node-stemmer

What is Node Stemmer?

Node Stemmer is a Node.js interface to the stemming algorithms from the Snowball project, largely inspired by Richard Boulton's PyStemmer. It uses ffi-rs and expects to find the file libstemmer.so (a version of Libstemmer compiled as shared library) in LD_LIBRARY_PATH.
In order to set-up this kind of environment you can take a look at docker-node-libstemmer Dockerfile or you can use the corresponding docker image: amaccis/node-libstemmer

Installation

You can install Node Stemmer using npm.

npm i node-stemmer

Compatibility

node-stemmer	libstemmer
3.0.2	3.0.0, 3.0.1
1.0.0	2.0.0, 2.1.0, 2.2.0

Usage

import { Stemmer, CharacterEncoding } from 'node-stemmer';

const algorithms = Stemmer.algorithms();
console.log(algorithms);
/*
[
    'arabic',       'armenian',   'basque',
    'catalan',      'danish',     'dutch',
    'dutch_porter', 'english',    'esperanto',
    'estonian',     'finnish',    'french',
    'german',       'greek',      'hindi',
    'hungarian',    'indonesian', 'irish',
    'italian',      'lithuanian', 'nepali',
    'norwegian',    'porter',     'portuguese',
    'romanian',     'russian',    'serbian',
    'spanish',      'swedish',    'tamil',
    'turkish',      'yiddish'
]
*/

const algorithm = 'english';
const word = Buffer.from('cycling');
const stemmer = new Stemmer(algorithm); // default character encoding is UTF-8
const stem = stemmer.stemWord(word);
console.log(stem);
/*
cycl
*/

const algorithm = 'basque';
const word = Buffer.from('aberatsenetakoa');
const stemmer = new Stemmer(algorithm, CharacterEncoding.ISO_8859_1);
const stem = stemmer.stemWord(word);
console.log(stem);
/*
aberatse
*/

License

All files are MIT © Andrea Maccis.