typanion

What is typanion?

The typanion npm package is a comprehensive solution for runtime input validation in JavaScript and TypeScript applications. It allows developers to ensure that the data their applications process meets certain criteria, thereby preventing unexpected errors and improving data integrity. Typanion offers a flexible and expressive API for defining validation rules, making it suitable for a wide range of use cases.

What are typanion's main functionalities?

Basic Type Validation

This feature allows for basic type validation, such as checking if a value is a number. The code sample demonstrates how to use the `isNumber` function to validate numeric values.

{"const {isNumber} = require('typanion');
const validate = isNumber();
console.log(validate(42)); // {ok: true, errors: []}
console.log(validate('hello')); // {ok: false, errors: [...]}"}

Complex Object Validation

This feature enables complex object validation, allowing developers to validate nested objects and apply multiple validation rules. The code sample shows how to validate a user object with both `name` and `email` fields.

{"const {isObject, isString, applyCascade} = require('typanion');
const validateUser = isObject({
  name: isString(),
  email: applyCascade(isString(), [isEmail()])
});
console.log(validateUser({name: 'John Doe', email: '[e-mail]'})); // {ok: true, errors: []}"}

Custom Validation Rules

Typanion allows for the creation of custom validation rules. This feature is particularly useful for domain-specific validations. The code sample illustrates how to define and use a custom validator to check if a person is an adult.

{"const {createValidator} = require('typanion');
const isAdult = createValidator({
  test: (value) => value >= 18,
  message: (value) => `${value} is not an adult`
});
console.log(isAdult(21)); // {ok: true, errors: []}
console.log(isAdult(16)); // {ok: false, errors: [...]}"}

Other packages similar to typanion

joi

Joi is a powerful schema description language and data validator for JavaScript. It offers a similar range of functionalities for validating data structures but with a slightly different API design. Compared to typanion, Joi might be considered more feature-rich but also more complex for simple use cases.

yup

Yup is a lean JavaScript schema builder for value parsing and validation. It integrates well with form libraries and is often used in front-end applications. While it provides similar validation capabilities, Yup focuses more on simplicity and ease of use compared to the more flexible and comprehensive approach of typanion.

validator

Validator is a library of string validators and sanitizers. Unlike typanion, which offers a wide range of validation types and custom validation logic, Validator focuses primarily on string validation, making it more specialized but less versatile for different data types.

README

Typanion

Static and runtime type assertion library with no dependencies

Installation

yarn add typanion

Why

• Typanion can validate nested arbitrary data structures
• Typanion is type-safe; it uses type predicates
• Typanion allows you to derive types from your schemas
• Typanion can report detailed error reports

Compared to yup, Typanion has a better inference support for TypeScript + supports isOneOf. Its functional API makes it very easy to tree shake, which is another bonus (although the library isn't very large in itself).

Usage

First define a schema using the builtin operators:

import * as t from 'typanion';

const isMovie = t.isObject({
    title: t.isString(),
    description: t.isString(),
});

Then just call the schema to validate any unknown value:

const userData = JSON.parse(input);

if (isMovie(userData)) {
    console.log(userData.title);
}

Passing a second parameter allows you to retrieve detailed errors:

const userData = JSON.parse(input);
const errors: string[] = [];

if (!isMovie(userData, {errors})) {
    console.log(errors);
}

You can also apply coercion over the user input:

const userData = JSON.parse(input);
const coercions: Coercion[] = [];

if (isMovie(userData, {coercions})) {
    // Coercions aren't flushed by default
    for (const [p, op] of coercions) op();

    // All relevant fields have now been coerced
    // ...
}

You can derive the type from the schema and use it in other functions:

import * as t from 'typanion';

const isMovie = t.isObject({
    title: t.isString(),
    description: t.isString(),
});

type Movie = t.InferType<typeof isMovie>;

// Then just use your alias:
const printMovie = (movie: Movie) => {
    // ...
};

Schemas can be stored in multiple variables if needed:

import * as t from 'typanion';

const isActor = t.isObject({
    name: t.isString();
});

const isMovie = t.isObject({
    title: t.isString(),
    description: t.isString(),
    actors: t.isArray(isActor),
});

API

Type predicates

• isArray(values) will ensure that the values are arrays whose values all match the specified schema.

• isBoolean() will ensure that the values are all booleans. Prefer isLiteral if you wish to specifically check for one of true or false. This predicate supports coercion.

• isDate() will ensure that the values are proper Date instances. This predicate supports coercion via either ISO8601, or raw numbers (in which case they're interpreted as the number of seconds since epoch, not milliseconds).

• isDict(values, {keys?}) will ensure that the values are all a standard JavaScript objects containing an arbitrary number of fields whose values all match the given schema. The keys option can be used to apply a schema on the keys as well (this will always have to be strings, so you'll likely want to use applyCascade(isString(), [...]) to define the pattern).

• isLiteral(value) will ensure that the values are strictly equal to the specified expected value. It's an handy tool that you can combine with oneOf and object to parse structures similar to Redux actions, etc.

• isNumber() will ensure that the values are all numbers. This predicate supports coercion.

• isObject(props, {extra?}) will ensure that the values are plain old objects whose properties match the given shape. Extraneous properties will be aggregated and validated against the optional extra schema.

• isString() will ensure that the values are all regular strings.

• isUnknown() will accept whatever is the input without validating it, but without refining the type inference either. Note that by default isUnknown will forbid undefined and null, but this can be switched off by explicitly allowing them via isOptional and isNullable.

• isInstanceOf(constructor) will ensure that the values are instances of a given constructor.

Helper predicates

• applyCascade(spec, [specA, specB, ...]) will ensure that the values all match spec and, if they do, run the followup validations as well. Since those followups will not contribute to the inference (only the lead schema will), you'll typically want to put here anything that's a logical validation, rather than a typed one (cf the Cascading Predicates section).

• isOneOf([specA, specB]) will ensure that the values all match any of the provided schema. As a result, the inferred type is the union of all candidates. The predicate supports an option, exclusive, which ensures that only one variant matches.

• isOptional(spec) will add undefined as an allowed value for the given specification.

• isNullable(spec) will add null as an allowed value for the given specification.

Cascading predicates

Cascading predicate don't contribute to refining the value type, but are handful to assert that the value itself follows a particular pattern. You would compose them using applyCascade (cf the Examples section).

• hasExactLength will ensure that the values all have a length property exactly equal to the specified value.

• hasMaxLength will ensure that the values all have a length property at most equal to the specified value.

• hasMinLength will ensure that the values all have a length property at least equal to the specified value.

• hasUniqueItems will ensure that the values only have unique items (map will transform before comparing).

• isAtLeast will ensure that the values compare positively with the specified value.

• isAtMost will ensure that the values compare positively with the specified value.

• isBase64 will ensure that the values are valid base 64 data.

• isHexColor will ensure that the values are hexadecimal colors (alpha will allow an additional channel).

• isInExclusiveRange will ensure that the values compare positively with the specified value.

• isInInclusiveRange will ensure that the values compare positively with the specified value.

• isInteger will ensure that the values are round safe integers (unsafe will allow unsafe ones).

• isJSON will ensure that the values are valid JSON, and optionally match them against a nested schema.

• isLowerCase will ensure that the values only contain lowercase characters.

• isNegative will ensure that the values are at most 0.

• isPositive will ensure that the values are at least 0.

• isISO8601 will ensure that the values are dates following the ISO 8601 standard.

• isUpperCase will ensure that the values only contain uppercase characters.

• isUUID4 will ensure that the values are valid UUID 4 strings.

• matchesRegExp will ensure that the values all match the given regular expression.

Examples

Validate that an unknown value is a port protocol:

const isPort = t.applyCascade(t.isNumber(), [
    t.isInteger(),
    t.isInInclusiveRange(1, 65535),
]);

isPort(42000);

Validate that a value contains a specific few fields, regardless of the others:

const isDiv = t.isObject({
    tagName: t.literal(`DIV`),
}, {
    extra: t.isUnknown(),
});

isDiv({tagName: `div`, appendChild: () => {}});

Validate that a specific field is a specific value, and that others are all numbers:

const isModel = t.isObject({
    uid: t.String(),
}, {
    extra: t.isDict(t.isNumber()),
});

isModel({uid: `foo`, valA: 12, valB: 24});

License (MIT)

Copyright © 2020 Mael Nison

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.