sentencex

1.0.23 verified Fri May 01 auth: no python

A sentence segmentation library supporting ~300 languages, powered by Rust for performance. Version 1.0.23, MIT license, maintained by Wikimedia. Releases are frequent with multilingual improvements.

pip install sentencex

Common errors

error TypeError: 'generator' object is not subscriptable ↓

cause segment returns a generator, not a list.

fix

sentences = list(segment('en', text))

error ValueError: Unsupported language code: 'English' ↓

cause Language code must be ISO 639-1 two-letter code, not language name.

fix

Use 'en' instead of 'English'.

Warnings

gotcha The segment function returns a generator, not a list. Must be consumed (e.g., list()) to see all sentences. ↓

fix Wrap with list() or iterate.

gotcha Language code must be lowercase (e.g., 'en', 'zh') but not all ISO codes are supported. Check support list. ↓

fix Use 'en' for English; see README for supported codes.

Imports

segment

wrong

import sentencex
sentencex.segment()

correct

from sentencex import segment

Function is in module, not as attribute of package.

Quickstart

Basic usage: segment English text into sentences.

from sentencex import segment

text = "Hello world! This is a test. And another sentence."
sentences = segment('en', text)
print(list(sentences))