Skip to content

Latest commit

 

History

History
86 lines (54 loc) · 3.73 KB

README.md

File metadata and controls

86 lines (54 loc) · 3.73 KB

Espells: Hunspell ported to Python ported to JavaScript

Pure JS/TS spellchecker, using Hunspell dictionaries. Direct port of the Spylls library. Without zverok's (the author of Spylls) work, this library couldn't exist.

Espells makes no use of platform specific features that would prevent it from running within Node, a browser, or even a web worker. Effectively, it's just a pure spellchecking library and it's up to you to connect it to whatever interface you want.

Espells was created as part of my work on Wikijump.

Why?

Two main reasons:

  1. You can't access the browser's spellchecking functionality in JS.
  2. There wasn't a fully featured spellchecker for clientside JS.

To elaborate on that second point: libraries using Hunspell, or are compatible with Hunspell dictionaries, do exist. However, they're either incapable of handling many dictionaries, aren't fully featured, can only run in Node, or bastardize Hunspell to run in WASM, with mixed results. Espells doesn't have those problems.

Features

Espells supports everything that Spylls support, which supports almost everything that Hunspell does. It reads the .dic and .aff formats that Hunspell expects. It can handle very complex languages, like the usual Italian .aff, without crashing or using excessive memory.

Installation

npm install espells

You can get dictionaries very easily from NPM as well, see here for that.

Usage

Usage of Espells is incredibly simple. Its primary interface is the Espells class exported by the module, which can be instantiated like so:

import { Espells } from "espells"
// the .aff and .dic should be given to Espells as a string or Uint8Array
// this special import just represents getting this data
import { aff, dic, dic2 } from "your-dictionary"

const spellchecker = new Espells({ aff, dic })

// alternatively, you can use a special async method to instantiate from URLs
const urlSpellchecker = await Espells.fromURL({
  aff: "someurl.aff",
  dic: "someurl.dic"
})

// finally, Espells supports loading multiple dictionaries at once
// this can be done with either the fromURL method or the normal constructor
const multipleSpellchecker = new Espells({ aff, dic: [dic, dic2] })

Espells is ready to use immediately. You can spellcheck a word like so:

const { correct, forbidden, warn } = spellchecker.lookup("word")

The forbidden and warn properties are special properties a "correctly spelled" word may have in Hunspell. Forbidden words are usually correct syntax wise, but aren't really considered real words, like "decreated". A word with warn: true is technically a correctly spelled word, but usage of that specific word is usually a mistake.

Getting suggestions is simple as well:

// an array of strings
const suggestions = spellchecker.suggest("wodr")

There are quite a few other methods made available, specifically: addDictionary, add, remove, stems, and data. These are fully documented in detailed tsdoc comments.

Contributing

Espells is welcome to PRs, issues, etc. Any contributions must maintain "parity" with Hunspell, although there is obviously some nuance to that idea.

If you want to understand how Espells work, the source files are fully documented. Additionally, you can take a look at Spylls' documentation for additional perspective and reasoning about certain technical details.

Building

Espells is built using TypeScript, with no special build tools. You can use the following command:

npm run build

License

MPL 2.0. See the license file for more details.