pyokaka

Simple tool for converting [Roma-ji]¹ sentences into [Hiragana]² one.
ローマ字の文を平仮名に変換するシンプルなツールです。

The origin of package name

An homage to pykakashi that provides highly function to convert Kana-Kanji into Roma-ji. Okaka(おかか) is easy Japanese word that mean bonito flakes.

Demo

As command line tool
Use as REPL just by calling from terminal. To quit, send EOF.

$ python -m pyokaka.okaka

Roman >>> ohayougozaimasu
JKana ... おはようございます
Roman >>> kon'nichiwa
JKana ... こんにちわ
Roman >>> oyasuminasai
JKana ... おやすみなさい

You can indicate file you want to convert.

$ cat sample.txt
Ima wa mukashi, taketori no okina to iu mono ari keri.

$ python -m pyokaka.okaka sample.txt
いま わ むかし, たけとり の おきな と いう もの あり けり.

To apply additional rule, load utf-8 encoded json file.

{
    "ら": ["la"], "り": ["li"], "る": ["lu"], "れ": ["le"], "ろ": ["lo"],
    "ふぁ": ["pha", "hua"], "ふぃ": ["phi"]
}

$ cat sample.txt
elephant
lalallalalla

$ python -m pyokaka.okaka sample.txt
えlえpはんt
lあlあllあlあllあ

$ python -m pyokaka.okaka sample.txt --load sample.json
load for sample.json...
えれふぁんt
ららっららっら

For more information, view python -m pyokaka.okaka --help.

As library

>>> from pyokaka import okaka
>>> okaka.convert('katsuobushi')
'かつおぶし'

You can add more vocabulary as described below.

>>> okaka.convert('philipps')
'pひlいpps'
>>>
>>> okaka.update_convert_dct({
...     'p': 'ぷ', 's': 'す'
... })
>>>
>>> okaka.convert('philips')
'ぷひlいぷす'
>>>
>>> import json
>>> with open('sample.json', encoding='utf-8') as fin:
...     table = json.load(fin)
...
>>> okaka.update_transtable(table)
>>> okaka.convert('philips')
'ふぃりぷす'

Notes

• You cannot reset convert table without restart.

• Though converter ignores letter what can be not interpret as a part of Roma-ji, remaining letters always be converted.

  $ python -m pyokaka.okaka
  Roman >>> Oh dear, this is English!
  JKana ... おh であr, tひs いs えんglいsh!
  

• Hyphen always be replaced with Cho'onpu.

  $ python -m pyokaka.okaka
  Roman >>> Roma-ji
  JKana ... ろまーじ
  

• Converter never analyze sentence structure. So it cannot recognize 'wa', 'o' and 'e' as postpositional particle.

  $ python -m pyokaka.okaka
  Roman >>> Watashi wa depa-to e enpitsu o kai ni ikimashita.
  JKana ... わたし わ でぱーと え えんぴつ お かい に いきました.
  

• Conversion is based on greedy algorithm. Single quote can be used as separater if you need.

  Roman >>> honya
  JKana ... ほにゃ
  Roman >>> honnya
  JKana ... ほっにゃ
  Roman >>> honnnya
  JKana ... ほんにゃ
  
  Roman >>> hon'ya
  JKana ... ほんや
  

Install

This module is registered at PyPI. PyPI - pyokaka

$ pip install pyokaka

License

MIT

Author

LouiS0616