A Grunt plugin to help bind Grunt tasks to Git hooks
This plugin requires at least Grunt ~0.4.1
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-githooks --save-dev
Once the plugin has been installed, it may be enabled inside your Gruntfile with this line of JavaScript:
grunt.loadNpmTasks('grunt-githooks');
In your project's Gruntfile, add a section named githooks
to the data
object passed into grunt.initConfig()
.
grunt.initConfig({
githooks: {
options: {
// Task-specific options go here.
},
all: {
options: {
// Target-specific options go here
},
// Hook definitions go there
}
},
})
Hooks are listed as keys of your target configuration. Any key other than
options
is considered the name of a hook you want to create. The simplest way
to define a hook is to provide a space-separated list of the tasks you want
the hook to run as the value.
For example:
grunt.initConfig({
githooks: {
all: {
// Will create `./git/hooks/pre-commit` file which will be used at every commit,
// so that to run the `jshint` and `test:unit` tasks before commit really happen.
'pre-commit': 'jshint test:unit',
}
}
});
The plugin warns you if the name matches one of the hooks announced in the Git documentation. It will still create the hook, though, in case Git introduces new hooks in the future.
If you need to override a few options for a given hook only, you can use an
Object instead of a String.
The taskNames
property will then correspond to the tasks you want to run.
Any other key will be merged into the options.
grunt.initConfig({
githooks: {
all: {
options: {
template: 'path/to/a/template'
},
// Will bind the jshint and test:unit tasks
// with the template specified above
'pre-commit': 'jshint test:unit',
// Will bind the bower:install task
// with a specific template
'post-merge': {
taskNames: 'bower:install',
template: 'path/to/another/template'
}
}
}
})
If you happen to have existing hooks in your hook folder, the plugin appends the code launching Grunt at the end of your hooks. You can also insert marker comments in your hooks to specify exactly where you want them inserted. Your existing hook would look something like this:
// Some code run before Grunt starts
// GRUNT-GITHOOKS START // GRUNT-GITHOOKS END
// Some code run after Grunt starts
The markers get automatically inserted when the plugin appends code, so hooks
get updated cleanly the next time you run grunt githooks
.
By default, the plugin generate NodeJS scripts for the hooks. Reasoning behind this is that creating Shell scripts won't work well for people using Windows. Plus, NodeJS is already installed as Grunt kinda needs it. However, you're not tied to it and you can customise the generated script entirely. In case of a Shell script:
grunt.initConfig({
githooks: {
all: {
options: {
// Customize the hashbang to say 'Shell script'
hashbang: '#!/bin/sh',
// Plugin comes in with a sheel script template already. Handy, innit?
template: './node_modules/grunt-githooks/templates/shell.hb',
// Customize the markers so comments start with #
startMarker: '## LET THE FUN BEGIN',
endMarker: '## PARTY IS OVER'
}
}
}
});
In the template, you've got access to the following variables:
String
with the name of the current hookString
with the name of the command to runString
with the name of the tasks to be runString
with the list of arguments to provide to the taskPretty annoying when you're using a library that's missing the exact extension
point you need to tweak its functionalities? grunt-githooks
is based on a lot
of small functions and most of them are exposed so you can override them. If you
need feel, free to tinker with the internals (at your own risk though ;)).
Could be something along:
var gruntGithooks = require('grunt-githooks/tasks/githooks');
var originalFunction = gruntGithooks.internals.Hook.prototype.getHookContent;
gruntGithooks.internals.Hook.prototype.getHookContent = function () {
console.log('Loading content of an existing hook');
originalFunction.apply(this, arguments);
};
Type: String
Defaults: grunt
The command that will be run by the hook. This has initially been introduced to allow specifying the full path to Grunt in some specific cases. It can also allow you to run another command than Grunt if you need.
Type: String
A space separated list of tasks that will be run by the hook.
Type: String
Additional CLI arguments to be passed to the command run by the hook.
Type: String
Defaults: '#!/usr/bin/env node'
The hashbang that will be used at the top of the hook script file. If a hook already exist, the hashbang will be used to check if its ok to append/insert code in it (to avoid inserting Node code in a Python hook for example).
Type: String
Path to the Handlebars template used to generate the code that will run Grunt
in the hook. Default template is the node.js.hb
file located in the templates
folder of the plugin. It also contains a shell.hb
file with the template for a
shell script hook.
Note: Handlebars escapes HTML special characters if you use only two curly braces to insert a variable in your template. Make sure you use three
{{{my_var}}}
if you need to insert variable that contains quotes, chevrons or anything that would be HTML escaped
Type: String
Default: '// GRUNT-GITHOOKS START'
Type: String
Default: '// GRUNT-GITHOOKS END'
startMarker
and endMarker
are markers the plugin use to know where to insert
code if a hook already exist. If the existing hook doesn't have these markers,
the code will simply be appended.
Type: Boolean
Default false
By default, the inserted code will exit the process after Grunt has run, using a -1 exit code if the task(s) failed. If you're inserting the code running Grunt in the middle of an existing hook, you might want to disable this so any code after what was inserted by the plugin runs.
Type: String
Default value: '.git/hooks'
This option allows you to choose in which directory the hooks should be generated. Comes in handy if your Gruntfile is not at the root of your Git project.
In lieu of a formal style-guide, take care to maintain the existing coding style. Please file a Pull Request along your issues.
The process actually is quite simple:
issue-{$integer}
and commit against that one.dev
using --no-ff
so we have a separate commit for that merge.dev
, we can merge to master
, again using --no-ff
.master
will get tagged. We use the SemVer standard, so no leading v
.If your PR is successful, you will get added as contributor to the repo. We
trust you after your first PR made it into the repo and you then have access
for further changes, handling issues, etc. So the important thing is to add
your name to the package.json
array of contributors
when changing or adding
some code for a PR. Please do that in a separate commit.
See Changelog for details or the Release list - (see latest for recent updates).