBack applications to integrate with backend systems. It's a factory for data access logic around model classes. With the ability to plug in various connectors, DataSource provides the necessary abstraction to interact with databases or services to decouple the business logic from plumbing technologies.

Creating a DataSource

The DataSource constructor is available from loopback-datasource-juggler module:

var DataSource = require('loopback-datasource-juggler').DataSource;

DataSource constructor accepts two arguments:

• connector: The name or instance of the connector module
• settings: An object of properties to configure the connector

var dataSource = new DataSource({
    connector: require('loopback-connector-mongodb'),
    host: 'localhost',
    port: 27017,
    database: 'mydb'
});

connector

The connector argument passed the DataSource constructor can be one of the following:

• The connector module from require(connectorName)
• The full name of the connector module, such as 'loopback-connector-oracle'
• The short name of the connector module, such as 'oracle', which will be converted to 'loopback-connector-'
• A local module under ./connectors/ folder

var ds1 = new DataSource('memory');
var ds2 = new DataSource('loopback-connector-mongodb');
var ds3 = new DataSource(require('loopback-connector-oracle'));

Note: LoopBack provides a built-in connector named as memory to use in-memory store for CRUD operations.

settings

The settings argument configures the connector. Settings object format and defaults depends on specific connector, but common fields are:

• host: Database host
• port: Database port
• username: Username to connect to database
• password: Password to connect to database
• database: Database name
• debug: Turn on verbose mode to debug db queries and lifecycle

For connector-specific settings refer to connector's readme file.

Creating a Model

DataSource extends from ModelBuilder, which is a factory for plain model classes that only have properties. DataSource connected with specific databases or other backend systems using Connector.

var DataSource = require('loopback-datasource-juggler').DataSource;
var ds = new DataSource('memory');

var User = ds.define('User', {
    name: String,
    bio: String,
    approved: Boolean,
    joinedAt: Date,
    age: Number
});

All model classes within single data source shares same connector type and one database connection or connection pool. But it's possible to use more than one data source to connect with different databases.

Alternatively, a plain model constructor created from ModelBuilder can be attached a DataSource.

var ModelBuilder = require('loopback-datasource-juggler').ModelBuilder;
var builder = new ModelBuilder();

var User = builder.define('User', {
        name: String,
        bio: String,
        approved: Boolean,
        joinedAt: Date,
        age: Number
});

var DataSource = require('loopback-datasource-juggler').DataSource;
var ds = new DataSource('memory');

User.attachTo(ds); // The CRUD methods will be mixed into the User constructor

More DataSource Features

In addition to data access functions mixed into the model class, DataSource also provides APIs to interact with the underlying backend system.

Discovering model definitions from the data source

Some connectors provide discovery capability so that we can use DataSource to discover model definitions from existing database schema.

The following APIs allow UI or code to discover database schema definitions that can be used to build LoopBack models.

// List database tables and/or views
ds.discoverModelDefinitions({views: true, limit: 20}, cb);

// List database columns for a given table/view
ds.discoverModelProperties('PRODUCT', cb);
ds.discoverModelProperties('INVENTORY_VIEW', {owner: 'STRONGLOOP'}, cb);

// List primary keys for a given table
ds.discoverPrimaryKeys('INVENTORY',  cb);

// List foreign keys for a given table
ds.discoverForeignKeys('INVENTORY',  cb);

// List foreign keys that reference the primary key of the given table
ds.discoverExportedForeignKeys('PRODUCT',  cb);

// Create a model definition by discovering the given table
ds.discoverSchema(table, {owner: 'STRONGLOOP'}, cb);

You can also discover and build model classes in one shot:

// Start with INVENTORY table and follow the primary/foreign relationships to discover associated tables
ds.discoverAndBuildModels('INVENTORY', {visited: {}, relations: true}, function (err, models) {

    // Now we have an object of models keyed by the model name
    // Find the 1st record for Inventory
    models.Inventory.findOne({}, function (err, inv) {
        if(err) {
            console.error(err);
            return;
        }
        console.log("\nInventory: ", inv);

        // Follow the product relation to get information about the product
        inv.product(function (err, prod) {
            console.log("\nProduct: ", prod);
            console.log("\n ------------- ");
        });
    });
});

In addition to the asynchronous APIs, DataSource also provides the synchronous ones. Please refer to the DataSource API references.

Synchronizing model definitions against the data source

DataSource instance have two methods for updating db structure: automigrate and autoupdate for relational databases.

The automigrate method drop table (if exists) and create it again, autoupdate method generates ALTER TABLE query. Both method accepts an optional array of model names and a callback function to be called when migration/update done. If the models argument is not present, all models are checked.

In the following example, we create first version of the CustomerTest model, use automigrate to create the database table, redefine the model with second version, and use autoupdate to alter the database table.

// Create the 1st version of 'CustomerTest'
ds.createModel(schema_v1.name, schema_v1.properties, schema_v1.options);

// Create DB table for the model
ds.automigrate(schema_v1.name, function () {

    // Discover the model properties from DB table
    ds.discoverModelProperties('CUSTOMER_TEST', function (err, props) {
        console.log(props);

        // Redefine the 2nd version of 'CustomerTest'
        ds.createModel(schema_v2.name, schema_v2.properties, schema_v2.options);

        // Alter DB table
        ds.autoupdate(schema_v2.name, function (err, result) {
            ds.discoverModelProperties('CUSTOMER_TEST', function (err, props) {
                console.log(props);
            });
        });
    });
});

To check if any db changes required use isActual method. It accepts and a callback argument, which receive boolean value depending on db state:

• false if db structure outdated

• true when dataSource and db is in sync

  dataSource.isActual(models, function(err, actual) { if (!actual) { dataSource.autoupdate(models, function(err, result) { ... }); } });

LoopBack Connector

Connectors implement the logic to integrate with specific backend systems, such as databases or REST services.

LoopBack Connector Modules

Type	Package Name
Memory	Built-in
MongoDB	loopback-connector-mongodb
Oracle	loopback-connector-oracle
REST	loopback-connector-rest
MySQL	loopback-connector-mysql

Interaction between DataSource and Connector

Initializing connector

The connector module can export an initialize function to be called by the owning DataSource instance.

exports.initialize = function (dataSource, postInit) {

    var settings = dataSource.settings || {}; // The settings is passed in from the dataSource

    var connector = new MyConnector(settings); // Construct the connector instance
    dataSource.connector = connector; // Attach connector to dataSource
    connector.dataSource = dataSource; // Hold a reference to dataSource
    ...
};

The DataSource calls the initialize method with itself and an optional postInit callback function. The connector receives the settings from the dataSource argument and use it to configure connections to backend systems.

Please note connector and dataSource set up a reference to each other.

Upon initialization, the connector might connect to database automatically. Once connection established dataSource object emit 'connected' event, and set connected flag to true, but it is not necessary to wait for 'connected' event because all queries cached and executed when dataSource emit 'connected' event.

To disconnect from database server call dataSource.disconnect method. This call is forwarded to the connector if the connector have ability to connect/disconnect.

Accessing data/services

The connector instance can have an optional property named as DataAccessObject that provides static and prototype methods to be mixed into the model constructor. DataSource has a built-in DataAccessObject to support CRUD operations. The connector can choose to use the CRUD DataAccessObject or define its own.

When a method is invoked from the model class or instance, it's delegated to the DataAccessObject which is backed by the connector.

For example,

User.create() --> dataSource.connector.create() --> Oracle.prototype.create()

Building your own connectors

LoopBack connectors provide access to backend systems including databases, REST APIs and other services. Connectors are not used directly by application code. We create a DataSource to interact with the connector.

For example,

var DataSource = require('loopback-datasource-juggler').DataSource;
var oracleConnector = require('loopback-connector-oracle');

var ds = new DataSource(oracleConnector, {
    host : 'localhost',
    database : 'XE',
    username : 'username',
    password : 'password',
    debug : true
});

Implementing a generic connector

A connector module can implement the following methods to interact with the data source.

exports.initialize = function (dataSource, postInit) {

    var settings = dataSource.settings || {}; // The settings is passed in from the dataSource

    var connector = new MyConnector(settings); // Construct the connector instance
    dataSource.connector = connector; // Attach connector to dataSource
    connector.dataSource = dataSource; // Hold a reference to dataSource

    /**
     * Connector instance can have an optional property named as DataAccessObject that provides
     * static and prototype methods to be mixed into the model constructor. The property can be defined
     * on the prototype.
     */
    connector.DataAccessObject = function {};

    /**
     * Connector instance can have an optional function to be called to handle data model definitions.
     * The function can be defined on the prototype too.
     * @param model The name of the model
     * @param properties An object for property definitions keyed by propery names
     * @param settings An object for the model settings
     */
    connector.define = function(model, properties, settings) {
        ...
    };

    connector.connect(..., postInit); // Run some async code for initialization
    // process.nextTick(postInit);
}

Another way is to directly export the connection function which takes a settings object.

module.exports = function(settings) {
    ...
}

Implementing a CRUD connector

To support CRUD operations for a model class that is attached to the dataSource/connector, the connector needs to provide the following functions:

/**
 * Create a new model instance
 */
CRUDConnector.prototype.create = function (model, data, callback) {
};

/**
 * Save a model instance
 */
CRUDConnector.prototype.save = function (model, data, callback) {
};

/**
 * Check if a model instance exists by id
 */
CRUDConnector.prototype.exists = function (model, id, callback) {
};

/**
 * Find a model instance by id
 */
CRUDConnector.prototype.find = function find(model, id, callback) {
};

/**
 * Update a model instance or create a new model instance if it doesn't exist
 */
CRUDConnector.prototype.updateOrCreate = function updateOrCreate(model, data, callback) {
};

/**
 * Delete a model instance by id
 */
CRUDConnector.prototype.destroy = function destroy(model, id, callback) {
};

/**
 * Query model instances by the filter
 */
CRUDConnector.prototype.all = function all(model, filter, callback) {
};

/**
 * Delete all model instances
 */
CRUDConnector.prototype.destroyAll = function destroyAll(model, callback) {
};

/**
 * Count the model instances by the where criteria
 */
CRUDConnector.prototype.count = function count(model, callback, where) {
};

/**
 * Update the attributes for a model instance by id
 */
CRUDConnector.prototype.updateAttributes = function updateAttrs(model, id, data, callback) {
};