rehype plugins to minify HTML.
Contents
What is this?
This GitHub repository is a monorepo that contains a couple utilities, ±30
plugins, and a preset with good and safe default, to minify HTML:
In
<!doctype html>
<html lang="en-GB">
<head>
<meta charset="utf-8">
<meta http-equiv="content-language" content="en-US">
<script src="index.js" type="text/javascript" language="javascript">
alert(true);
</script>
<link rel="stylesheet" href="index.css" type="text/css">
<title>Foo & bar</title>
</head>
<body>
<h1 class="foo foo">bar bar</h1>
<p id="alfred" id="alfred"> <strong>foo</strong> <em>bar</em> </p>
<button type="BUTTON" onclick="javascript:return false">Alpha</button>
</body>
</html>
Out
<!doctypehtml><html lang=en><meta charset=utf8><script src=index.js></script><link href=index.css rel=stylesheet><title>Foo & bar</title><h1 class=foo>bar bar</h1><p id=alfred><strong>foo</strong> <em>bar</em></p><button onclick=return!1 type=button>Alpha</button>
When should I use this?
This project is useful when you want to improve the size of HTML documents.
It’s particularly useful when you’re already using rehype (and remark?) to
process HTML.
Presets
Presets are combinations of plugins.
One preset is maintained here:
Plugins
The following plugins maintained here are included in the above preset.
The following plugins are not included in the preset as they are potentially
dangerous, can make sites slower in certain cases, or need extra
configuration.
Read their readmes before using:
Also: pass allowDangerousCharacters
to rehype-stringify
if you trust your content.
Benchmark
Here’s a benchmark comparing the results from html-minifier
and rehype-minify
.
To summarize: differences are negligible, in fact, minifying HTML doesn’t matter
much.
What does matter is using good compressions, such as gzip.
Note: html-minifier
sometimes crashes, such as on amazon.
name | raw | gzip |
---|
original | rehype-minify | html-minifier | original | rehype-minify | html-minifier |
---|
amazon | 567.85KB | 5.71% | 💥 | 128.65KB | 5.07% | 💥 |
---|
bbc | 531.54KB | 4.70% | 4.58% | 58.09KB | 1.85% | 1.96% |
---|
bootstrap | 33.96KB | 22.36% | 22.82% | 8.12KB | 5.17% | 5.36% |
---|
compat-table | 8.67MB | 16.34% | 16.74% | 275.06KB | 8.26% | 10.06% |
---|
ecmascript | 6.62MB | 10.86% | 11.56% | 789.66KB | 6.39% | 6.60% |
---|
eloquent | 80.08KB | 6.43% | 6.45% | 19.13KB | 1.25% | 1.26% |
---|
github | 233.14KB | 13.39% | 14.51% | 38.86KB | 7.48% | 7.91% |
---|
google | 253.98KB | 3.13% | 3.20% | 80.56KB | 3.64% | 3.63% |
---|
guardian | 765.54KB | 6.75% | 8.71% | 112.21KB | 3.96% | 4.38% |
---|
html | 12.72MB | 3.15% | 3.19% | 1.85MB | 5.49% | 5.58% |
---|
linkedin | 120.77KB | 32.29% | 32.34% | 14.4KB | 13.99% | 13.91% |
---|
microsoft | 174.26KB | 22.44% | 34.12% | 26.45KB | 12.96% | 15.45% |
---|
nbc | 1.64MB | 5.94% | 6.28% | 221.08KB | 1.86% | 1.92% |
---|
npm | 27.98KB | 3.03% | 2.61% | 8.28KB | 2.22% | 2.01% |
---|
nytimes | 822.38KB | 11.80% | 11.80% | 137.47KB | 2.71% | 2.72% |
---|
slack | 205.45KB | 4.76% | 4.55% | 49.63KB | 2.06% | 1.98% |
---|
stackoverflow | 169.8KB | 18.64% | 19.10% | 45.4KB | 6.42% | 6.67% |
---|
vice | 337.93KB | 3.48% | 3.22% | 92.58KB | 2.61% | 2.52% |
---|
wikipedia | 768.18KB | 6.08% | 6.76% | 144KB | 1.58% | 1.77% |
---|
total | 34.63MB | 31.9MB | 31.83MB | 4.05MB | 3.85MB | 3.85MB |
---|
Huge differences in results are suspicious and may point to bugs.
💥 — Crash.
Security
Use of rehype-preset-minify
is safe by default if the tree is already safe.
As rehype works on HTML and improper use of HTML can open you up to a
cross-site scripting (XSS) attack, use of rehype can also be unsafe.
Use rehype-sanitize
to make the tree safe.
To further optimize the result disregarding security, use the extra plugins
listed above and pass allowDangerousCharacters
to
rehype-stringify
.
Related
Contribute
See contributing.md
in rehypejs/.github
for ways
to get started.
See support.md
for ways to get help.
This project has a code of conduct.
By interacting with this repository, organization, or community you agree to
abide by its terms.
License
MIT © Titus Wormer