language-detection

Why we included this project

PHP teams that need to know what language a chunk of text is written in have a small, honest option here. The library builds N-gram profiles from its bundled training samples in 110 languages, then scores an input string against those profiles and returns a guess with a confidence value. When the default corpus falls short for specialized vocabulary, a training mode lets you feed your own text samples, which fixes cases where generic models get confused. It installs as a Composer package with a plain API, so wiring it into moderation queues, search indexing, or front-end localization is an afternoon task. One catch: it needs the mbstring extension, so check that before committing.