Leia

Leia is a testing utility that tests code blocks in documentation. This makes tests easy to write and also ensures documentation is up to date and working. Behind the scenes documentation is parsed and run as a series of mocha tests.

Leia will

• Consolidate code examples and tests into a single, easy to understand and write markdown file
• Write functional tests quickly in an accessible and lowest common denominator language (eg sh/bash/dash etc)
• Pass on exit status code 0, fail on anything else
• Work cross platform-ish, with some caveats, see Shell Considerations below
• Keep Lando honest so he can be a real hero who doesn't betray his friends again

Installation

# With npm
npm install @lando/leia

Basics

A very basic example of a valid Leia test is below. It must have a single H1 header, at least one H2 header and then a code block where the comment is the human readable test description and the command below is the test.

# Some Example

## Testing

# A description of my test
the command i am running

Usage

You can invoke leia as a command line tool or directly require it in a module.

CLI

npx leia

Cleverly converts markdown files into mocha cli tests

USAGE
  $ leia <files> <patterns> [--cleanup-header=<cleanup-headers>] [--debug] [--help] [--ignore=<patterns>] [--retry=<count>] [--setup-header=<setup-headers>] [--test-header=<test-headers>]
  [--shell=<bash|cmd|powershell|pwsh|sh|zsh>] [--stdin] [--timeout=<seconds>] [--version]

ARGUMENTS
  TESTS  files or patterns to scan for test

OPTIONS
  -c, --cleanup-header=cleanup-header      [default: Clean,Tear,Burn] considers these h2 sections as cleanup commands
  -i, --ignore=ignore                      files or patterns to ignore
  -r, --retry=retry                        [default: 1] retries tests the given amount
  -s, --setup-header=setup-header          [default: Start,Setup,This is the dawning] considers these h2 sections as setup commands
  -t, --test-header=test-header            [default: Test,Validat,Verif] considers these h2 sections as tests
  -v, --version                            shows version info
  --debug                                  shows debug output
  --help                                   shows help
  --shell=bash|cmd|powershell|pwsh|sh|zsh  [default: /opt/homebrew/bin/zsh] runs tests with given shell, autodetected by default
  --stdin                                  attachs stdin when the test is run
  --timeout=timeout                        [default: 1800] seconds before tests time out

EXAMPLES
  leia README.md
  leia README.md "examples/**/*.md" --retry 6 --test-header Tizzestin
  leia "examples/*.md" --ignore BUTNOTYOU.md test --stdin --timeout 5
  leia README.md --shell cmd

Module

# Instantiate a new leia
const Leia = require('@lando/leia');
const leia = new Leia();

// Find some tests
const files = leia.find(['examples/**.md']);
// Parse those files into leia test metadata
const sources = leia.parse(files);
// Generate the mocha tests
const tests = leia.generate(sources);
// Run the tests
const runner = leia.run(tests);
runner.run((failures) => process.exitCode = failures ? 1 : 0);

For more details on specific options check out the code docs

• leia.find
• leia.generate
• leia.parse
• leia.run

Markdown Syntax

In order for your markdown file to be recognized as containing functional tests it needs to have at least the following

1. A h1 Header

# Something to identify these tests

2. A h2 Header

By default our parser will look for a section that beings with the word "Testing". This section will contain your tests.

## Testing

You can customize the word(s) that leia will look for to identify the testing section(s) using the --test-header option. You can also run npm leia --help to get a list of default words.

3. A code block with at least one command and comment

Under the above h2 sections you need to have a triple tick markdown code block that contains at least one comment and one command. The comment will be the human readable description of what the test does.

Here is a basic code block that runs one test

# Should cat a file
cat test.txt

If you want to learn more about the syntax and how leia puts together the above, check out this example

Skipping

You can also skip tests. This is useful if you want to stub out a test for later.

# Should write this test later and dont want to forget it
skip

Environment Variables

leia will also set the following environment variables for each test that is running so you can use them for stuff.

Here are the values you would expect for the Should set envvars with the test number test in examples/environment.md running on Leia version v1.0.0 with --retry=1.

# generic vars
LEIA=true
LEIA_ENVIRONMENT=true
LEIA_VERSION=1.0.0

# test vars
LEIA_TEST_RUNNING=true
LEIA_TEST_ID=environment
LEIA_TEST_NUMBER=4
LEIA_TEST_RETRY=1
LEIA_TEST_STAGE=test

Note: LEIA_TEST_STAGE can be either setup, test or cleanup and LEIA_TEST_NUMBER resets to 1 for each LEIA_TEST_STAGE.

Shell considerations

leia will autodetect your shell and use a bashy one if available.

• On POSIX systems it will prefer bash or zsh if available with a fallback to sh.
• On Windows systems it will prefer bash if available with a fallback to cmd.

You can also explicitly tell leia what shell to use with the --shell option. However, currently only bash, sh, zsh, cmd, powershell and pwsh are supported options.

In most use cases it's best to just let leia decide the shell to use automatically.

Advanced Usage

Leia also allows you to specify additional h2 sections in your markdown for setup and cleanup commands that run before and after your core tests. You can tell leia what words these headers should start with in order to be flagged as setup and cleanup commands using the --setup-header and --cleanup-header options.

Here is an example of a markdown file with Setup, Testing and Cleanup sections. And here is a whole directory of examples that we test on every commit.

Issues, Questions and Support

If you have a question or would like some community support we recommend you join us on Slack. Note that this is the Slack community for Lando but we are more than happy to help with this module as well!

If you'd like to report a bug or submit a feature request then please use the issue queue in this repo.

Changelog

We try to log all changes big and small in both THE CHANGELOG and the release notes.

Development

• Requires Node 18+

git clone https://github.com/lando/leia.git && cd leia
npm install

If you dont' want to install Node 18 for whatever reason you can install Lando and use that:

git clone https://github.com/lando/leia.git && cd leia
# Install deps and get node
lando start

# Run commands
lando node
lando npm install
lando npx leia

Testing

# Lint the code
npm run lint

# Run unit tests
npm run test:unit

Releasing

To deploy and publish a new version of the package to the npm registry you need only create a release on GitHub with a semver tag.

Note that prereleases will get pushed to the edge tag on the npm registry.

Maintainers

• @pirog
• @reynoldsalec

Contributors

Made with contributors-img.

Legacy Version

You can still install the older version of Leia eg leia-parser.

npm install leia-parser

And its documentation lives on here.

Other Resources

• Mountain climbing advice