YouCab: Converts MeCab Parsing Results to Python Objects

Installation

Install MeCab

MeCab is required for YouCab to work. If it is not already installed, install MeCab first.

Install YouCab

$ pip install youcab

Tokenize Japanese sentence

In this example code, we generate a tokenizer with MeCab's default dictionary and run tokenization. The tokenizer converts text into a list of Word objects.

from youcab import youcab

tokenize = youcab.generate_tokenizer()
words = tokenize("本を読んだ")
for word in words:
    print("surface: " + word.surface)
    print("pos    : " + str(word.pos))
    print("base   : " + word.base)
    print("c_type : " + word.c_type)
    print("c_form : " + word.c_form)
    print("")

surface: 本
pos    : ['名詞', '一般']
base   : 本
c_type : 
c_form : 

surface: を
pos    : ['助詞', '格助詞', '一般']
base   : を
c_type : 
c_form : 

surface: 読ん
pos    : ['動詞', '自立']
base   : 読む
c_type : 五段・マ行
c_form : 連用タ接続

surface: だ
pos    : ['助動詞']
base   : だ
c_type : 特殊・タ
c_form : 基本形

Available for any MeCab dictionary

Dictionaries such as IPAdic, UniDic and neologd are available.

from youcab import youcab

tokenize = youcab.generate_tokenizer(dicdir="/path/to/mecab/dic/dir/")