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

zora-dev-server

v0.0.5

Published

test runner for browser programs, using zora testing library

Downloads

24

Readme

zora-dev-server

A dev server which serves your project files and run your tests in any modern browser without any configuration or extra build step.

Motivation

One of the most spread and targeted Javascript environment is the browser. Javascript being an interpreted language, you don't need to compile your source code before the browser can run it. This is very handy to develop and test your program at a high pace.

However, in practice most of the programs which target the browser's environment do have compilation steps, at least for their production release:

  1. you may want to bundle different structured source files together
  2. you will probably optimize your code with minifiers and other tools

These steps are usually not mandatory to run tests on your source code and iterate quickly, yet there are very few tools at your reach to test javascript program directly in a browser in an efficient way. Most of the times, in order to use these tools you will have to pre process your source code files which brings yet a different set of tools or plugins to configure (and the troubles that go with), while slowing down the whole development cycle unnecessarily.

Why is that so ?

Browser and Node come with different interfaces on top of Javascript. They are sometimes incompatible: a browser can not have access to the Operating System low level APIs such the file system; while at the opposite Node does not come with a Document Object Model (DOM). So if you want to run with Node a program which references DOM interfaces, you'll need to mock them or implement them in the Node world (that's what jsdom do for instance), however it will remain different than running your program in a real browser and you might face some unwanted discrepancies. On the other side you might have code that can technically run in both environments but because other part the program references Node's native modules, it becomes incompatible with a browser environment. To tackle this issue you have to statically analyze your code and replace or remove references to Node's modules.

I believe the aforementioned problems are not very common (or easily workaround) because most of the time you'll develop a part of a program either for the browser either for Node and won't need to test that part in both environments. For the parts which are meant to be platform agnostic, then you should not bump into these problems as they will likely contain only EcmaScript code.

But there is another major pain point: the way modules are defined, packaged and distributed through the registry used in most of the cases together with Node (NPM). A browser has no clue of the Commonjs module format defined by Node and does not have any require method. While Node has the ability to understand and use EcmaScript module format, and even if people now are tending to distribute their libraries in both module formats (CJS and ESM), you still need to be careful when writing an import statement. The browser only understands a module definition through URLs while Node will be able to resolve packages by name. You could workaround this problem by taking the habit of always requiring module by their path (with extension!) even with your dependencies in the node_modules. But in practice, this is not doable: you would need to go through every dependency to find where an eventual ESM module file is and hope every package you use is written the same way. Therefore it is way easier to stick to the package.json contract and write code such

import {test} from 'zora';
// rather than
import {test} from 'node_modules/zora/dist/bundle/module.js';

then transforming that code with some tool.

The idea of this package is to let you write source code in the convenient way while serving file in a way browsers can understand and run your tests without any compilation or extra step. This should allow you to iterate quickly on your development without extra set up and configuration cost.

Installation

npm install --save-dev zora-dev-server or globally npm install -g zora-dev-server

Usage

// see the video for a tour of what you can do

Write your test files with a default export function taking a zora's assertion object as argument.

Note: interestingly, this convention allows you to run your test files with pta test runner(for Node environment) as well.

// test/answer.js
export default t => {
    t.test('some test', t=>{
        t.eq(40+2,42, 'this answer');
    });
}

Now if you want to run your test in the browser, all you have to do is to start a zora dev server and visit http://localhost:3000/test/answer.test with the browser you want to use.

Note: Your browser must be modern and understand module scripts (it is the case for every evergreen browser: Firefox, Chrome, Edge, Opera, Safari, etc)

See the reporters section to understand how to analyze and troubleshoot your tests the most efficiently.

CLI

zds will start the server rooted to the current working directory

configuration

The defaults will allow you to work with no configuration in most of the cases. But if for some reason you wish to change the default settings you can pass options to the cli or with a config file

port (--port)

Change the port of the server (default is 3000)

config (--config, -c)

Point to a file whose properties are any option you can pass to the CLI plus

dependenciesMap (through config file only)

This will overwrite the way the dev server resolve a given package.

For example the file server will resolve the package zora to the URL /node_modules/zora/dist/bundle/module.js. If for some reason you want to change it or if the file server fails to resolve a given package to a valid ESM module file, you can tell it to use a specific URL

{
  "port": 4000,
  "dependenciesMap": {
    "package_a": "/path/to/es/module.js", // in that case the root "/" match the current working directory 
    "package_b": "/path/to/other/es/module.js", 
    "zora": "https://unpkg.com/zora@different_version/dist/bundle/module.js"
  }
}

Endpoints

file system

Once the dev server is started it mirrors your file system. You can for example visit the default origin url (http://localhost:/) and it will match the current working directory. You should then be able to navigate through your project.

Often you can pass parameters or options to a specific endpoint thanks to query parameters.

source files

If you visit a source file (.js) you will see the source code. However while serving source files, zora-dev-server may transform on the fly import statements to point to an actual source file so the browser won't see a package's name but an actual URL. You can see the raw file if you wish by appending ?format=raw to the URL

test files

If a source file is actually a test file you want to run, you can replace the extension name (.js) by .test.

For example if you have a test file at cwd/test/my_test.js which respects the default export convention you can run those tests in the browser by visiting the url http://localhost:<port>/test/my_test.test

test runners

If you wish to run multiple test files at the same time (matching glob patterns), you can use a special endpoint.

// todo isolation with iframe if necessary

tip: zora-dev-server comes with a test runner which aims at simplifying your experience but zora alone is also pretty easy to use for browser testing.

Just drop in an html file

<script type="module">
import {test} from 'path/to/zora.js'; // or simply 'zora' if you serve the file with zora-dev-server
import spec from 'path/to/file';

test(`my test`, spec);
</script>
<script type="module">
import {test} from 'path/to/zora.js'; // or simply 'zora' if you serve the file with zora-dev-server
import spec from 'path/to/other/file';

test(`my ohter test`, spec);
</script>
<!--etc-->

And that's it!

And if you don't use the convention of the default export and import zora directly in your test file, it's even easier: you just need to import the test file nothing more:

// test.js
import {test} from 'path/to/zora.js' // or simply 'zora' if you serve the file with zora-dev-server

test('some test', t=> {
    t.eq('foo','bar')
});
<script type="module" src="test.js"/>

reporters

By default you will see a micro app with a summary of the tests (that is the summary-app reporter). But you will have in the console more details, stack traces, etc to troubleshoot your failing tests (that is the console reporter). If for some reason you wish to change the reporter, you can use the query parameters reporter. You can have at the same time two reporters. This may be useful if you want to have something in the browser and pass info at the same time with the devTool protocol (through console calls) or a network protocol (websocket etc).

example: http://localhost:<port>/some/test.test?reporter=tap&reporter=summary-app

summary-app

// todo

console

// todo

tap

// todo

tap-indent

// todo

Log

By default zora-dev-server is silent. But it uses the debug package. So if you wish to have more info on entering requests or thrown errors you can start zora dev server with the DEBUG env variable set to zora-dev-server:*.

Caching

The browser comes with a variety of caches. zora-dev-server tries to take advantage of it to be the fastest as possible. You should probably not deactivate the caches while developing, if a file is changed it will be detected. However, once you want to swap to another project we recommend to purge the caches especially if you use the same port.

As part of CI

// todo puppeteer

// idea web socket (like karma ?)

Security

The server is meant for development purpose and is not in any way a secure server you would want to expose to the network.

icons

The icons used are from the awesome icomoon application