validate

Performs validation and normalization on JavaScript objects, through schemata also defined as JavaScript objects, using a MongoDB-query-like structure.

Install with: npm install @hlorenzi/validate

There's a battery of tests at npm test, and test coverage checks at npm run test-coverage.

Example

import { Validator } from "@hlorenzi/validate"

const schema =
{
    name:   { $type: "string", $maxLen: 30 },
    age:    { $type: "int", $min: 0 },
    height: { $type: "float" },
    
    address:
    {
        number: { $type: "int" },
        street: { $type: "string" }
    },
    
    info: { $optional: true, $type: "string" }
}

const object =
{
    name: "Bob",
    age: 25,
    height: 1.79,
    address:
    {
        number: 10,
        street: "One Street"
    }
}

const normalizedObject = Validator.validateOrThrow(schema, object)

Validator API

Validator.validateOrThrow(schema, object, options = {})
Validator.validateOrNull(schema, object, options = {})

These are the main entry points for validation. They all return a normalized object containing only the relevant fields:

• undefined or null values, where accepted, are removed from the object.
• By using { discardUnknownFields: true } for options, extraneous fields not defined in the schema will also be simply removed, instead of being considered an error.

Schema Structure

{
    field1: { $type: "..." },
    field2: { $type: "..." },
    field3:
    {
        nestedField1: { $type: "..." },
        nestedField2: { $type: "..." },
    }
}

Fields not starting with a $ are treated as being expected to exist in the object under validation.

The fields on the object have their values validated by using the directives contained in the schema's corresponding fields, such as the { $type: "..." } declarations above.

A basic directive can only specify the type, as in { $type: "string" }, but depending on the type, they can also have extra constraints. These are independent from one another, so you can specify only the ones that are relevant:

{ $type: "string",
    $minLen: 0,
    $maxLen: 30,
    $in: [ "apple", "orange", "banana" ] }
    
{ $type: "int",
    $min: 0,
    $max: 10,
    $minExclusive: 0,
    $maxExclusive: 10 }
    
{ $type: "float",
    $min: 0,
    $max: 10,
    $minExclusive: 0,
    $maxExclusive: 10,
    $acceptInfinity: true,
    $acceptNaN: true }
    
{ $type: "bool" }

{ $type: "array", $of: { /* nested directive */ },
    $minLen: 1,
    $maxLen: 5 }

Fields can also be declared with the $optional directive, in which case the validator will accept missing, undefined, and null values. For example:

{
    field1: { $type: "string", $optional: true }
}

For multiple possible types, you can use the $either directive:

{
    field1:
    {
        $either:
        [
            { $type: "string", $maxLen: 30 },
            { $type: "float", $max: 10 },
            { $type: "float", $min: 20 }
        ]
    }
}

Multiple errors can be detected in a single validation pass, and the validation functions will provide an array of errors upon unsuccessful validation.

• validateOrThrow(schema, object, options = {}) will return the normalized object or throw the array of errors.
• validateOrNull(schema, object, options = {}) will return the normalized object or return null on error, never throwing.

When the validation throws, you can check whether the error is coming from an actual validation error against the schema, as opposed to some other unexpected exception, by using:

• isValidatorErrorArray(err) will return true if the given exception is an array of validation errors.

Each object in the array of validation errors contains two fields:

• path is a string describing the name of the field that is invalid, possibly using dot syntax for nested fields, such as "address.street".
• failure is a string describing the internal reason for the field's value being rejected.