grunt-markdown-to-json

Extract YAML front-matter from Markdown files to a single JSON file.

The use case for this markdown-to-yaml-to-json task is somewhat narrow. I use it to strip the YAML front-matter off a set of blog posts written in Markdown. The metadata for each file is combined into a single object, then emitted as a JSON file.

Along the way, a few extra fields are created for each article:

• an ISO 8601 formatted date
• a preview of the actual body content
• the basename of the file, used as a key to get back to the metadata

It wraps the markdown-to-json npm module.

I realize a more descriptive name for this task would be markdown-yaml-frontmatter-to-json but that's pretty wordy. And you can't have dashes in tasks, so the shortname is m2j.

Incidentally, I never had Grunt figured out until I wrote this contrib module. It finally made sense. Try pulling down the source for this and running grunt test for yourself, and inspect the test input (test\fixtures) and output (test\expected).

Getting Started

This plugin requires 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-markdown-to-json --save-dev

The plugin should load automatically, since your Gruntfile.js parses your package.json, which has a reference to the library.

The "m2j" task

Pretend you have a folder structure like this:

.
├── Gruntfile.coffee
├── component.json
├── package.json
└── source
    ├── articles
    │   ├── bellflower.md    <--
    │   ├── fiddler.md       <--
    │   └── lottery.md       <--
    ├── favicon.ico
    ├── index.jade
    ├── style.styl
    ├── styles
    │   ├── h5bp.css
    │   ├── main.css
    │   └── normalize.css
    └── templates
        └── h5bp.jade

Each Markdown file in the articles directory has a bit of YAML metadata, like the title of the article, author, and tags. We want just the metadata from all three to be combined into a single JSON stringified file, called articles.json.

Now grunt release will build a release folder that looks like this:

.
├── Gruntfile.coffee
├── component.json
├── package.json
├── release
│   ├── articles
│   │   ├── bellflower.html
│   │   ├── fiddle.html
│   │   └── lottery.html
│   ├── articles.json            <---
│   ├── favicon.ico
│   ├── index.html
│   └── style.css
└── source

Configuring your m2j task

In your project's Gruntfile, add a section named m2j to the data object passed into grunt.initConfig().

m2j: {
	release: {
		options: {
			minify: true,
			width: 60
		},
		files: {
			'release/articles.json': ['source/articles/*.md']
		},
	}
}

Options

options.minify

Type: Boolean Default value: false

If true, then the JSON.stringify is instructed to strip unnecessary linebreaks, making the resulting .json file smaller.

options.width

Type: Int Default value: 70

No more than width charactes from the Markdown file's body is saved in the preview element. Trailing ellipses are added.

options.files

This is the common source / destination pairing you see in all Grunt tasks. Note that you have just one destination, one or more sources, per pair.

See the docs, specifically the Compact Format and Files Object Format for examples.

Contributing

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.

Release History

Version 0.4.0 is the initial version, which matches 0.4.1 of the npm module m2j.

Scott Stanfield
[e-mail]
April 2014