rehype-parse

What is rehype-parse?

rehype-parse is a plugin for the unified processor that parses HTML into a syntax tree. It is part of the rehype ecosystem, which is a toolset for transforming HTML with plugins. This package is particularly useful for processing and manipulating HTML content programmatically.

What are rehype-parse's main functionalities?

Basic HTML Parsing

This feature allows you to parse a simple HTML string into a syntax tree. The code sample demonstrates how to parse an HTML string and inspect the resulting tree structure.

const unified = require('unified');
const parse = require('rehype-parse');
const inspect = require('unist-util-inspect');

const html = '<h1>Hello, world!</h1>';
const tree = unified()
  .use(parse)
  .parse(html);

console.log(inspect(tree));

Parsing with Options

This feature allows you to parse HTML with specific options. In this example, the 'fragment' option is set to true, which allows parsing of HTML fragments instead of full documents.

const unified = require('unified');
const parse = require('rehype-parse');
const inspect = require('unist-util-inspect');

const html = '<h1>Hello, world!</h1>';
const tree = unified()
  .use(parse, { fragment: true })
  .parse(html);

console.log(inspect(tree));

Integration with Other Plugins

This feature demonstrates how to integrate rehype-parse with other plugins in the rehype ecosystem. The code sample shows how to parse HTML and then stringify it back to HTML.

const unified = require('unified');
const parse = require('rehype-parse');
const stringify = require('rehype-stringify');

const html = '<h1>Hello, world!</h1>';
const output = unified()
  .use(parse)
  .use(stringify)
  .processSync(html)
  .toString();

console.log(output);

Other packages similar to rehype-parse

htmlparser2

htmlparser2 is a fast and forgiving HTML/XML parser. It is more low-level compared to rehype-parse and provides a SAX-style parser, which can be more complex to use but offers more control over the parsing process.

parse5

parse5 is a highly compliant HTML parser that closely follows the WHATWG HTML specification. It is similar to rehype-parse in terms of compliance and ease of use but is a standalone parser without the plugin ecosystem that rehype offers.

jsdom

jsdom is a JavaScript implementation of the DOM and HTML standards. It is more heavyweight compared to rehype-parse and is used for simulating a browser environment, making it suitable for more complex DOM manipulations and testing.

README

rehype-parse

Parser for unified. Parses HTML to a HAST syntax tree. Used in the rehype processor.

Installation

npm:

npm install rehype-parse

Usage

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:

<!doctypehtml>
<title class="a" class="b">Hello…</title>
<h1/>World!</h1>

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')

unified()
  .use(parse, {emitParseErrors: true, duplicateAttribute: false})
  .use(rehype2remark)
  .use(stringify)
  .process(vfile.readSync('example.html'), function(err, file) {
    console.error(report(err || file))
    console.log(String(file))
  })

Now, running node example yields:

example.html
  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!

API

processor.use(parse[, options])

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

options

options.fragment

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.

options.space

⚠️ 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'.

options.emitParseErrors

⚠️ 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:

• abandonedHeadElementChild — unexpected metadata element after head (example)
• abruptClosingOfEmptyComment — unexpected abruptly closed empty comment (example)
• abruptDoctypePublicIdentifier — unexpected abruptly closed public identifier (example)
• abruptDoctypeSystemIdentifier — unexpected abruptly closed system identifier (example)
• absenceOfDigitsInNumericCharacterReference — unexpected non-digit at start of numeric character reference (example)
• cdataInHtmlContent — unexpected CDATA section in HTML (example)
• characterReferenceOutsideUnicodeRange — unexpected too big numeric character reference (example)
• closingOfElementWithOpenChildElements — unexpected closing tag with open child elements (example)
• controlCharacterInInputStream — unexpected control character (example)
• controlCharacterReference — unexpected control character reference (example)
• disallowedContentInNoscriptInHead — disallowed content inside `<noscript>` in `<head>` (example)
• duplicateAttribute — unexpected duplicate attribute (example)
• endTagWithAttributes — unexpected attribute on closing tag (example)
• endTagWithTrailingSolidus — unexpected slash at end of closing tag (example)
• endTagWithoutMatchingOpenElement — unexpected unopened end tag (example)
• eofBeforeTagName — unexpected end of file (example)
• eofInCdata — unexpected end of file in CDATA (example)
• eofInComment — unexpected end of file in comment (example)
• eofInDoctype — unexpected end of file in doctype (example)
• eofInElementThatCanContainOnlyText — unexpected end of file in element that can only contain text (example)
• eofInScriptHtmlCommentLikeText — unexpected end of file in comment inside script (example)
• eofInTag — unexpected end of file in tag (example)
• incorrectlyClosedComment — incorrectly closed comment (example)
• incorrectlyOpenedComment — incorrectly opened comment (example)
• invalidCharacterSequenceAfterDoctypeName — invalid sequence after doctype name (example)
• invalidFirstCharacterOfTagName — invalid first character in tag name (example)
• misplacedDoctype — misplaced doctype (example)
• misplacedStartTagForHeadElement — misplaced `<head>` start tag (example)
• missingAttributeValue — missing attribute value (example)
• missingDoctype — missing doctype before other content (example)
• missingDoctypeName — missing doctype name (example)
• missingDoctypePublicIdentifier — missing public identifier in doctype (example)
• missingDoctypeSystemIdentifier — missing system identifier in doctype (example)
• missingEndTagName — missing name in end tag (example)
• missingQuoteBeforeDoctypePublicIdentifier — missing quote before public identifier in doctype (example)
• missingQuoteBeforeDoctypeSystemIdentifier — missing quote before system identifier in doctype (example)
• missingSemicolonAfterCharacterReference — missing semicolon after character reference (example)
• missingWhitespaceAfterDoctypePublicKeyword — missing whitespace after public identifier in doctype (example)
• missingWhitespaceAfterDoctypeSystemKeyword — missing whitespace after system identifier in doctype (example)
• missingWhitespaceBeforeDoctypeName — missing whitespace before doctype name (example)
• missingWhitespaceBetweenAttributes — missing whitespace between attributes (example)
• missingWhitespaceBetweenDoctypePublicAndSystemIdentifiers — missing whitespace between public and system identifiers in doctype (example)
• nestedComment — unexpected nested comment (example)
• nestedNoscriptInHead — unexpected nested `<noscript>` in `<head>` (example)
• nonConformingDoctype — unexpected non-conforming doctype declaration (example)
• nonVoidHtmlElementStartTagWithTrailingSolidus — unexpected trailing slash on start tag of non-void element (example)
• noncharacterCharacterReference — unexpected noncharacter code point referenced by character reference (example)
• noncharacterInInputStream — unexpected noncharacter character (example)
• nullCharacterReference — unexpected NULL character referenced by character reference (example)
• openElementsLeftAfterEof — unexpected end of file (example)
• surrogateCharacterReference — unexpected surrogate character referenced by character reference (example)
• surrogateInInputStream — unexpected surrogate character
• unexpectedCharacterAfterDoctypeSystemIdentifier — invalid character after system identifier in doctype (example)
• unexpectedCharacterInAttributeName — unexpected character in attribute name (example)
• unexpectedCharacterInUnquotedAttributeValue — unexpected character in unquoted attribute value (example)
• unexpectedEqualsSignBeforeAttributeName — unexpected equals sign before attribute name (example)
• unexpectedNullCharacter — unexpected NULL character (example)
• unexpectedQuestionMarkInsteadOfTagName — unexpected question mark instead of tag name (example)
• unexpectedSolidusInTag — unexpected slash in tag (example)
• unknownNamedCharacterReference — unexpected unknown named character reference (example)

options.verbose

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,
     properties:
      { src:
         { start: { line: 1, column: 6, offset: 5 },
           end: { line: 1, column: 13, offset: 12 } },
        alt:
         { start: { line: 1, column: 14, offset: 13 },
           end: { line: 1, column: 17, offset: 16 } } } } }

parse.Parser

Access to the parser, if you need it.

License

MIT © Titus Wormer