This module has moved and is now available at https://github.com/jonkemp/inline-css/tree/master/packages/extract-css. This repository is no longer maintained.

Extract the CSS from an HTML document.

Install with npm

npm install --save extract-css

var extractCss = require('extract-css');
var options = {
      url: './',
      applyStyleTags: true,
      removeStyleTags: true,
      applyLinkTags: true,
      removeLinkTags: true,
      preserveMediaQueries: false
  };

extractCss(document, options, function (err, html, css) {
    console.log(html);
    console.log(css);
});

Type: Boolean

Whether to inline styles in <style></style>.

Type: Boolean

Whether to resolve <link rel="stylesheet"> tags and inline the resulting styles.

Type: Boolean

Whether to remove the original <style></style> tags after (possibly) inlining the css from them.

Type: Boolean

Whether to remove the original <link rel="stylesheet"> tags after (possibly) inlining the css from them.

Type: String

How to resolve hrefs. Required.

Type: Boolean

Preserves all media queries (and contained styles) within <style></style> tags as a refinement when removeStyleTags is true. Other styles are removed.

Type: Object
Default: { EJS: { start: '<%', end: '%>' }, HBS: { start: '{{', end: '}}' } }

An object where each value has a start and end to specify fenced code blocks that should be ignored during parsing. For example, Handlebars (hbs) templates are HBS: {start: '{{', end: '}}'}. Note that codeBlocks is a dictionary which can contain many different code blocks, so don't do codeBlocks: {...} do codeBlocks.myBlock = {...}.

When a data-embed attribute is present on a <style></style> tag, extract-css will not inline the styles and will not remove the <style></style> tags.

This can be used to embed email client support hacks that rely on css selectors into your email templates.

The code for this module was originally taken from the Juice library.

MIT © Jonathan Kemp