Latest commit 45441d4 Nov 17, 2018
Type Name Latest commit message Commit time
Failed to load latest commit information.
errors.json rehype-parse: add support for `<noscript>` parse errors Jun 10, 2018
index.js rehype-parse: add `space` option Jul 17, 2018
package.json rehype-parse: 6.0.0 Nov 17, 2018 Update ``s Nov 17, 2018


Build Coverage Downloads Size Chat

Parser for unified. Parses HTML to hast syntax trees. Used in the rehype processor but can be used on its own as well.

Announcing the unified collective! 🎉 Read more about it on Medium »



🥇 Gatsby

🥉 Compositor





npm install rehype-parse


This example shows how we parse HTML with this module and configure it to emit parse errors except for duplicate attributes. Then we transform HTML to markdown with rehype-remark and finally compile that markdown with remark-stringify.

Say we have the following file, example.html, with a few errors:

<title class="a" class="b">Hello…</title>

And our script, example.js, looks as follows:

var vfile = require('to-vfile')
var report = require('vfile-reporter')
var unified = require('unified')
var parse = require('rehype-parse')
var rehype2remark = require('rehype-remark')
var stringify = require('remark-stringify')

  .use(parse, {emitParseErrors: true, duplicateAttribute: false})
  .process(vfile.readSync('example.html'), function(err, file) {
    console.error(report(err || file))

Now, running node example yields:

  1:10-1:10  warning  Missing whitespace before doctype name                      missing-whitespace-before-doctype-name                 parse-error
    3:1-3:6  warning  Unexpected trailing slash on start tag of non-void element  non-void-html-element-start-tag-with-trailing-solidus  parse-error

⚠ 2 warnings
# World!


processor.use(parse[, options])

Configure the processor to read HTML as input and process a hast syntax tree.


Specify whether to parse a fragment (boolean, default: false), instead of a complete document. In document mode, unopened html, head, and body elements are opened in just the right places.

⚠️ rehype is not an XML parser. It support SVG as embedded in HTML, but not the features available in the rest of XML/SVG. Passing SVG files could strip useful information, but fragments of modern SVG should be fine.

Whether the document is in the 'html' or 'svg' space ('svg' or 'html', default: 'html').

If an svg element is found in the HTML space, toHTML automatically switches to the SVG space when entering the element, and switches back when leaving.

Note: make sure to set fragment: true if space: 'svg'.


⚠️ Parse errors are currently being added to HTML. Not all errors emitted by parse5 (or rehype-parse) are specced yet. Some documentation may still be missing.

Emit parse errors while parsing on the vfile (boolean, default: false).

Setting this to true starts emitting HTML parse errors.

Specific rules can be turned off by setting them to false (or 0). The default, when emitParseErrors: true, is true (or 1), and means that rules emit as warnings. Rules can also be configured with 2, to turn them into fatal errors.

The specific parse errors that are currently supported are detailed below:


Patch extra positional information (boolean, default: false). If specified, the following element:

<img src="#" alt>

...has the following data:

{ position:
   { opening:
      { start: { line: 1, column: 1, offset: 0 },
        end: { line: 1, column: 18, offset: 17 } },
     closing: null,
      { src:
         { start: { line: 1, column: 6, offset: 5 },
           end: { line: 1, column: 13, offset: 12 } },
         { start: { line: 1, column: 14, offset: 13 },
           end: { line: 1, column: 17, offset: 16 } } } } }


Access to the parser, if you need it.


MIT © Titus Wormer