@joinbox/node-database
v1.3.0
Published
In-memory database abstraction for consuming relational data sources
Downloads
8
Readme
Node-Database
A simple in-memory database abstraction for consuming relational data sources
Features:
- Read-only implementations for Databases, Collections and Models
- Loading of data using over HTTP for RESTful JSON and SOAP resources
- Refreshing data from the data source
- Data Model definition standard
- Relation resolution between models: hasMany and belongsToMany
- Injection of custom data loaders for implementing alternative protocols
- Injection of custom Model implementations
- Injection of custom Collection implementations
API
A Database has many collections containing each one type of Models. The database accepts a data model definition as a parameter of its constructor. The data model definition defines which models can be loaded from where and which relations they have between each other.
Data Model Definition
The data model definition consists of an array describing the different entities of the database. Each entity can implement different data loaders, models and collections so that the database can be composed of different data sources using different functionality.
const dataModelDefintion = [{
// name of the entity
name: 'user',
// define which data loader to use for the data of this entity, pass it's configuration
RESTDataLoader: {
type: 'RESTDataLoader',
config: {
hostname: 'https://l.dns.porn:2341',
pathname: '/user',
},
},
// define which model implementation to use for this entity
model: {
type: 'Model',
},
// define which collection implementation to use for this entity
collection: {
type: 'Collection',
// define here all relations between the collections
relations: [{
// the type of relation
type: 'hasMany',
// the collection this relations relates to
collection: 'company',
// the key on the other collection this collection is referring to
theirKey: 'id',
// the key on our collection that contains the matching key to the other collection
ourKey: 'companyIds',
// the name of the property where the references models are stored on
name: 'companies',
}],
},
}, {
name: 'companyIds',
RESTDataLoader: {
type: 'RESTDataLoader',
config: {
hostname: 'https://l.dns.porn:2341',
pathname: '/company',
},
},
model: {
type: 'Model',
},
collection: {
type: 'Collection',
relations: [{
type: 'belongsToMany',
collection: 'user',
theirKey: 'companyIds',
ourKey: 'id',
name: 'users',
}],
},
}];
Database Class
Database constructor
The database class is used to set up a full database. It's the main interface for loading the classes, the data model definition and the data.
// set up the database using the data model definition defined above
import Database from '@joinbox/node-database';
const database = new Database({
name: 'users',
dataModelDefintion,
});
database.registerModel(name, Constructor), database.registerRESTDataLoader(name, Constructor), database.registerCollection(name, Constructor) methods
After instantiating the database, you may inject your custom RESTDataLoader, Model and Collection implementations. We're going to use the built in ones for now.
// set up the database using the data model definition defined above
import {
Collection,
Model,
RESTDataLoader,
} from '@joinbox/node-database';
// the first argument of the register functions is the name of the component as it's defined in the data model
// definition
database.registerRESTDataLoader('RESTDataLoader', RESTDataLoader);
database.registerModel('Model', Model);
database.registerCollection('Collection', Collection);
database.setup() and database.load() methods
Now it's time to set up the database and load the data. The Setup method compiles the data model definition and builds the required infrastructure while the the load method loads the data from the remote source.
await database.setup();
await database.load();
After those two calls were issued, the database is ready for use.
database.reload() method
This method reloads all data from the data source. At it's current state it does not refresh the data inside of the models but replaces them all with newly created models.
database.has(collectionName) method
Checks if a collection is part of the database
database.get(collectionName) method
Gets a collection from the database
Collection Class
A collection represents an entity and contains N models.
collection.addModel(modelInstance) method
A private member for adding model instances to the collection
collection.createAndAddModel(dataObject) method
A private member for adding a model by passing the data for the model
collection.setModels(modelArray) method
A private member for replacing all models on the collection
collection.getModels() method
Returns all models from the collection as an array
collection.deleteModel(modelInstance) method
Removes the model passed as parameter from the collection
collection.hasModel(modelInstance) method
Checks if the given model is part of the collection
collection.findModel(propertyName, propertyValue) method
Finds the first model matching the parameters
collection.findById(propertyValue) method
Finds the first model matching the passed id
Model class
A model represents one instance of an entity
model.setData(dataObject) method
Replaces all data on the model with the data passed to the method
model.getData() method
Returns all data of the model as object
model.set(propertyName, value) method
Sets the data of one property
model.get(propertyName) method
Returns the value of one property
model.has(propertyName) method
Checks if a property was set. Returns also true if the value is falsy (null, undefined, false, 0)
Implementing your own Models, Collections or RESTDataLoaders
Please have a look at the following classes which you can extend and inject to the database using the register* method of the database class:
- ./src/model/Model.js
- ./src/collection/Collections.js
- ./src/loader/RESTDataLoaders.js
Debugging and Logging
You may us the extended loggin features of logd to print logs. In order to do so, you have to add the following code to the main file of your application:
import logd from 'logd';
import ConsoleTransport from 'logd-console-transport';
logd.transport(new ConsoleTransport());
When starting the application, you may configure the logger to print logs of different levels via command line parameters
--l
log everything--log-level=debug+
log all level from debug upwards (you may also use the levelsinfo
,warn
anderror
)--log-module=moduleName
log only logs from a given module (the name is printed in violet)