Schematiq - JavaScript object schematiqs

A lightweight library that allows for object handling based on various schemas. The library includes object validation, (de-)normalization and transformation.

Object validation

For object validation one defines rules for each property of the object. Combined, rules define the validation schema. Each rule must contain a type. Other settings are optional.

type Rule = {
  type: 'string' | 'boolean' | 'number' | 'array'; // type error
  required?: boolean; // required error
  message?: 'my custom message';
  regexp?: RegExp; // format error, only when type = 'string'
  rule?: (value, obj) => boolean; // other error
  each?: Schema; // required when type = 'array'
};

type Schema = {
  [key: string]: Rule; // also for nested properties, e.g. "nested.property"
};

when

An object can be validated by using the validate function of schematiq. It returns an object indicating which properties of the object have errors. It also indicates the type of error, unless you set a custom error message.

import { validate } from 'schematiq';

const errors = validate(obj, schema);
// { "nested.property": "type" | "required" | "format" | "other" | "my custom message" }

Object normalization and denormalization

Normalization allows for deep nested objects to be transformed into a flat dictionary, while denormalization reverses this. See the example below.

const input = [
  {
    likes: [
      { id: 1, name: 'Like 1' },
      { id: 2, name: 'Like 2' }
    ],
    company: { id: 1, name: 'Company 1', likes: [{ id: 1, name: 'Like 1' }] }
  },
  {
    id: 2,
    name: 'User 2',
    likes: [
      { id: 1, name: 'Like 1' },
      { id: 3, name: 'Like 3' }
    ]
  }
];

const output = {
  users: [
    { id: 1, name: 'User 1', likes: [1, 2], company: 1 },
    { id: 2, name: 'User 2', likes: [1, 2] }
  ],
  likes: [
    { id: 1, name: 'Like 1' },
    { id: 2, name: 'Like 2' },
    { id: 3, name: 'Like 3' }
  ],
  companies: [{ id: 1, name: 'Company 1', likes: [1] }]
};

To allow for (de-)normalization, a schema has to be created indicating how objects relate to each other. You can build these schemas using the schema(name, relations) helper from schematiq. The name parameter indicates in which dictionary of the output the objects will be stored. The relations parameter is an object representing those parts of the schema that indicates relationships to other objects.

import { schema, normalize, denormalize } from 'schematiq';

const likeSchema = schema('likes');
const companySchema = schema('companies', { likes: [likeSchema] });
const userSchema = schema('users', {
  likes: [likeSchema],
  company: companySchema
});

const output = normalize(input, userSchema);
const input = denormalize(output, userSchema);

Object helpers

In addition, some simple helpers around objects are also present.

• get(obj, path, defaultValue = undefined): retrieves a value in a nested object, defined by its path. Optionally, the default value can be provided;
• set(obj, path, value): sets a value in a nested object defined by its path. If the structure of the object does not suffice, this function will make it so;