unified

Project: syntax-tree/hast-util-to-nlcst

Package: hast-util-to-nlcst@3.0.0

  1. Dependents: 0
  2. hast utility to transform to nlcst
  1. util 145
  2. utility 141
  3. unist 132
  4. html 124
  5. rehype 88
  6. hast 74
  7. hast-util 46
  8. retext 42
  9. nlcst 15
  10. language 12
  11. natural 9

hast-util-to-nlcst

Build Coverage Downloads Size Sponsors Backers Chat

hast utility to transform to nlcst.

Contents

What is this?

This package is a utility that takes a hast (HTML) syntax tree as input and turns it into nlcst (natural language).

When should I use this?

This project is useful when you want to deal with ASTs and inspect the natural language inside HTML. Unfortunately, there is no way yet to apply changes to the nlcst back into hast.

The mdast utility mdast-util-to-nlcst does the same but uses a markdown tree as input.

The rehype plugin rehype-retext wraps this utility to do the same at a higher-level (easier) abstraction.

Install

This package is ESM only. In Node.js (version 16+), install with npm:

npm install hast-util-to-nlcst

In Deno with esm.sh:

import {toNlcst} from 'https://esm.sh/hast-util-to-nlcst@4'

In browsers with esm.sh:

<script type="module">
  import {toNlcst} from 'https://esm.sh/hast-util-to-nlcst@4?bundle'
</script>

Use

Say our document example.html contains:

<article>
  Implicit.
  <h1>Explicit: <strong>foo</strong>s-ball</h1>
  <pre><code class="language-foo">bar()</code></pre>
</article>

…and our module example.js looks as follows:

import {fromHtml} from 'hast-util-from-html'
import {toNlcst} from 'hast-util-to-nlcst'
import {ParseEnglish} from 'parse-english'
import {read} from 'to-vfile'
import {inspect} from 'unist-util-inspect'

const file = await read('example.html')
const tree = fromHtml(file)

console.log(inspect(toNlcst(tree, file, ParseEnglish)))

…now running node example.js yields (positional info removed for brevity):

RootNode[2] (1:1-6:1, 0-134)
├─0 ParagraphNode[3] (1:10-3:3, 9-24)
│   ├─0 WhiteSpaceNode "\n  " (1:10-2:3, 9-12)
│   ├─1 SentenceNode[2] (2:3-2:12, 12-21)
│   │   ├─0 WordNode[1] (2:3-2:11, 12-20)
│   │   │   └─0 TextNode "Implicit" (2:3-2:11, 12-20)
│   │   └─1 PunctuationNode "." (2:11-2:12, 20-21)
│   └─2 WhiteSpaceNode "\n  " (2:12-3:3, 21-24)
└─1 ParagraphNode[1] (3:7-3:43, 28-64)
    └─0 SentenceNode[4] (3:7-3:43, 28-64)
        ├─0 WordNode[1] (3:7-3:15, 28-36)
        │   └─0 TextNode "Explicit" (3:7-3:15, 28-36)
        ├─1 PunctuationNode ":" (3:15-3:16, 36-37)
        ├─2 WhiteSpaceNode " " (3:16-3:17, 37-38)
        └─3 WordNode[4] (3:25-3:43, 46-64)
            ├─0 TextNode "foo" (3:25-3:28, 46-49)
            ├─1 TextNode "s" (3:37-3:38, 58-59)
            ├─2 PunctuationNode "-" (3:38-3:39, 59-60)
            └─3 TextNode "ball" (3:39-3:43, 60-64)

API

This package exports the identifier toNlcst. There is no default export.

toNlcst(tree, file, Parser)

Turn a hast tree into an nlcst tree.

👉 Note: tree must have positional info and file must be a VFile corresponding to tree.

Parameters
Returns

NlcstNode.

Notes
Implied paragraphs

The algorithm supports implicit and explicit paragraphs, such as:

<article>
  An implicit paragraph.
  <h1>An explicit paragraph.</h1>
</article>

Overlapping paragraphs are also supported (see the tests or the HTML spec for more info).

Ignored nodes

Some elements are ignored and their content will not be present in nlcst: <script>, <style>, <svg>, <math>, <del>.

To ignore other elements, add a data-nlcst attribute with a value of ignore:

<p>This is <span data-nlcst="ignore">hidden</span>.</p>
<p data-nlcst="ignore">Completely hidden.</p>
Source nodes

<code> elements are mapped to Source nodes in nlcst.

To mark other elements as source, add a data-nlcst attribute with a value of source:

<p>This is <span data-nlcst="source">marked as source</span>.</p>
<p data-nlcst="source">Completely marked.</p>

ParserConstructor

Create a new parser (TypeScript type).

Type
type ParserConstructor = new () => ParserInstance

ParserInstance

nlcst parser (TypeScript type).

For example, parse-dutch, parse-english, or parse-latin.

Type
type ParserInstance = {
  parse(value?: string | null | undefined): NlcstRoot
  tokenize(value?: string | null | undefined): Array<NlcstSentenceContent>
  tokenizeParagraph(value?: string | null | undefined): NlcstParagraph
  tokenizeParagraphPlugins: Array<(node: NlcstParagraph) => undefined | void>
  tokenizeSentencePlugins: Array<(node: NlcstSentence) => undefined | void>
}

Types

This package is fully typed with TypeScript. It exports the additional types ParserConstructor and ParserInstance.

Compatibility

Projects maintained by the unified collective are compatible with maintained versions of Node.js.

When we cut a new major release, we drop support for unmaintained versions of Node. This means we try to keep the current release line, hast-util-to-nlcst@^4, compatible with Node.js 16.

Security

hast-util-to-nlcst does not change the original syntax tree so there are no openings for cross-site scripting (XSS) attacks.

Contribute

See contributing.md in syntax-tree/.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