What is betterkhmer?

BetterKhmer is a text normalization library for the Khmer language, used in Cambodia, ported to 18 programming languages. Its job is to ensure that Khmer text which looks visually identical is also encoded identically, so that searching, storage, and security checks produce consistent results. The core problem it solves is that Khmer syllables are made up of consonants and vowel marks that can be arranged in multiple different orders in a text file yet render identically on screen. The same written word can exist as several distinct byte sequences. This creates real world issues: a search engine can return different results for visually identical queries typed in different apps, domain names that look the same can route to different servers, and variable names in code that appear identical can differ in encoding, making malicious substitutions invisible to code reviewers. The library provides a single normalize function in every language port. Calling it on any Khmer string produces a canonical form regardless of how the original was encoded. It supports both modern Khmer and Middle Khmer. Rather than a package installed from a registry, each language port is a single self-contained source file that you copy directly into your project. The ports are all verified against a shared test suite of 10,085 input and output pairs sampled from real Khmer text. The README includes a benchmark showing throughput across all 18 ports. The library is based on a Khmer encoding structure proposed to Unicode and is released under the MIT license.