npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2024 – Pkg Stats / Ryan Hefner

okej

v0.0.9

Published

A bare-metal Result type for Javascript/TypeScript

Downloads

442

Readme

okej

okej a bare-metal implemention of Result types in JavaScript.

This library introduces some super simple helper functions so that data and errors can easily be passed throughout an application without the need to add try/catch statements everywhere.

See for yourself:

const result = doSomething(); // returns ok() or err()
if (result.ok) {
  // yes we have an Ok value. Let's examine the data
  console.log(result.data);
} else {
  // no looks like we have an Err value
  console.log(result.errMessage); // something went wrong
  console.log(result.errCode); // figure out what to do based on the code value
}

What are Result objects?

A Result object is simply an object that wraps itself around data as well as error information. Result objects exist in Rust, Elm, Kotlin, etc..., but not in JavaScript.

Here's an example of how they work:

import { err, ok } from "okej";

// Ok result (when something is valid)
const goodResult = ok("data");

// Err result (when something invalid or fails)
const badResult = err("something failed", "ERROR_500");

Now for any code that deals with a result object, it must deal with the result object and explicity deal with it's success and failure (ok/err) states.

// result is either `goodResult` or `badResult`
if (result.ok) {
  // valid state (continue along with the happy path)
} else {
  // invalid state (deal with the error)
}

Result objects are useful for the following reasons:

  • They force the developer to plan out each error and figure out how to handle them
  • Result objects can be returned up the chain to other functions
  • All Result objects have an error code as well as an error message
  • They clearly separate system/runtime errors from logic errors
  • try/catch statements can be used for truly unexpected situations (i.e. network errors or actual code issues)

Want more information? Click here for more

License

MIT.

Installation

First install it into your project.

# via NPM
npm install okej

# or Yarn
yarn add okej

# or PNPM
pnpm add okej

Then import it as needed:

// ESM / MJS
import { err, ok, Result } from "okej";

// CommonJS / old school Node
const { err, ok, Result } = require("okej");

Usage

Below is a example of how data is processed

import { err, ok, Result } from "okej";

// let's imagine this function is calling some remote operation
// and will return either a Ok or Err result once complete.
async function okOrErr(): Promise<Result> {
  return (await operationSucceeds())
    ? ok({ message: "yes!" })
    : err("didn't work out today", 500);
}

// we can check the data that is returned
const result = await okOrErr();
if (result.ok) {
  console.log(result.data); // { message: "yes!" }
} else {
  console.log(result.errMessage); // didn\'t work out today
  console.log(result.errCode); // 500
}

Docs

The main functions are ok() and err(), but there are also various helper functions that can be used as well.

ok()

When called, a Result object with ok=true/err=false will be returned.

import { ok } from "okej";

// with no data
let okResult = ok();
console.log(okResult.ok); // true
console.log(okResult.err); // false

// with data
okResult = ok({ id: "123", name: "foo" });
console.log(okResult.ok); // true
console.log(okResult.err); // false
console.log(okResult.data.id); // "123"
console.log(okResult.data.name); // "foo"

err()

When called, a Result object with ok=false/err=true will be returned.

import { err } from "okej";

// with no data
let result = err();
console.log(result.ok); // false
console.log(result.err); // true

// with data
result = err({
  errCode: 123,
  errMessage: "something broke",
  context: { customDataRelatedToError: "cool" },
});
console.log(result.ok); // false
console.log(result.err); // true
console.log(result.errCode); // 123
console.log(result.errMessage); // "something broke"
console.log(result.errContext); // { customDataRelatedToError: "cool" }

// by other means...
result = err(10); // result.errCode is 10
result = err("something broke"); // result.errMessage is "something broke"
result = err("something broke", 10); // result.errMessage/result.errCode set
result = err("something broke", "BAD ERROR"); // yes string-based error codes work too
result = err("something broke", 10, { extraData: true }); // result.errMessage/result.errCode/result.errContext set

We can also pass in a JavaScript Error instance directly:

import { err } from "okej";

try {
  //...
} catch (e) {
  result = err(e, { errCode: 10 });
  console.log(result.ok); // false
  console.log(result.errCode); // 10 (default is 0)
  console.log(result.errMessage); // (will be inferred from the `e` param)
  console.log(result.errContext); // null
  console.log(result.errException); // the provided `e` value
}

When TypeScirpt is used, err() can be used to deal with error code enums:

const enum HttpError {
  BadRequest = 400,
  NotFound = 404,
  ServerError = 500,
  Forbidden = 403,
}

interface ApiData {
  id: string;
  totalUsers: number;
  userNames: string[];
}

async function callApiUsersServer(
  url: string,
): Promise<Result<ApiData, HttpError>> {
  // call the api server and deal with the error
  try {
    const req = await fetch(url);
    if (response.status == 500)
      return err("server error", HttpError.ServerError);
    if (response.status == 403) return err("forbidden", HttpError.Forbidden);
    if (response.status == 404) return err("not found", HttpError.NotFound);

    const json = await req.json();
    if (!json || !isValidResponse(json)) {
      // this might be a 404 in the real world... This is just an example!
      return err("bad request", HttpError.BadRequest);
    }

    return ok(json);
  } catch (e) {
    return err("bad request", HttpError.BadRequest);
  }
}

const request = await callApiUsersServer();
if (request.err) {
  // deal with the error (TypeScript will know that `errCode` is an instance of HttpError)
  switch (request.errCode) {
    case HttpError.BadRequest:
      // handle the bad request error
      break;

    case HttpError.NotFound:
      // handle the not found error
      break;

    case HttpError.ServerError:
      // handle the not server error
      break;

    case HttpError.Forbidden:
      // handle the not forbidden error
      break;

    default:
      // just incase a new error was introduced later
      break;
  }
}

Helper Functions

The helper functions below can be used to deal with collections of result data.

import { allErr, allOk, isResult, merge } from "okej";

// true if all have ok=true/err=false (false if an empty array)
allOk(results);

// true if all have ok=false/err=true (false if an empty array)
allErr(results);

// true if some have ok=true/err=false
someOk(results);

// true if some have ok=false/err=true
someErr(results);

There also some functions that check to see if a result is indeed a result. Note that each of the functions listed in the code example below adhere to TypeScript type narrowing as well as assertions.

import { allErr, allOk, isResult, merge } from "okej";

// true if Ok or Err
isResult(someValue);

// true if Ok
isOkResult(someValue);

// true if Err
isErrResult(someValue);

// throws an error if not a Ok or Err value
assertIsResult(someValue);

// throws an error if not a Ok value
assertIsOkResult(someValue);

// throws an error if not a Err value
assertIsErrResult(someValue);

When dealing with a series of result values, the from function can be used to combine everything together.

import { from } from "okej";

const results = [ok(1), ok("yes"), ok({ data: "exists" })];

// a Ok value will be returned if all input values are also Ok
const result = from(results);
console.log(result.ok); // true
console.log(result.err); // true
console.log(result.data); // [ 1 , "yes" , { data: "exists" } ]

However if there are any values that are an instance of Err then the from function will return an Err response.

import { from } from "okej";

const results = [ok(1), err("noooo"), err(100)];

// a Ok value will be returned if all input values are also Ok
const result = from(results);
console.log(result.ok); // false
console.log(result.err); // false
console.log(result.errContext.results); // [ ok(1), err("noooo"), err(100) ]

More on JavaScript's error handling issues

As mentioned at the start of the README, error handling in JavaScript is not so great.

Why can't we use try/catch everywhere

Try/catch does the job in JavaScript to capture and deal with errors, but the actual integrity of Error objects is a complete mess in the language.

Let's look at some custom form validation:

function validateForm(formData: { [key: string]: unknown }): boolean {
  if (!formData.username) {
    throw new Error("you didn't enter a username");
  }
  if (!formData.email) {
    throw new Error("you didn't enter an email");
  }

  // yes it looks good
  return true;
}

Now when the validateForm data is called, how does that look?

const myFormData = { ... }; // collected from some form
let isValid = false;
try {
  isValid = validateForm(myFormData);
} catch (e) {
  // `e` could be an error or even a string. How do we figure out what the error actually is?
  if (e instanceof Error) {
    if (/username/i.test(e.message)) {
      // show the error in the form for the username
      // what if there are various errors here?
    } else if (/email/i.test(e.message)) {
      // show the error in the form for the email
    } else {
      // what now?
    }
  } else {
    // oh crap... what now?
  }
}

if (isValid) {
  // finally.
} else {
  // wait there's another error case here too?
}

As made clear in the example code above, how does one deal with the error programmatically? Well using a Result object we can clean this up quite a lot.

import {ok, err, Result} from "okej";

interface FormError {
  EmptyUsername,
  BadUsername,
  EmptyEmail,
  BadEmail,
}

function validateForm(formData: {[key: string]: unknown): Result {
  if (!formData.username) {
    return err("you didn't enter a username", FormError.EmptyUsername);
  }

  if (!formData.email) {
    return err("you didn't enter an email", FormError.EmptyEmail);
  }

  return ok();
}

And dealing with this error is just a matter of looking at the result response

const myFormData = { ... }; // collected from some form
const validationResponse = validateForm(myFormData);
if (validationResponse.ok) {
  // cool it works
} else {
  switch (validationResponse.errCode) {
    case FormError.EmptyUsername:
      // deal with the missing username
      break;

    case FormError.EmptyEmail:
      // deal with the missing email
      break;

    default:
      // catch all
      break;
  }
}

Why does this library exist?

The ts-results library is the only other library out there in JS land that does something similar. This library does the job, however it closely follow's Rust's Result implementation which means that it brings in a lot of extra features that might not be necessary (i.e. Option values, Some/None types, Rx and mapping support as well as unwrapping). These features are essential in Rust, but they do introduce complexity and lead JS code into a direction that isn't common.

This library is just a small utility library that aims to stay lean and below the 2kb mark.

There is also the NPM library called results, but the README says to consider other tools.

It works with HTTP

Another good reason for this library is because it works with HTTP requests out of the box since no special function/object code is added around result objects.

// somewhere in the backend
import express from "express";
import { err, ok } from "okej";

const app = express();
app.get("/users", (req, res) => {
  res.send(ok([{ username: "user123" }, { username: "user456" }]));
});

app.post("/users", (req, res) => {
  // Ok or Err depending on if the user was created
  const result = createUser(req.body.user);
  res.send(result);
});

And the front-end can just consume the Ok and Err Result values.