Getting started

Ferrolex checks words against dictionaries and returns deterministic suggestions. It is pre-1.0: public APIs can change in minor releases, and Hunspell support has explicit compatibility boundaries. Start with the Rust library or command-line tool.

Install the command-line tool

cargo install --locked ferrolex-cli

Create a UTF-8 file named words.txt, with one word per line:

ferrolex
Ferris

Check a word or request suggestions:

ferrolex check --dictionary words.txt ferrolex
ferrolex suggest --dictionary words.txt ferolex

The engine and CLI reference covers plain text, Hunspell sources and compiled artifacts.

Use the Rust library

cargo add ferrolex

The Rust quick start shows dictionary import and suggestions. The published API documentation is available on docs.rs.

Choose dictionary data explicitly

Ferrolex does not bundle dictionary data. The reviewed catalog downloads pinned sources into a cache you choose; ordinary checks never download or update data implicitly. Follow the dictionary acquisition guide and check the locale compatibility evidence for the exact source pair you intend to use.

Integrate at the word boundary

Your application extracts words from its document or language format. Ferrolex owns dictionary import, recognition and suggestions. The product-boundary decision explains why document parsing stays with format-aware tools.

The Node.js binding is implemented but not yet published to npm. C ABI, Python, LSP and editor work remain prototypes; see the support tiers.