Skip to content

flix-tech/fp-ts-type-check

master
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
src
 
 
 
 
 
 
 
 
 
 
 
 
 
 

fp-ts-type-check is a library for runtime type validation of variables where you can't say their type for sure i.e. data you get via some API. It's somewhat similar to Typescript's Type Guards except it uses type system to make sure this data is properly validated.

Features:

  • Type safe. Typescript compiler makes sure your types and your type checkers are always in sync.
  • Detailed error reports. If parsing of a big deep-nested structure failed - you'll know where exactly in that structure you have invalid data and why it was considered invalid
  • Composable. Type checkers for big structures are created by composing checkers for simple structures.
  • Functional. Each type checker is a pure function returning Either type from fp-ts. It's up for you to decide how to handle errors.
  • Tree-shakeable. Bundle only functions you really use.

Installation

To install the stable version:

npm install '@flix-tech/fp-ts-type-check@~0.2.0'

While the major version number is 0 changes in minor version number bay break backward compatibility so you should stick to a fixed minor version.

Usage

Meet Parser[T], the main type of this library. Parser[T] is a function that takes any variable and checks if it's a variable of type T. It either returns the same value as T or ParseError object with parse error details. Parser's type definition looks somewhat like this:

type Parser<A> = (x: unknown): Either<ParseError, A>;

interface ParseError {
  path: string; // Path to the property causing error in deep nested structures
  message: string; // Parsing error message
}

We're using the type Either from fp-ts library.

Parsers for some complex structures you'd want to validate are composed from small parsers like here:

import * as P from 'fp-ts-type-check';

interface ShoppingListItem {
  name: string;
  amount: { count: number; unit?: string };
}
const shoppingListItemParser: P.Parser<ShoppingListItem> = P.type({
  name: P.string,
  amount: P.type({
    count: P.number,
    unit: P.optional(P.string),
  }),
});

shoppingListItem({name: "Apple", amount: {count: 5}}); // Right<ShoppingListItem>({name: "Apple", amount: {count: 5}})

shoppingListItem({name: "Apple", amount: {count: "some"}}); // Left<ParseError>({path: ".amount.count", message: "expected number, got string"})

As you can see, ParseError data is for your eyes only, user should get a generalized error message appropriate in this case.

You can reuse your parsers to construct parsers for even bigger structures:

type ShoppingList = array<ShoppingListItem>;

const shoppingListParser: P.Parser<ShoppingList> = P.arrayOf(shoppingListItemParser);

const validShoppingList = [
  {name: "Apple", amount: {count: 5}},
  {name: "Milk", amount: {count: 500, unit: 'ml'}},
];
shoppingListParser(validShoppingList); // Right<ShoppingList>(...)

const invalidShoppingList = [
  {name: "Apple", amount: {count: 5}},
  {name: "Milk", amount: {unit: 'ml'}}, // No count here
];
shoppingListParser(invalidShoppingList); // Left<ParseError>({path: "[1].amount.count", message: "expected number, got undefined"})

API documentation

Simple type parsers

  • string(): Parser<string> - checks that value is string.
  • boolean(): Parser<boolean> - checks that value is boolean.
  • number(): Parser<number> - checks that value is number.
  • object(): Parser<object> - checks that value is any object.
  • any(): Parser<any> - does not check anything.
  • exact<A>(expected: A): Parser<A> - checks that value is same as expected one.
  • oneOf<A>(allowed: A[]): Parser<A> - checks that value is one of allowed ones.
  • keyOf<A extends object>(allowed: A): Parser<keyof A> - checks that value is string and also is a key of object allowed.

Copositional parsers

  • type<A>(propertyParsers: { [K in keyof A]: Parser<A[K]> }): Parser<A> - checks that values is an object and validates each it's property with corresponding parser.

  • arrayOf<A>(parseBody: Parser<A>): Parser<A[]> - checks that value is an array and every item matches parseBody parser.

  • optional<A>(parseBody: Parser<A>): Parser<A | undefined> - checks that value is either undefined or matches parseBody parser.

  • nullable<A>(parseBody: Parser<A>): Parser<A | null> - checks that value is either null or matches parseBody parser.

  • discriminatedUnion<A extends {type: string;}>(parsers: {(type value): (type parser)}): Parser<A> - checks that value is a discriminated union with discriminant in type property. Example usage:

    type Foo = { type: 'foo'; foo: number };
    type Bar = { type: 'bar'; bar: number };
    type FooBar = Foo | Bar;
    
    const parser: P.Parser<FooBar> = P.discriminatedUnion({
      foo: P.type({ foo: P.number }),
      bar: P.type({ bar: P.number }),
    });

Logic combination parsers

  • and(parserA: Parser<A>, parserB<B>): Parser<A & B> - checks that value matches both types A and B.
  • or(parserA: Parser<A>, parserB<B>): Parser<A | B> - checks that value matches any of types A or B.

License

The MIT License (MIT)

About

runtime type validation library for Typescript

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published