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
markdown
filesh/bash/dash
etc)0
, fail on anything else# With npm
npm install @lando/leia
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
You can invoke leia
as a command line tool or directly require
it in a module.
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
# 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
In order for your markdown
file to be recognized as containing functional tests it needs to have at least the following
# Something to identify these tests
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.
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
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
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
.
leia
will autodetect your shell and use a bashy
one if available.
bash
or zsh
if available with a fallback to sh
.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.
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.
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.
We try to log all changes big and small in both THE CHANGELOG and the release notes.
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
# Lint the code
npm run lint
# Run unit tests
npm run test:unit
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.
Made with contributors-img.
You can still install the older version of Leia eg leia-parser
.
npm install leia-parser
And its documentation lives on here.