datumo

Lightweight data modelling for Node.js

Basic Usage

Datumo requires Node.js 4.0 or later.

$ npm install --save datumo

let Datumo = require('datumo')

class Person extends Datumo.Model {
  static get schema () {
    return {
      givenName: { type: 'string', required: true },
      middleName: { type: 'string' },
      familyName: { type: 'string', required: true },
      email: { type: 'string', format: 'email' }
    }
  }
}

let amanda = new Person({
  givenName: 'Amanda',
  familyName: 'Bryson',
  email: '[email protected]'
})

amanda.givenName === 'Amanda'
amanda.familyName === 'Bryson'
amanda.email === '[email protected]'

Person.validate(amanda).valid === true

Why

ES6 provides a lot of new conveniences, with classes arguably as the most notable addition. Datumo makes data modelling using ES6 classes incredibly easy and syntactically compact.

Details

You can do a number of nifty things with Datumo:

• Validate data - harness the power of lx-valid for quick-and-easy validation
• Subclass models - easily create models for subsets of your data
• Extend models - take advantage of ES6 classes to easily extend models
• Map data properties - easily handle data from a 3rd-party that has the data you need, but with different property names
• Avoid extraneous data - Datumo model instances only allow setting properties defined on the schema

Validation

Datumo uses lx-valid for validation. This makes it incredibly easy for you to define your model's schema and solve the problem of validating your data in one shot.

Because Datumo delegates validation to lx-valid, Datumo supports all of the options lx-valid supports on schemas.

let Datumo = require('datumo')

class Person extends Datumo.Model {
  static get schema () {
    return {
      givenName: { type: 'string', required: true },
      middleName: { type: 'string' },
      familyName: { type: 'string', required: true },
      email: { type: 'string', format: 'email' }
    }
  }
}

let jeff = new Person({
  givenName: 'Jeff',
  email: 'Don\'t send me any e-mails!'
})
let validationResults = Person.validate(jeff)

validationResults.valid === false
validationResults.errors.length === 2
// familyName is required
// email is not a valid e-mail address

Validating defined properties only

If you're working with a database, occassionally you may want only to validate properties that are defined as a part of an update or a patch request. With the definedOnly option, the validate function treats all properties on the schema as though they are not required, effectively ignoring any undefined properties.

let vanessa = new Person({
  givenName: 'Vanessa',
  email: 'thisisnotavalidemail'
})
let results = Person.validate(vanessa)
let resultsDefinedOnly = Person.validate(vanessa, { definedOnly: true })

results.valid === false
results.errors.length === 2
// familyName is required
// email is not a valid e-mail address

resultsDefinedOnly.valid === false
resultsDefinedOnly.errors.length === 1
// email is not a valid e-mail address

Extension

Because Datumo models are ES6 classes, you can easily extend them:

let Datumo = require('datumo')

class Person extends Datumo.Model {
  static get schema () {
    return {
      givenName: { type: 'string', required: true },
      middleName: { type: 'string' },
      familyName: { type: 'string', required: true },
      email: { type: 'string', format: 'email' }
    }
  }
}

class Worker extends Person {
  static get schema () {
    let schema = super.schema
    Object.assign(schema, {
      position: { type: 'string', required: true },
      company: { type: 'string', required: true }
    })
    return schema
  }
}

let alex = new Worker({
  givenName: 'Alex',
  familyName: 'Ueltzhöfer',
  position: 'Lead Architectural Engineer',
  company: 'ACME Company'
})

Subclassing

Say you have a model that may contain sensitive information. Or, perhaps, you have a model that has all the data you need, but has a bit too much. With Datumo, it's easy to create a new model that has only the data you need.

Let's assume you have a Person model defined as such:

let Datumo = require('datumo')

class Person extends Datumo.Model {
  static get schema () {
    return {
      givenName: { type: 'string', required: true },
      middleName: { type: 'string' },
      familyName: { type: 'string', required: true },
      email: { type: 'string', format: 'email' }
    }
  }
}

Now, let's create an instance of Person.

let amanda = new Person()

amanda.givenName = 'Amanda'
amanda.middleName = 'Barrett'
amanda.familyName = 'Bryson'
amanda.email = '[email protected]'

Let's assume you don't want certain code to ever see a person's e-mail address. Simple! We can exclude the email property on a new, subclassed model.

class OfflinePerson extends Person.exclude('email') {}

let offlineAmanda = new OfflinePerson(amanda)

offlineAmanda.givenName === 'Amanda'
offlineAmanda.middleName === 'Barrett'
offlineAmanda.familyName === 'Bryson'
offlineAmanda.email === undefined

You can also subclass a subset of the data inclusively instead of exclusively. For example:

class Friend extends Person.subset('givenName', 'email') {}

let friendAmanda = new Friend(amanda)

friendAmanda.givenName === 'Amanda'
friendAmanda.middleName === undefined
friendAmanda.familyName === undefined
friendAmanda.email === '[email protected]'

Mapping

Web services are ubiquitous, and with something widespread, comes the issue of consistency. Many services have certain sets of data that they may share in common. For example, the concept of a user is universal. However, the format in which services serve user data isn't.

Imagine that the property names used for data, both by your app and the services, looks like this:

Your app | Service A | Service B | Service C --- | --- | --- | --- givenName | given_name | firstName | name middleName | middle_name | middleName | middle_initial familyName | family_name | lastName | surname email | email | emailAddress | mail

Datumo helps you navigate the SNAFU of the modern world of data with property mapping. Simply define the mappings on the model:

let Datumo = require('datumo')

class User extends Datumo.Model {
  static get schema () {
    return {
      givenName: { type: 'string', required: true },
      middleName: { type: 'string' },
      familyName: { type: 'string', required: true },
      email: { type: 'string', format: 'email' }
    }
  },
  static get mappings () {
    return {
      serviceA: {
        givenName: 'given_name',
        middleName: 'middle_name',
        familyName: 'family_name'
      },
      serviceB: {
        givenName: 'firstName',
        familyName: 'lastName',
        email: 'emailAddress'
      },
      serviceC: {
        givenName: 'name',
        middleName: 'middle_initial',
        familyName: 'surname',
        email: 'mail'
      }
    }
  }
}

Now, when you receive data from any of these services, accepting that data is as easy as:

apiClient.getUserInfo()
  .then(userInfo => new User(userInfo, { mapping: 'serviceA' }))
  .then(user => db.users.save(user))
  // ...

You can also map data out of a Datumo model to use it in a service. Only properties defined in the mapping will be in the output object. For example, using the sample User class above:

let amanda = new User({
  givenName: 'Amanda',
  middleName: 'Barrett',
  familyName: 'Bryson',
  email: '[email protected]'
})

let apiFriendlyAmanda = User.map(amanda, 'serviceC')

apiFriendlyAmanda.name === 'Amanda'
apiFriendlyAmanda.middle_initial === 'Barrett'
apiFriendlyAmanda.surname === 'Bryson'
apiFriendlyAmanda.mail === '[email protected]'

apiClient.updateUserInfo(amanda.email, apiFriendlyAmanda)
// ...

Expressions

NOTE: Expressions work great with mapping data into a Datumo model. However, they're not designed to work with mapping data out of a Datumo model. If you're using the map function, make sure that you are using a mapping that uses simple property names instead of expressions.

Property mappings support a limited subset of Javascript that you can use when facing data that requires more complex logic to be mapped properly.

For example:

let Datumo = require('datumo')

class User extends Datumo.Model {
  static get schema () {
    return {
      givenName: { type: 'string', required: true },
      // ...
    }
  },
  static get mappings () {
    return {
      serviceA: {
        givenName: 'cn || given_name',
        // ...
      },
      // ...
    }
  }
}

License

MIT