langdetect-zh 1.0.4

Google's langdetect modified for Chinese texts

langdetect_zh

Installation

$ pip install langdetect_zh

Supported Python versions 2.7, 3.4+.

Languages

langdetect_zh supports 2 languages out of the box (ISO 639-1 codes):

zh-cn, zh-tw

Basic usage

Directly output the most similar language code：

>>> from langdetect_zh import detect
>>> detect("这是一段中文文本")
'zh-cn'

To find out the probabilities for the top languages:

>>> from langdetect_zh import detect_langs
>>> detect_langs("这是一段中文文本")
[zh-cn:0.999997316441747]

NOTE

Language detection algorithm is non-deterministic, which means that if you try to run it on a text which is either too short or too ambiguous, you might get different results everytime you run it.

To enforce consistent results, call following code before the first language detection:

from langdetect_zh import DetectorFactory
DetectorFactory.seed = 0

Original project

This package is an optimization of langdetect. The specific optimization measure is to subdivide simplified Chinese and traditional Chinese under the condition of pure Chinese.