Grunt plugin for automating all the release steps of your node lib or bower component, with optional publishing to npm.
Releasing a new version of your killer Node/Bower/Component/JS lib looks something like this:
- bump the version in your
package.json
file. - stage the package.json file's change.
- commit that change with a message like "release 0.6.22".
- create a new git tag for the release.
- push the changes out to GitHub.
- also push the new tag out to GitHub.
- create a .zip release on GitHub.
- publish the new version to npm.
Cool, right? No! What's wrong with you? Automate all that:
grunt release
Done. No more GitHub issues from angry people reminding you how often you forget to do one or more of the steps.
If you haven't used Grunt before, be sure to check out the Getting Started guide, as it explains how to create a Gruntfile as well as install and use Grunt plugins. Once you're familiar with that process, you may install this plugin with this command:
npm install grunt-release --save-dev
Once the plugin has been installed, it may be enabled inside your Gruntfile with this line of JavaScript:
grunt.loadNpmTasks('grunt-release');
Patch Release:
grunt release
or
grunt release:patch
Minor Release:
grunt release:minor
Major Release:
grunt release:major
Specific Version Release:
grunt release:1.2.3
Pre-release
grunt release:prerelease
prerelease
will just update the number after MAJOR.MINOR.PATCH
(eg: 1.0.0-1
)
If you want to add an alphanumeric identifier, you will need to add it by hand.
Example: add -alpha.0
to get something like 1.0.0-alpha.0
. Calling grunt release:prerelease
will just update the last number to 1.0.0-alpha.1
.
Releasing Unstable/Beta Versions
Sometimes it is useful to publish an 'unstable' or 'beta' version to npm
, while leaving your last stable release as the default that gets installed on an npm install
.
npm
accomplishes this using the --tag myUnstableVersion
flag. You can enable this flag in grunt-release either by setting the npmtag
option:
release: {
options: {
npmtag: 'canary',
}
}
or by passing the CLI arg:
grunt release --npmtag canary
NOTE: If the tag you pass is true, then the tag will be the new version number after the bump. Otherwise it will be the string you provided.
Bump multiple files at once
Sometimes you may need to bump multiple files while releasing.
release: {
options: {
additionalFiles: ['bower.json']
}
}
You can also provide multiple files in this array or provide a string with multiple file paths separated by comma (,
).
The version to bump is set in the master file defined with option 'file' (default : package.json). This version will be propagated to every additionalFiles.
Dry Run:
To see what grunt-release does, without really changing anything, use --no-write
option.
grunt release --no-write
You'll see something like:
>> Release dry run
>> bumped version to 0.8.0
>> staged package.json
>> committed package.json
>> created new git tag: 0.8.0
>> pushed to remote git repo
>> pushed new tag 0.8.0 to remote git repo
>> published version 0.8.0 to npm
>> created 0.8.0 release on github.
Done, without errors.
The following are all the release steps, you can disable any you need to:
release: {
options: {
bump: false, //default: true
changelog: true, //default: false
changelogText: '<%= version %>\n', //default: '### <%= version %> - <%= grunt.template.today("yyyy-mm-dd") %>\n'
file: 'component.json', //default: package.json
add: false, //default: true
commit: false, //default: true
tag: false, //default: true
push: false, //default: true
pushTags: false, //default: true
npm: false, //default: true
npmtag: true, //default: no tag
indentation: '\t', //default: ' ' (two spaces)
folder: 'folder/to/publish/to/npm', //default project root
tagName: 'some-tag-<%= version %>', //default: '<%= version %>'
commitMessage: 'check out my release <%= version %>', //default: 'release <%= version %>'
tagMessage: 'tagging version <%= version %>', //default: 'Version <%= version %>',
beforeBump: [], // optional grunt tasks to run before file versions are bumped
afterBump: [], // optional grunt tasks to run after file versions are bumped
beforeRelease: [], // optional grunt tasks to run after release version is bumped up but before release is packaged
afterRelease: [], // optional grunt tasks to run after release is packaged
updateVars: [], // optional grunt config objects to update (this will update/set the version property on the object specified)
github: {
apiRoot: 'https://git.example.com/v3', // Default: https://github.com
repo: 'geddski/grunt-release', //put your user/repo here
accessTokenVar: 'GITHUB_ACCESS_TOKE', //ENVIRONMENT VARIABLE that contains GitHub Access Token
// Or you can use username and password env variables, we discourage you to do so
usernameVar: 'GITHUB_USERNAME', //ENVIRONMENT VARIABLE that contains GitHub username
passwordVar: 'GITHUB_PASSWORD' //ENVIRONMENT VARIABLE that contains GitHub password
}
}
}
If you want to use multiline commit messages just pass an array to the commitMessage
option instead of a string.
By default, the queued tasks ran through the beforeBump
, afterBump
, beforeRelease
, afterRelease
options will not inherit the flags specified when the release
task is ran. To preserve those flags, you can optionally pass an object to those arrays in the following format:
{
name: 'taskName',
preserveFlags: true // defaults to false
}
You can still pass only strings to that array, or mix the two as need be. For example:
release: {
options: {
beforeRelease: ['oneTask', { name: 'anotherTask', preserveFlags: true }]
}
}
- Yes, you have to use environment variables.
- The GitHub Releases API is still unstable and may change in the future.
- You should use an environment variable to set an access token and "link" it via
accessTokenVar
. - You can use environment variables for username (referenced in
usernameVar
) and password (passwordVar
). - We don't encourage you to use username and password for the GitHub release.
For node libs, leave file
option blank as it will default to package.json
. For Bower components, set it to bower.json
.
MIT