Table of Contents
In general, the conventions used by this task are as follows:
Templates
- Files with extension
.tmpl.md
are generally templates that will be compiled one-to-one into documents - Files with extension
.md
are generally intended to be used as includes. {%= _.doc("foo") %}
is used to included files from your project's./docs
directory{%= _.include("foo") %}
is used to include boilerplate files from grunt-readme
To change the plugin's defaults, add a section to your project's Gruntfile named readme
to the data object passed into grunt.initConfig()
:
grunt.initConfig({
// The "repos" task
repos: {
options: {}
},
// The "readme" task
readme: {
options: {
metadata: {}
}
}
});
grunt.loadNpmTasks('grunt-readme');
grunt.registerTask('default', ['readme']);
Add YAML front matter to documents to extend the metadata that is supplied to your project's templates.
---
username: jonschlinkert
---
This is probably most useful when:
- You need to use the same or similar templates on a number of different projects
- You want to supply data to the templates that won't typically be found in package.json
Code comments may be used in markdown templates, and they will be stripped from the rendered README as long as they adhere to the following syntax:
This task automatically adjusts heading levels in included templates. For example, #
is adjusted to ##
, so that heading levels "line up" properly after the README is built.
This can cause problems if you're using hashes for a reason other than headings, such as CSS Id's in code comments. So to prevent grunt-readme from converting #id {}
to ##id {}
, just add a single backtick before the hash: `#id {}
.
To prevent Lo-Dash from attempting to evaluat templates that shouldn't be (as with code examples), just use square brackets instead of curly braces in any templates that have similar patterns to these: {%= .. %}
, {%- .. %}
, and {% .. %}
. The square brackets will be replaced with curly braces in the rendered output.
readme: {
options: {
readme: '',
templates: '',
metadata: '',
sep: '\n',
prefixes: [],
contributing: true
}
}
Type: String
Default: ./node_modules/grunt-readme/tasks/templates/README.tmpl.md
By default, if no options are specified the task will look for a README.md.tmpl
template to use, if none is found the task will use the "starter" file supplied by grunt-readme
(more detail below). Example:
readme: {
options: {
readme: 'path/to/custom/README.md.tmpl'
}
}
- If the
readme
options is defined, the task will use that custom template. - If (1) is undefined, the task uses the directory defined by
options: { docs: ''}
- If (2) is undefined, the task checks if
README.tmpl.md
exists in the./docs
directory (without having to define it in the options) - if (3) is undefined,
options: { resolve: { readme: ''}}
attempts to automagically use aREADME.tmpl.md
template fromnode_modules
. The module must must be defined indevDependencies
. Note that for a README template to resolve properly fromnode_modules
, themain
property in thepackage.json
of the module being referenced must specify the path to the template. This option is probably most useful when you plan to use the same README template on a number of projects. - If (4) is undefined, the task uses the "starter" README template from
grunt-readme
.
Type: Object
Default: package.json
Optional source of metadata to extend the data object that is passed as context into the templates. Context of the data object is the value of this
, and properties in package.json
will be ignored when matching properties are defined on the metadata
object. Example:
readme: {
options: {
metadata: {
name: 'Foo',
description: 'This is foo.'
}
}
}
Or specify the path or paths to any .json
or .yml
files to use. Any of the following formats will work:
readme: {
options: {
metadata: 'docs/metadata.json'
}
}
Array of files:
readme: {
options: {
metadata: ['docs/one.json', 'docs/two.yml'],
}
}
minimatch (wilcard/globbing) patterns:
readme: {
options: {
metadata: ['docs/*.{json,yml}', 'foo.json']
}
}
Since context is the value of "this", the metadata
path is not required in templates, only property names:
{%= name %}
(e.g. not{%= metadata.name %}
) =>Foo
{%= description %}
=>This is foo.
Type: String
Default: ./docs/
Override the default directory for files included using {%= _.doc('foo.md') %}
. This defaults to the ./docs
directory in the root of your project.
readme: {
options: {
docs: 'foo/'
}
}
Type: String
Default: ./node_modules/grunt-readme/tasks/templates/
(relative to your project)
Override the default cwd
for files included by using {%= _.include('foo.md') %}
. By default, the include
mixin will look for files in ./node_modules/grunt-readme/tasks/templates
directory, where some starter templates are stored. (Also see examples →)
readme: {
options: {
templates: 'bar/'
}
}
Type: Array
Default: grunt|helper|mixin
Any string defined in the remove will be removed from the content passed in using the {%= _.shortname() %}
template. Example:
readme: {
options: {
remove: ["foo", "bar", "baz"]
}
}
Given a package.json
with the following property:
{
"name": "foo-module"
}
when referenced in a template like this:
## {%= _.titleize(_.shortname(name)) %}
will renders to:
## Module
Type: Boolean
Default: True
By default, the README task copies a basic CONTRIBUTING.md
file to the root of your project. If one exists, the task will skip this. If you wish to prevent the task from adding this file to your project, set the contributing
option to false
.
Type: String
Default: \n
Separator to use between sections of content that is included using the include
or doc
mixins (more about these in the "Mixins" section below). This option is more useful when you use minimatch patterns to specify the files to include.
The sep
option can either be defined in the task options:
readme: {
options: {
sep: '\n***\n'
}
}
or as a second parameter in the include
or doc
mixins.
{%= _.include("docs-*.md", "***") %}
(more below...){%= _.doc("*.md", "\n***\n") %}
(more below...)
Mixins use the following formats:
_.mixin()
: when used in JavaScript{%= _.mixin() %}
: when used in templates
Three different mixins are built into the task for including "external" content:
include
,doc
andresolve
. Each is used for a different purpose.
Here is a summary of what they do (settings for the include
and doc
mixins can be customized in the task options):
{%= _.include("file.md") %}
: include a file (or files using minimatch patterns) from the./templates/
directory of the grunt-readme task.{%= _.doc("file.md") %}
: include a file (or files using minimatch patterns) from the./docs/
directory of your project.{%= _.resolve("file.md") %}
: include a specific file from node_modules`.{%= _.contrib("file.md") %}
: include a file (or files using minimatch patterns) from the./contrib/
directory of the grunt-readme task. This mixin is for the Assemble.
Use the include
mixin in templates to pull in content from other files:
{%= _.include("examples.md") %}
Minimatch patterns may also be used:
{%= _.include("docs-*.md") %}
Unless overridden in the templates
option, the include
mixin will use the ./node_modules/grunt-readme/tasks/templates/
directory (from the root of your project) as the cwd
for templates.
Same as the include
mixin but is hard-coded to use the docs/
folder of your project as the cwd
for templates.
Use the resolve
mixin in templates to include content from named NPM modules listed in devDependencies
:
{%= _.resolve("my-boilerplate-readme") %}
where my-boilerplate-readme
is the name of a devDependency
currently installed in node_modules
.
For the resolve
mixin to work:
- The referenced file must be listed in the
devDependencies
of your project'spackage.json
, - It must be installed in
node_modules
, and - The referenced project must have the file defined in the
main
property of that project'spackage.json
. - Last, in your templates make sure you use the name of the module, not the name of the file to "include".
example
In the package.json
of the project that will store your templates, you might do something like:
{
"name": "my-boilerplate-readme",
"main": "README.tmpl.md"
}
Get the value of any property in package.json
. Example:
{%= _.meta('name') %}
{%= _.meta('version') %}
{%= _.meta('contributors') %}
{%= _.meta('keywords') %}
A second paramter can be passed in to set the indentation on returned JSON: {%= _.meta('contributors', 4) %}
. This only works for stringified objects.
Also, if left undefined ({%= _.meta() %}
) the mixin will return the entire metadata object (by default, this is the entire contents of package.json
):
Parse and extract comments from specified JavaScript files to generate output for each code comment block encountered.
{%= _.jsdocs("tasks/readme.js") %}
Currently, only the block is output and a link to the block in the source code is provided. This needs to be updated to only generate the markdown for jsdoc comments and to do something to make them more readable.
Add a copyright statement, including the name of the author and the year, or range of years, the copyright is in effect. The primary advantage of using this is to ensure that the copyright dates are correct.
Parameters:
Number
: Optionally define the start year of the project.
Examples:
{%= _.copyright() %}
// => Copyright (c) 2013 Jeffrey Herb, contributors.
{%= _.copyright('2011') %}
// => Copyright (c) 2011-2013 Jeffrey Herb, contributors.
Add a "license statement" to the README, using the license(s) specified in package.json. If you maintain a number of projects, some of which might have more than one license, while others only have one, you can use the _.license()
mixin to automate the process of adding license info.
Examples:
{%= _.license() %}
Released under the MIT license
Customize the output:
{%= _.license('Licensed under the ') %}
Licensed under the MIT license
Render contributors listed in the project's package.json.
Extract the username or org from URLs in the project's package.json. The mixin will extract the username from the homepage
property if it exists. If not, it will try to extract the username from the git.url
property.
Extract the homepage URL from the project's package.json. If a homepage
property doesn't exist, the mixin will create a homepage
URL using the value defined in the git.url
property.
Copy/paste any of these examples into your templates as a starting point.
{%= name %}
grunt-readme
{%= version %}
v{%= version %}
{%= version ? " v" + version : "" %}
{%= version ? " * @version " + version + "\\n" : "" %}
0.1.3 v0.1.3 v0.1.3
- @version 0.1.3\n
{%= description %}
{%= description ? " * " + description + "\\n" : "" %}
Generate your README from a template. If you already use Grunt, this is a no brainer.
- Generate your README from a template. If you already use Grunt, this is a no brainer.\n
{%= homepage ? " | " + homepage : "" %}
{%= homepage ? " * " + homepage + "\n" : "" %}
{%= homepage ? " * @docs " + homepage + "\\n" : "" %}
[AUTHORS](NPM https://npmjs.org/doc/json.html)
If there is an
AUTHORS
file in the root of your package, npm will treat each line as aName <email> (url)
format, where email and url are optional. Lines which start with a # or are blank, will be ignored. [-- NPM]((NPM https://npmjs.org/doc/json.html)
To use author
data from package.json
:
[{%= author.name %}]({%= author.url %})
{%= author.name ? " * @author " + author.name + "\\n" : "" %}
{%= author.url ? " * @link " + author.url + "\\n" : "" %}
- @author Jon Schlinkert\n
- @link https:/jonschlinkert\n
Or, if you prefer to use an AUTHORS
file in the root of the project:
[{%= authors[0].name %}]({%= authors[0].url %})
{%= grunt.template.today() %}
Tue Sep 17 2013 18:38:42
{%= grunt.template.today("yyyy") %}
2013
{%= grunt.template.today("yyyy-mm-dd") %}
2013-09-17
_This file was generated on {%= grunt.template.date("fullDate") %}._
This file was generated on Monday, September 30, 2013.
/*!
* {%= name %} v{%= version %}, {%= grunt.template.today("yyyy-mm-dd") %}
* {%= homepage %}
* Copyright (c) {%= grunt.template.today("yyyy") %} {%= author %}, contributors.
* {%= _.license() %}.
*/
/*!
- grunt-readme v0.1.3, 2013-09-22
- https:/assemble/grunt-readme
- Copyright (c) 2013 [object Object], contributors.
- Released under the MIT license. */
{%= _.include("docs-changelog.md") %}
- 2013-09-21 v0.1.3 Completely refactored. Adds a lot of documentation.
- 2013-09-19 v0.1.0 First commmit.
Or:
* {%= grunt.template.today('yyyy') %} v0.1.0 First commit
- 2013 v0.1.0 First commit
{%= _.license() %}
Released under the MIT license.
{%= _.contributors() %}
Jon Schlinkert Brian Woodward
You can mix and match formats in the metadata
option, all of the following shoulw work:
grunt.initConfig({
pkg: 'package.json',
foo: 'package.json',
bar: grunt.file.readJSON('package.json'),
qux: grunt.file.readJSON('test/fixtures/data/one.json'),
baz: ['<%= bar %>'],
config: {
one: 'test/fixtures/data/one.json',
two: 'test/fixtures/data/two.yml',
three: 'test/fixtures/data/three.json',
pkg: grunt.file.readJSON('package.json'),
qux: grunt.file.readJSON('test/fixtures/data/one.json')
},
// Obviously you can't have duplicate properties on an
// object, so this is just for illustrative purposes
// The point is.. you can get just about as crazy as you want.
readme: {
options: {
metadata: ['<%= pkg %>', '<%= qux %>'],
metadata: ['<%= config.pkg %>', '<%= config.qux %>'],
metadata: ['<%= pkg %>', {foo: 'bar'}],
metadata: ['<%= pkg %>', 'test/fixtures/data/*.{json,yml}'],
metadata: '<%= config.one %>',
metadata: 'test/fixtures/data/one.json',
metadata: ['test/fixtures/data/one.json', 'test/fixtures/data/two.yml'],
metadata: ['test/fixtures/data/two.yml', {description: 'Foo', name: 'Bar'}, '<%= pkg %>', 'test/fixtures/data/*.json', {alpha: 1, beta: 2 }, {kappa: 3, gamma: 4 }, {zed: {orange: 5, apple: 6 } }, '<%= config.one %>', {name: 'New'}, {quux: '<%= qux %>'}, ['one', {pkg: '<%= config.pkg %>'}, 'three'], {arr: ['one', 'two', 'three']}],
metadata: ['<%= config.one %>', '<%= config.two %>'], metadata: 'test/fixtures/data/*.{json,yml}',
metadata: ['test/fixtures/data/*.{json,yml}'],
metadata: ['test/fixtures/data/*.json', 'test/fixtures/data/*.yml'],
metadata: ['test/fixtures/data/*.json', '<%= config.two %>'],
metadata: {
description: 'Foo',
name: 'Bar'
}
}
}
}
Find a bug? Have a feature request? Please create an Issue.
In lieu of a formal styleguide, take care to maintain the existing coding style. Add unit tests for any new or changed functionality. Lint and test your code using grunt, and build the documentation with grunt-readme.
Pull requests are also encouraged, and if you find this project useful please consider "starring" it to show your support! Thanks!
Jon Schlinkert
Brian Woodward
Copyright (c) 2013 Jon Schlinkert, contributors. Released under the MIT license
This file was generated by grunt-readme on Saturday, December 14, 2013.