node-html-parser is both an executable binary that can be run, and a library that can be used in Rust programs.

Installing the command-line executable

Assuming you have Rust/Cargo installed, run this command in a terminal:

cargo install node-html-parser

It will make the node-html-parser command available in your PATH if you've allowed the PATH to be modified when installing Rust. cargo uninstall node-html-parser uninstalls.

Adding `node_html_parser` library as a dependency

Run this command in a terminal, in your project's directory:

cargo add node-html-parser

To add it manually, edit your project's Cargo.toml file and add to the [dependencies] section:

node-html-parser = "0.1.0"

The node_html_parser library will be automatically available globally. Read the node_html_parser library documentation.

Back to the crate overview.

Readme

Fast HTML Parser

Fast HTML Parser is a very fast HTML parser. Which will generate a simplified DOM tree, with element query support.

Per the design, it intends to parse massive HTML files in lowest price, thus the performance is the top priority. For this reason, some malformatted HTML may not be able to parse correctly, but most usual errors are covered (eg. HTML4 style no closing <td> etc).

Install

cargo add node-html-parser

Note: when using Fast HTML Parser in a Typescript project the minimum Typescript version supported is ^4.1.2.

Performance

-- 2022-08-10

html-parser     :24.1595 ms/file ± 18.7667
htmljs-parser   :4.72064 ms/file ± 5.67689
html-dom-parser :2.18055 ms/file ± 2.96136
html5parser     :1.69639 ms/file ± 2.17111
cheerio         :12.2122 ms/file ± 8.10916
parse5          :6.50626 ms/file ± 4.02352
htmlparser2     :2.38179 ms/file ± 3.42389
htmlparser      :17.4820 ms/file ± 128.041
high5           :3.95188 ms/file ± 2.52313
node-html-parser:2.04288 ms/file ± 1.25203
node-html-parser (last release):2.00527 ms/file ± 1.21317

Tested with htmlparser-benchmark.

Usage

import { parse } from 'node-html-parser';

const root = parse('<ul id="list"><li>Hello World</li></ul>');

console.log(root.firstChild.structure);
// ul#list
//   li
//     #text

console.log(root.querySelector('#list'));
// { tagName: 'ul',
//   rawAttrs: 'id="list"',
//   childNodes:
//    [ { tagName: 'li',
//        rawAttrs: '',
//        childNodes: [Object],
//        classNames: [] } ],
//   id: 'list',
//   classNames: [] }
console.log(root.toString());
// <ul id="list"><li>Hello World</li></ul>
root.set_content('<li>Hello World</li>');
root.toString();	// <li>Hello World</li>

var HTMLParser = require('node-html-parser');

var root = HTMLParser.parse('<ul id="list"><li>Hello World</li></ul>');

Global Methods

parse(data[, options])

Parse the data provided, and return the root of the generated DOM.

data, data to parse

options, parse options

{
  lowerCaseTagName: false,		// convert tag name to lower case (hurts performance heavily)
  comment: false,           		// retrieve comments (hurts performance slightly)
  fixNestedATags: false,    		// fix invalid nested <a> HTML tags 
  parseNoneClosedTags: false, 	// close none closed HTML tags instead of removing them 
  voidTag: {
    tags: ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr'],	// optional and case insensitive, default value is ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr']
    closingSlash: true	// optional, default false. void tag serialisation, add a final slash <br/>
  },
  blockTextElements: {
    script: true,		// keep text content when parsing
    noscript: true,		// keep text content when parsing
    style: true,		// keep text content when parsing
    pre: true			// keep text content when parsing
  }
}

valid(data[, options])

Parse the data provided, return true if the given data is valid, and return false if not.

Class

classDiagram
direction TB
class HTMLElement{
	this trimRight()
	this removeWhitespace()
	Node[] querySelectorAll(string selector)
	Node querySelector(string selector)
	HTMLElement[] getElementsByTagName(string tagName)
	Node closest(string selector)
	Node appendChild(Node node)
	this insertAdjacentHTML('beforebegin' | 'afterbegin' | 'beforeend' | 'afterend' where, string html)
	this setAttribute(string key, string value)
	this setAttributes(Record string, string attrs)
	this removeAttribute(string key)
	string getAttribute(string key)
	this exchangeChild(Node oldNode, Node newNode)
	this removeChild(Node node)
	string toString()
	this set_content(string content)
	this set_content(Node content)
	this set_content(Node[] content)
	this remove()
	this replaceWith((string | Node)[] ...nodes)
	ClassList classList
	HTMLElement clone()
	HTMLElement getElementById(string id)
	string text
	string rawText
	string tagName
	string structuredText
	string structure
	Node firstChild
	Node lastChild
	Node nextSibling
	HTMLElement nextElementSibling
	Node previousSibling
	HTMLElement previousElementSibling
	string innerHTML
	string outerHTML
	string textContent
	Record<string, string> attributes
	[number, number] range
}
class Node{
	<<abstract>>
	string toString()
	Node clone()
	this remove()
	number nodeType
	string innerText
	string textContent
}
class ClassList{
	add(string c)
	replace(string c1, string c2)
	remove(string c)
	toggle(string c)
	boolean contains(string c)
	number length
	string[] value
	string toString()
}
class CommentNode{
	CommentNode clone()
	string toString()
}
class TextNode{
	TextNode clone()
	string toString()
	string rawText
	string trimmedRawText
	string trimmedText
	string text
	boolean isWhitespace
}
Node --|> HTMLElement
Node --|> CommentNode
Node --|> TextNode
Node ..> ClassList

HTMLElement Methods

trimRight()

Trim element from right (in block) after seeing pattern in a TextNode.

removeWhitespace()

Remove whitespaces in this sub tree.

querySelectorAll(selector)

Query CSS selector to find matching nodes.

Note: Full range of CSS3 selectors supported since v3.0.0.

querySelector(selector)

Query CSS Selector to find matching node. null if not found.

getElementsByTagName(tagName)

Get all elements with the specified tagName.

Note: Use * for all elements.

closest(selector)

Query closest element by css selector. null if not found.

before(...nodesOrStrings)

Insert one or multiple nodes or text before the current element. Does not work on root.

after(...nodesOrStrings)

Insert one or multiple nodes or text after the current element. Does not work on root.

prepend(...nodesOrStrings)

Insert one or multiple nodes or text to the first position of an element's child nodes.

append(...nodesOrStrings)

Insert one or multiple nodes or text to the last position of an element's child nodes. This is similar to appendChild, but accepts arbitrarily many nodes and converts strings to text nodes.

appendChild(node)

Append a node to an element's child nodes.

insertAdjacentHTML(where, html)

Parses the specified text as HTML and inserts the resulting nodes into the DOM tree at a specified position.

setAttribute(key: string, value: string)

Set value to key attribute.

setAttributes(attrs: Record<string, string>)

Set attributes of the element.

removeAttribute(key: string)

Remove key attribute.

getAttribute(key: string)

Get key attribute. undefined if not set.

exchangeChild(oldNode: Node, newNode: Node)

Exchanges given child with new child.

removeChild(node: Node)

Remove child node.

toString()

Same as outerHTML

set_content(content: string | Node | Node[])

Set content. Notice: Do not set content of the root node.

remove()

Remove current element.

replaceWith(...nodes: (string | Node)[])

Replace current element with other node(s).

classList

classList.add

Add class name.

classList.replace(old: string, new: string)

Replace class name with another one.

classList.remove()

Remove class name.

classList.toggle(className: string):void

Toggle class. Remove it if it is already included, otherwise add.

classList.contains(className: string): boolean

Returns true if the classname is already in the classList.

classList.value

Get class names.

clone()

Clone a node.

getElementById(id: string): HTMLElement | null

Get element by it's ID.

HTMLElement Properties

text

Get unescaped text value of current node and its children. Like innerText. (slow for the first time)

rawText

Get escaped (as-is) text value of current node and its children. May have & in it. (fast)

tagName

Get or Set tag name of HTMLElement. Note that the returned value is an uppercase string.

structuredText

Get structured Text.

structure

Get DOM structure.

childNodes

Get all child nodes. A child node can be a TextNode, a CommentNode and a HTMLElement.

children

Get all child elements, so all child nodes of type HTMLELement.

firstChild

Get first child node. undefined if the node has no children.

lastChild

Get last child node. undefined if the node has no children.

firstElementChild

Get the first child of type HTMLElement. undefined if none exists.

lastElementChild

Get the first child of type HTMLElement. undefined if none exists.

childElementCount

Get the number of children that are of type HTMLElement.

innerHTML

Set or Get innerHTML.

outerHTML

Get outerHTML.

nextSibling

Returns a reference to the next child node of the current element's parent. null if not found.

nextElementSibling

Returns a reference to the next child element of the current element's parent. null if not found.

previousSibling

Returns a reference to the previous child node of the current element's parent. null if not found.

previousElementSibling

Returns a reference to the previous child element of the current element's parent. null if not found.

textContent

Get or Set textContent of current element, more efficient than set_content.

attributes

Get all attributes of current element. Notice: do not try to change the returned value.

range

Corresponding source code start and end indexes (ie [ 0, 40 ])

WASM / 浏览器 & Node.js 使用

本仓库同时支持：

作为普通 Rust 库 (native)
编译为 WebAssembly，在浏览器或 Node.js 中调用

构建前准备

安装 wasm 目标

rustup target add wasm32-unknown-unknown

（可选）安装 wasm-pack
```
cargo install wasm-pack
```

Feature 说明

Feature	作用	默认	备注
parallel	启用 rayon 并行	启用	wasm 线程需额外配置，建议 wasm 下关闭
wasm	启用 wasm-bindgen + panic hook	关闭	wasm 构建时启用

在 wasm 场景下为避免 rayon 线程问题，建议使用：--no-default-features --features wasm。

使用 wasm-pack 构建（推荐）

wasm-pack build --release --target bundler --no-default-features --features wasm
# 或 --target web / nodejs / deno 视使用环境而定

输出目录 pkg/ 中包含：

node_html_parser_bg.wasm
对应 JS 绑定与类型声明

浏览器示例

<script type="module">
import init, { hello } from './pkg/node_html_parser.js';
await init();
console.log(hello('Browser'));
</script>

Node.js 示例 (ESM)

import init, { hello } from './pkg/node_html_parser.js';
await init();
console.log(hello('Node'));

直接使用 cargo 构建 (无 JS 包装层)

cargo build --release --target wasm32-unknown-unknown --no-default-features --features wasm

你需要自己加载 .wasm 并处理导出（wasm-bindgen 生成包装时不需要）。

Panic 与调试

启用 wasm feature 后自动安装 console_error_panic_hook，可在浏览器控制台看到 Rust panic 堆栈。

并行策略

total_len 等内部函数在启用 parallel 时使用 rayon 并行；在 wasm 构建关闭 parallel 时会自动退化为串行。

未来计划（可选）

导出完整 HTML Parser API 到 wasm
提供更细粒度的 DOM 访问接口
带 wasm 线程支持（SharedArrayBuffer 环境）

Installing the command-line executable

Adding node_html_parser library as a dependency

Adding `node_html_parser` library as a dependency