lnx

An adaptable search engine API.

Make it
>
lightning fast. relevant. typo tolerant. search as you type. your way.

An open-source, lightning-fast and typo tolerant search engine empowering developers and users alike.

Community Driven

100% open-source, built for the community.

Insanely Fast

Query millions of documents in milliseconds with our fast-fuzzy system.

Typo Tolerant

lnx provides fuzzy query systems to give you a typo-tolerant query.

Multi-Language

Currently lnx supports 17 latin languages. Chinese, Japanese and Korean are planned.

Quick start with docker

Get started in seconds with our docker images , just run:

docker run -p "8000:8000" chillfish8/lnx:latest --host "0.0.0.0"

Once setup you can create indexes and start exploring everything
lnx has to offer with the docs.

The query kinds

The "normal" query

This mode exposes the Tantivy query parser which gives you a powerful query language for searching things like logs.

The "fuzzy" query

This mode uses the traditional Levenshtein distance to find close matches giving you typo-tolerance with your queries.

The "more-like-this" query

Get similar documents to a reference document, great for e-commerce or e-readers.

The "fast-fuzzy" query

The fast-fuzzy mode is a implementation of the symspell algorithm which provides spell corrections via pre-computation. This is the secret sauce behind being able to blow other engines out of the water when it comes to fuzzy searching.

Sorting By Field

Results can be sorted by any fast-field quickly and easily.

Stop Words

With a pre-made set, stop words can be removed to boost relevancy in queries.

Token Bearer Authorization

Secure your system giving permissions based access to specific users.

Boost Fields

Each fields can be individually adjusted to give the best searching experience.

Fast-Fuzzy Pre-Computation

This feature uses pre-computation to boost fuzzy query performance by over 10x.

© 2021 lnx search.

Docs FAQ Docker Images GitHub