html2json

A Rust port of
cheerio-json-mapper.

---

Overview

- Input: HTML source + Extractor spec (JSON)
- Output: JSON matching the structure defined in the spec
- Available as: Rust crate, CLI tool, and WebAssembly npm package

Installation

$3

``bash
npm install @qretaio/html2json
`

$3

`bash
cargo install html2json --features cli
`

$3

`bash
cargo install --path . --features cli


or from a git repository

`bash
just install
`

Usage

$3

`javascript
import { extract } from "@qretaio/html2json";

const html =

const result = await extract(html, spec);
console.log(result);
// {
// "title": "My Article",
// "author": "John Doe",
// "tags": [{"name": "rust"}, {"name": "wasm"}]
// }
`

$3

`bash


Extract from file

- --spec, -s - Path to JSON extractor spec file (required)
- --check, -c - Compare output against expected JSON file. Exits with 0 if match, 1 if differ (with colored diff).

Spec Format

The spec is a JSON object where each key defines an output field and each value defines a CSS selector to extract that field.

$3

`json
{
"title": "h1",
"description": "p.description"
}
`

$3

`json
{
"link": "a.main | attr:href",
"image": "img.hero | attr:src"
}
`

$3

`json
{
"title": "h1 | trim",
"slug": "h1 | lower | regex:\\s+-",
"price": ".price | regex:\\$(\\d+\\.\\d+) | parseAs:int"
}
`

Available pipes:

- trim - Trim whitespace
- lower - Convert to lowercase
- upper - Convert to uppercase
- substr:start:end - Extract substring
- regex:pattern - Regex capture (first group)
- parseAs:int - Parse as integer
- parseAs:float - Parse as float
- attr:name - Get attribute value
- void - Extract from void elements, useful for extracting xml

$3

`json
{
"items": [
{
"$": ".item",
"title": "h2",
"description": "p"
}
]
}
`

$3

`json
{
"$": "article",
"title": "h1",
"paragraphs": ["p"]
}
`

$3

`json
{
"title": "h1.main || h1.fallback || h1"
}
`

$3

`json
{
"title": "h1",
"description?": "p.description"
}
`

Optional fields that return null` are removed from the output.

LICENSE

MIT