Skip to main content
Version: 3.0

Schema

info

This concept is available for use only through the use of models without the use of decorators, see more.

Orionjs schema has two main purposes. First, it describes the structure of each Model, which will set automatically the GraphQL schema for that Model. Second, if you specify a Model in the setup of a Collection, it will validate the structure of a document before it's get inserted or updated into the MongoDB. You can also use the validate Model function which will throw a ValidationError if the object doesn't fit the given schema. It's important to note that validate is async.

Basic Usage

In Orionjs schemas are defined as a plain object with an specific structure.

Example

import {validate} from '@orion-js/schema'

const bookSchema = {
title: {
type: String,
label: 'Name'
},
author: {
type: String,
label: 'Author'
},
releaseDate: {
type: Date,
label: 'Release Date',
optional: true
}
}

// The following will pass
await validate(bookSchema, {title: 'The Book', author: 'The author'})
await validate(bookSchema, {title: 'The Book', author: 'The author', releaseDate: new Date()})

// The following will throw a ValidationError
await validate(bookSchema, {title: 'The Book'}) // The author is missing
await validate(bookSchema, {title: 'The Book', author: 'The author', releaseDate: 3}) // releaseDate should be type date

Schema rules

There are several options available, the only required is type.

The options available are:

Type

The type property has the following options:

  • String or 'string'.
  • Number or 'number'.
  • Date or 'date'.
  • Boolean or 'boolean'.
  • 'ID' allows you to save an id, which could be a string or an integer.
  • 'email' checks if the email inserted has a valid format.
  • 'integer' allows you to save integers.
  • 'blackbox' allows you to save any object.
  • Custom, you can set as a type any Model.

All the types can be used as array, for example if you want an array of String you should use [String].

Label

The label should be a String. It's used by orionjs-react-autoform.

description

The description should be a String. It will be the description for GraphQL and it's used by orionjs-react-autoform .

optional

By default, all keys are required. Set optional: true to make a field optional.

private

This attribute cannot be queried from the API, not exposed in Graphql.

By default, all keys are not private. Set private: true to make a field private.

min/max

  • If type is Number or 'integer', these rules define the minimum and maximum numeric value.
  • If type is String, these rules define the minimum or maximum string length.
  • If type is Date, these rules define the minimum or maximum date.

allowedValues

Should be an array of values.

  • If type is Number, String, Date or 'integer' it will allow only values specified in this field.
  • If type is an array of any type, it will accept only values specified in this field.

defaultValue

Defines the default value for the field if nothing was given.

validate

This receives a function - which could be async - with the following arguments:

  • fieldValue: It's the value for the given field.
  • options: An object with the following keys:
    • schema: The full schema you are validating against.
    • currentSchema: The subpart of the schema you are validating.
    • doc: The full object you are validating.
    • currentDoc: The subpart of the object you are validating.

With this function you can make a custom validation of the value, it should return the error message if any and nothing if it has been successfully validated

Example, we will make a custom email validation without the use of the type email

const schema = {
email: {
type: String,
validate(fieldValue) {
// Regex to test if an email is valid
const regex =
/^(([^<>()\[\]\\.,;:\s@"]+(\.[^<>()\[\]\\.,;:\s@"]+)*)|(".+"))@((\[[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}])|(([a-zA-Z\-0-9]+\.)+[a-zA-Z]{2,}))$/

if (!regex.test(fieldValue)) return 'notAnEmail'
}
}
}

clean

This receives a function - which could be async - with the following arguments:

  • fieldValue: It's the value for the given field.
  • options: An object with the following keys:
    • doc: The full object you are validating.
    • currentDoc: The subpart of the object you are validating.

With this function you can make a custom clean of the value, it should return the new value for this field