日本語

kuzukiri

A simple text segmenter

What's this?

This is a python library for text segmentation of Japanese text.

Features

• Text segmentation by simple rules,
  • rule-based, no machine learning,
  • so you can assume results.
• comparably fast. It's written in rust-lang.

Install

from PyPI

pip install kuzukiri

from source code

pip install setuptools-rust
python setup.py install

Usage

import kuzukiri

segmenter = kuzukiri.Segmenter()
text = "これはテストです。文分割します。"
sentences = segmenter.split(text)
print(sentences)  # => ['これはテストです。', '文分割します。']

For details, see examples and tests directories.

License

MIT

Dependencies

• PyO3 : to compile rust code for python.
• unicode_normalization crate : for NFKC normalization