Metadata-Version: 2.4
Name: mlnormalize
Version: 0.4.0
Classifier: Programming Language :: Rust
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Text Processing :: Linguistic
Classifier: License :: OSI Approved :: Apache Software License
Summary: Malayalam Unicode normalizer: bit-identical normalization for training and inference
Keywords: malayalam,unicode,normalization,nlp,tokenizer
Author-email: Jayashankar R <56070307+jayashankarvr@users.noreply.github.com>
License: Apache-2.0
Requires-Python: >=3.8
Description-Content-Type: text/markdown; charset=UTF-8; variant=GFM
Project-URL: Homepage, https://github.com/jayashankarvr/mlnormalize
Project-URL: Issues, https://github.com/jayashankarvr/mlnormalize/issues

# mlnormalize (Python wheel)

PyO3 bindings for the [`mlnormalize`](https://github.com/jayashankarvr/mlnormalize) Malayalam
Unicode normalizer. The wheel is a thin wrapper around the Rust core, so Python gets
byte-identical output. That identity is the point of the normalizer contract: one
implementation, two callers.

## Install

```sh
pip install mlnormalize
```

Also available as a Rust crate on crates.io (same byte-identical output): `cargo add mlnormalize`.

```python
import mlnormalize
mlnormalize.normalize("ൻ്റ")     # -> "ന്റ"  (byte-identical to the Rust core)
mlnormalize.stripped_key(text)    # -> str    (dedup/matching key; NOT training text)
mlnormalize.version()             # -> (0, 4, 0)
mlnormalize.__version__           # -> "0.4.0"
```

## Build

Built with [maturin](https://github.com/PyO3/maturin) from this directory:

```sh
maturin build --release     # produces a wheel in target/wheels/
maturin develop             # install into the active venv for development
```

