zhwi dictionary for rime

Install

git clone https://github.com/PinkD/rime-pinyin-zhwiki
cd rime-pinyin-zhwiki
make && make install
# or you can use `make VERSION=20200601` to build specified version instead of the latest

NOTE: you need to modify install dir in Makefile depending on your os release

Arch Linux

use aur

Usage

add zhwiki_pinyin into your dictionary table in rime config file

create a custom config named like luna_pinyin.custom.yaml under .config/fcitx/rime (may be different for fcitx5):

patch:
  translator/dictionary: luna_pinyin.extended

create luna_pinyin.extended.dict.yaml to define dictionary:

---
name: luna_pinyin.extended
version: 'v0.1-20200524'
sort: by_weight
use_preset_vocabulary: true
import_tables:
  - 'luna_pinyin'
  - 'zhwiki_pinyin'
...

NOTE: double_pinyin also uses luna_pinyin as dictionary

redeploy RIME and enjoy

Tested for luna_pinyin by me and double_pinyin by haruue

remove words by word frequency manually

# make raw
make zhwiki_pinyin.raw

# convert tc to sc because jieba doesn't support tc
opencc -c /usr/share/opencc/t2s.json -i zhwiki_pinyin.raw -o zhwiki_pinyin.sc

# count word frequency(requires jieba)
python word_frequency.py

# sort by count
sort -rn result.txt > result0.txt
sort -rn result.txt | awk -F ':' '{print $1":"$2}' > result1.txt

# use grep to exclude words you want to
grep -v "xxx" zhwiki_pinyin.sc > zhwiki_pinyin.txt

# convert sc back to tc
opencc -c /usr/share/opencc/s2t.json -i zhwiki_pinyin.txt -o zhwiki_pinyin.raw

# clean
rm -f *.txt

# make
make

Thanks

This project is inspired by felixonmars/fcitx5-pinyin-zhwiki

Change list

• latest
  • add word_frequency.py for excluding useless words manually
• 0.2:
  • remove manually convert tc to sc with opencc because RIME will do that automatically
  • limit word len from 2 to 8
• 0.1: first release