Skip to main content

Japanese Configuration

Follow along if you are a Japanese Learner.

Kagome (Japanese morphological analyzer)​


go install github.com/ikawaha/kagome/v2@latest

yomichan-jlpt-vocab (An Modified Japanese Tokenizer and Japanese to English offline dictionary)​

  • Compile the database, it is very easy, only three commands, no external python dependencies are involved. Also, it is a very small dictionary which only has JLPT words (14540 items). A file japanese.db will be generated.

git clone https://github.com/chenyanming/yomichan-jlpt-vocab
cd yomichan-jlpt-vocab
python make_dictionary_db.py

  • Set paw-jlpt-db and point to the location of japanese.db.
  • PC: Download mecap and unidic-lite or janome

pip install mecab-python3
pip install unidic-lite
# pip install janome # You can also use janome instead

  • Android: Download janome

pkg install tur-repo # https://github.com/termux-user-repository/tur 
pkg install python3.10 # install python 3.10
pip3.10 install janome # janome is much easier to install on Android

  • Setup paw-python-program if necessary, if the pip module is installed with different python version, for android, set it to python3.10

  • Enable paw-annotation-show-unknown-words-p

    • Tweak paw-jlpt-tags, part of or all among "n5 n4 n3 n2 n1"
    • Add words to known words file
      • Setup paw-jlpt-known-words-files and paw-jlpt-default-known-words-file, for example, I have two files, one is csv file downloaded from somewhere, the other is a plain text maintained manually.

(setq paw-jlpt-known-words-files `(,(expand-file-name "japanese.txt" org-directory)))
(setq paw-jlpt-default-known-words-file (expand-file-name "japanese.txt" org-directory))

Text wordlists or dictionaries​

  • You can also disable paw-annotation-show-unknown-words-p, and use your personal wordlist files (csv/txt file dictionaries) instead, by setting paw-jlpt-wordlist-files.
    • csv file requirement: just make sure first column is the word. Other columns will be automatically combined and separated by new line as explanation.
    • txt file requirement: the whole line should be the word, one line one word.

You can think of paw-jlpt-wordlist-files are external text dictionaries. Paw will query the word with the first column of text dictionaries inside paw-jlpt-wordlist-files, and use other columns as explanation. If no explanation, it will use paw-jlpt-db.

You can also think of paw-jlpt-wordlist-files are editable wordlists. All words inside them will be highlighted (Blue by default) in the buffer that are enabled paw-annoatation-mode, while words added into paw-db-file will be highlighted as Orange by default.