Interested in contributing? Great! Here are some suggestions to make it a good experience:
Start by opening an issue, whether to identify a problem or outline a change. That issue should be used to discuss the situation and agree on a plan of action before writing code or sending a pull request. Maybe the problem isn't really a problem, or maybe there are more things to consider. If so, it's best to realize that before spending time and effort writing code that may not get used.
Match the coding style of the files you edit. Although everyone has their own preferences and opinions, a pull request is not the right forum to debate them.
Do not add new dependencies
to package.json
. The Markdown
parsers markdown-it
and micromark
are the
project's only dependencies.
If developing a new rule, start by creating a custom rule in its
own project. Once written, published, and tested in real world scenarios, open
an issue to consider adding it to this project. For rule ideas, see issues
tagged with the new rule
label.
Add tests for all new/changed functionality. Test positive and negative scenarios. Try to break the new code now, or else it will get broken later.
Run tests before sending a pull request via npm test
in the usual
manner. Tests should all pass on all platforms. The test runner is
AVA and test cases are located in test/markdownlint-test*.js
. When
running tests, test/*.md
files are enumerated, linted, and fail if any
violations are missing a corresponding {MD###}
marker in the test file. For
example, the line ### Heading {MD001}
is expected to trigger the rule MD001
.
For cases where the marker text can not be present on the same line, the syntax
{MD###:#}
can be used to include a line number. If some-test.md
needs custom
configuration, a some-test.json
is used to provide a custom options.config
for that scenario. Tests run by markdownlint-test-scenarios.js
use AVA's
snapshot feature. To update snapshots (for example, after
modifying a test file), run npm run update-snapshots
and include the updated
files with the pull request.
Lint before sending a pull request by running npm run lint
. There should be no
issues.
Run a full continuous integration pass before sending a pull request via npm run ci
. Code coverage should always be 100%. As part of a continuous
integration run, generated files may get updated and fail the run - commit them
to the repository and rerun continuous integration.
Pull requests should contain a single commit. If necessary, squash multiple commits before creating the pull request and when making changes. (See Git Tools - Rewriting History for details.)
Open pull requests against the next
branch. That's where the latest changes
are staged for the next release. Include the text "(fixes #??)" at the end of
the commit message so the pull request will be associated with the relevant
issue. End commit messages with a period (.
). Once accepted, the tag fixed in next
will be added to the issue. When the commit is merged to the main branch
during the release process, the issue will be closed automatically. (See
Closing issues using keywords for details.)
Please refrain from using slang or meaningless placeholder words. Sample content can be "text", "code", "heading", or the like. Sample URLs should use example.com which is safe for this purpose. Profanity is not allowed.
In order to maintain the permissive MIT license this project uses, all contributions must be your own and released under that license. Code you add should be an original work and should not be copied from elsewhere. Taking code from a different project, Stack Overflow, or the like is not allowed. The use of tools such as GitHub Copilot that generate code from other projects is not allowed.
Thank you!