Markdown parser done right. Fast and easy to extend.
- Supports the CommonMark spec + syntax extensions + sugar (URL autolinking, typographer).
- Configurable syntax! You can add new rules and even replace existing ones.
- High speed!
- Community plugins on npm.
node.js:
npm install remarkable --save
browser (CDN):
import { Remarkable } from 'remarkable';
var md = new Remarkable();
console.log(md.render('# Remarkable rulezz!'));
// => <h1>Remarkable rulezz!</h1>
or with commonjs
const { Remarkable } = require('remarkable');
var md = new Remarkable();
console.log(md.render('# Remarkable rulezz!'));
// => <h1>Remarkable rulezz!</h1>
If installed globally with npm
:
cat myfile.md | remarkable
remarkable --file myfile.md
# get options
remarkable -h
See the docs directory for documentation on the following topics:
By default, remarkable is configured to be similar to GFM, but with HTML disabled. This is easy to change if you prefer different settings.
There are two ways to define options.
Define options in the constructor:
// Actual default values
var md = new Remarkable({
html: false, // Enable HTML tags in source
xhtmlOut: false, // Use '/' to close single tags (<br />)
breaks: false, // Convert '\n' in paragraphs into <br>
langPrefix: 'language-', // CSS language prefix for fenced blocks
// Enable some language-neutral replacement + quotes beautification
typographer: false,
// Double + single quotes replacement pairs, when typographer enabled,
// and smartquotes on. Set doubles to '«»' for Russian, '„“' for German.
quotes: '“”‘’',
// Highlighter function. Should return escaped HTML,
// or '' if the source string is not changed
highlight: function (/*str, lang*/) { return ''; }
});
console.log(md.render('# Remarkable rulezz!'));
// => <h1>Remarkable rulezz!</h1>
Or define options via the .set()
method:
import { Remarkable } from 'remarkable';
var md = new Remarkable();
md.set({
html: true,
breaks: true
});
Note: To achieve the best possible performance, don't modify a Remarkable
instance on the fly. If you need multiple configurations, create
multiple instances and initialize each with a configuration that is ideal for
that instance.
Remarkable offers some "presets" as a convenience to quickly enable/disable active syntax rules and options for common use cases.
Enable strict CommonMark mode with the commonmark
preset:
import { Remarkable } from 'remarkable';
var md = new Remarkable('commonmark');
Enable all available rules (but still with default options, if not set):
import { Remarkable } from 'remarkable';
var md = new Remarkable('full');
// Or with options:
var md = new Remarkable('full', {
html: true,
typographer: true
});
Apply syntax highlighting to fenced code blocks with the highlight
option:
import { Remarkable } from 'remarkable';
import hljs from 'highlight.js' // https://highlightjs.org/
// Actual default values
var md = new Remarkable({
highlight: function (str, lang) {
if (lang && hljs.getLanguage(lang)) {
try {
return hljs.highlight(lang, str).value;
} catch (err) {}
}
try {
return hljs.highlightAuto(str).value;
} catch (err) {}
return ''; // use external default escaping
}
});
Enabled by default:
Disabled by default:
- <sup> -
19^th^
- <sub> -
H~2~O
- abbreviations
- <ins> -
++inserted text++
(experimental) - <mark> -
==marked text==
(experimental)
HEADS UP!: Experimental extensions can be changed later for something like Critic Markup, but you will still be able to use old-style rules via external plugins if you prefer.
var md = new Remarkable();
md.inline.ruler.enable([ 'ins', 'mark' ]);
md.block.ruler.disable([ 'table', 'footnote' ]);
// Enable everything
md = new Remarkable('full', {
html: true,
typographer: true,
});
//
// Manually enable rules, disabled by default:
//
var md = new Remarkable();
md.core.ruler.enable([
'abbr'
]);
md.block.ruler.enable([
'footnote',
'deflist'
]);
md.inline.ruler.enable([
'footnote_inline',
'ins',
'mark',
'sub',
'sup'
]);
Although full-weight typographical replacements are language specific, remarkable
provides coverage for the most common and universal use cases:
import { Remarkable } from 'remarkable';
var md = new Remarkable({
typographer: true,
quotes: '“”‘’'
});
// Disable rules at all:
md.core.ruler.disable([ 'replacements', 'smartquotes' ]);
// Actual default replacements:
//
// '' → ‘’
// "" → “”. Set '«»' for Russian, '„“' for German, empty to disable
// (c) (C) → ©
// (tm) (TM) → ™
// (r) (R) → ®
// +- → ±
// (p) (P) -> §
// ... → … (also ?.... → ?.., !.... → !..)
// ???????? → ???, !!!!! → !!!, `,,` → `,`
// -- → –, --- → —
//
Of course, you can also add your own rules or replace the defaults with something more advanced or specific to your language.
Easily load plugins with the .use()
method:
var md = new Remarkable();
md.use(plugin1)
.use(plugin2, opts)
.use(plugin3);
Please refer to the plugin documentation to create your own plugins.
Autoconvert URL-like text to links
import { Remarkable } from 'remarkable';
import { linkify } from 'remarkable/linkify';
var md = new Remarkable().use(linkify);
UMD bundle provides linkify out of the box
const { Remarkable, linkify, utils } = window.remarkable;
Big thanks to John MacFarlane for his work on the CommonMark spec and reference implementations. His work saved us a lot of time during this project's development.
Related Links:
- https://github.com/jgm/CommonMark - reference CommonMark implementations in C & JS, also contains latest spec & online demo.
- http://talk.commonmark.org - CommonMark forum, good place to collaborate developers' efforts.
Parser consists of several responsibility chains filled with rules. You can reconfigure any of them as you wish. Renderer also can be modified and extended. See source code to understand details. Pay attention to these properties:
Remarkable.core
Remarkable.core.ruler
Remarkable.block
Remarkable.block.ruler
Remarkable.inline
Remarkable.inline.ruler
Remarkable.renderer
Remarkable.renderer.rules
Here is result of CommonMark spec parse at Core i5 2.4 GHz (i5-4258U):
$ benchmark/benchmark.js spec
Selected samples: (1 of 27)
> spec
Sample: spec.txt (110610 bytes)
> commonmark-reference x 40.42 ops/sec ±4.07% (51 runs sampled)
> current x 74.99 ops/sec ±4.69% (67 runs sampled)
> current-commonmark x 93.76 ops/sec ±1.23% (79 runs sampled)
> marked-0.3.2 x 22.92 ops/sec ±0.79% (41 runs sampled)
As you can see, remarkable
doesn't pay with speed for its flexibility. Because
it's written in monomorphic style and uses JIT inline caches effectively.
- Jon Schlinkert github/jonschlinkert
- Alex Kocharin github/rlidwka
- Vitaly Puzrin github/puzrin