npm i @hazae41/phobos

Node Package 📦

Philosophy 🧠

Phobos aims to be minimalist and to always work no matter the:

• runtime (Node, Deno, browser)
• module resolution (ESM, CommonJS)
• language (TypeScript, JavaScript)
• bundler (Rollup, Vite)

It's just a library you can import everywhere! That's it, no CLI, no configuration file, just JavaScript.

Features 🔥

Current features

• 100% TypeScript and ESM
• No external dependency
• Unit tested (by itself)
• Runnable in the browser
• Minimalist assertion helpers
• Asynchronous fork-join parallelism
• Function calls spying

Upcoming features

• Mocks
• Diffing

Usage 🚀

import { assert, test } from "@hazae41/phobos"

test("it should work", async () => {
  assert(false, "oh no")
})

ts-node --esm ./test.ts

Concurrent tests

Test blocks are always executed concurrently, unless you await them

import { assert, test } from "@hazae41/phobos"

test("it should work", async ({ test }) => {
  
  // run in sequence
  await test("first test", async () => {
    assert(true, "should be true")
  })
  
  // or in parallel
  test("second test", async () => {
    assert(true, "should be true")
  })
})

You can also use await wait() to forcefully join

import { assert, test } from "@hazae41/phobos"

test("it should work", async ({ test, wait }) => {
  
  test("first test", async () => {
    assert(true, "should be true")
  })
  
  test("second test", async () => {
    assert(true, "should be true")
  })

  // wait first and second tests
  await wait()

  test("third test", async () => {
    assert(true, "should be true")
  })
})

Spying function calls

You can spy on function calls using spy(function)

You can then .call() it and get a list of all its .calls

import { assert, test, spy } from "@hazae41/phobos"

test("it should work", async () => {
  const f = spy((param: boolean) => !param)

  const result = f.call(true)
  assert(result === false, `result should be false`)

  assert(f.calls.length === 1, `should have been called 1 time`)
  assert(f.calls[0].params[0] === true, `should have been called with true`)
  assert(f.calls[0].result === false, `should have resulted in false`)
})

Setting up 🔧

Most setups will just need a custom entry point that imports all your tests, that you either run as-is using ts-node, or that you transpile using your favorite bundler.

For example, the entry point index.test.ts imports:

• some-module/index.test.ts, which imports:
  • some-module/some-file.test.ts
  • some-module/some-other-file.test.ts
• some-other-module/index.test.ts, which imports:
  • some-other-module/some-file.test.ts
  • some-other-module/some-other-file.test.ts

You can see an example on this repository, all tests are imported in src/index.test.ts, then we use Rollup to transpile it into dist/test/index.test.cjs, which we then run using Node with node ./dist/test/index.test.cjs.

Running 🏎️

Using a bundler

node ./dist/test/index.test.cjs

Using ts-node with ESM

ts-node --esm ./src/index.test.ts

Using ts-node with ESM and ttypescript

ts-node --esm --compiler ttypescript ./src/index.test.ts

Using dynamic import

await import("index.test.ts")