Lints Pull Request commits with commitlint.
Create a github workflow in the .github
folder, e.g. .github/workflows/commitlint.yml
:
name: Lint Commit Messages
on: [pull_request]
permissions:
contents: read
pull-requests: read
jobs:
commitlint:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- uses: wagoid/commitlint-github-action@v6
Alternatively, you can run on other event types such as on: [push]
. In that case the action will lint the push event's commit(s) instead of linting commits from a pull request. You can also combine push
and pull_request
together in the same workflow.
GitHub's merge queue is a feature that allows you to queue pull requests for merging once they meet certain criteria. When using merge queues, you need to ensure that your workflows are set up to handle the merge_group event, which is triggered when pull requests are added to the merge queue.
To use the commitlint-github-action with merge queues, you need to set up a workflow that listens to the merge_group event. Here's an example of how to configure your workflow:
name: Lint Commit Messages in Merge Queue
on:
merge_group:
types:
- checks_requested
jobs:
commitlint:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
with:
ref: ${{ github.sha }}
- uses: wagoid/commitlint-github-action@v6
To ensure that the merge_group event triggers correctly, you need to have at least one workflow that responds to the pull_request event with a job named the same as the one in your merge_group workflow (commitlint in this example). This is necessary because the merge queue relies on the existence of status checks from the pull request context.
Here's a minimal pull_request workflow to satisfy this requirement:
name: Placeholder Workflow for Merge Queue
on:
pull_request:
jobs:
commitlint:
runs-on: ubuntu-latest
steps:
- name: Checkout Repository
uses: actions/checkout@v3
This workflow can also be a meaningful one that checks out the commits in your PR and runs other checks, but it must have a job named commitlint.
Before you can use merge queues, you need to enable the feature in your repository settings:
- Go to your repository's Settings > Branches.
- Under Branch protection rules, edit the rule for your target branch (e.g. master).
- Enable Require merge queue.
- Specify your new job (e.g. commitlint) and any other required status checks, that must pass before merging.
For more information on configuring merge queues, refer to the GitHub documentation on managing a merge queue.
You can supply these inputs to the wagoid/commitlint-github-action@v6
step.
The path to your commitlint config file.
Default: commitlint.config.mjs
If the config file doesn't exist, config-conventional settings will be loaded as a default fallback.
Details on the configuration file can be found on the commitlint website.
Note: commitlint.config.js
doesn't work with this action. If you use a JS config file, it's required to be an ES Module (.mjs
extension)
Whether you want to fail on warnings or not.
Default: false
Whether you want to fail on errors or not. Still outputs the results, just forces the action to pass even if errors are detected.
Default: true
Link to a page explaining your commit message convention.
default: https://github.com/conventional-changelog/commitlint/#what-is-commitlint
When set to a valid Integer value - X, considers only the latest X number of commits.
default: null
(Equivalent to linting all commits)
Personal access token (PAT) used to interact with the GitHub API. By default, the automatic token provided by GitHub is used. You can see more info about GitHub's default token here.
default: ${{ github.token }}
The error and warning messages for each one of the analyzed commits. This is useful if you want to use the commitlint results in a JSON format in other jobs. See the documentation on how to read JSON information from outputs.
Below you can see an example text output together with its corresponding JSON output:
You have commit messages with errors
⧗ input: wrong message
✖ subject may not be empty [subject-empty]
✖ type may not be empty [type-empty]
✖ found 2 problems, 0 warnings
ⓘ Get help: https://github.com/conventional-changelog/commitlint/#what-is-commitlint
⧗ input: chore: my message
⚠ body must have leading blank line [body-leading-blank]
⚠ found 0 problems, 1 warnings
ⓘ Get help: https://github.com/conventional-changelog/commitlint/#what-is-commitlint
[
{
"hash": "cb0f846f13b490c2fd17bd5ed0b6f65ba9b86c75",
"message": "wrong message",
"valid": false,
"errors": ["subject may not be empty", "type may not be empty"],
"warnings": [],
},
{
"hash": "cb14483cbde23b61322ffb8d3fcdc87f514a3141",
"message": "chore: my message\n\nsome context without leading blank line",
"valid": true,
"errors": [],
"warnings": ["body must have leading blank line"],
},
]
This is a Docker
action, and was made like this so that you can run it with minimum setup, regardless of your repo's environment. It comes packed with the most famous shared configurations that you can use in your commitlint config's extends
field:
- @commitlint/config-angular
- @commitlint/config-conventional
- @commitlint/config-lerna-scopes
- @commitlint/config-patternplate
- conventional-changelog-lint-config-canonical
- commitlint-config-jira
- commitlint-config-function-rules
Apart from the shared configurations that are included by default, you can also include extra dependencies for other configs and plugins that you want to use.
In order to do so, you can use NODE_PATH
env var to make the action take those dependencies into account. Below is an example workflow that does that.
name: Lint Commit Messages
on: [pull_request]
permissions:
contents: read
pull-requests: read
jobs:
commitlint:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- uses: actions/setup-node@v2
with:
node-version: '14'
- run: npm install
# Run the commitlint action, considering its own dependencies and yours as well 🚀
# `github.workspace` is the path to your repository.
- uses: wagoid/commitlint-github-action@v6
env:
NODE_PATH: ${{ github.workspace }}/node_modules
💡 You can see other ways to install your dependencies (including private ones) in the Setup Node action's docs.