gyrojeff's picture
feat: add exclusion rules
a9b34cd
raw
history blame
3.14 kB
# Language code specification: https://www.w3.org/International/articles/language-tags/
# Exclusion:
# - EN
# - SC -> TC
# - Regional
# - CJK Regional
# - Glyphs
# - Symbols
# - Numbers
# - Pure Katakana and Hiragana
# - Others
dataset:
path: ./dataset/fonts
exclusion:
- ./Founder Type(方正)/韩文/方正朝文中圆.TTF
- ./Founder Type(方正)/简繁/ttf/方正宋体S-超大字符集(SIP).TTF
specs:
- path:
- ./Adobe/CJK
rule: |
lambda x: \
x.split('/')[-1].split('\\')[-1].startswith('SourceHanSans-') \
or x.split('/')[-1].split('\\')[-1].startswith('SourceHanSerif-')
language: CJK
- path:
- ./Google(谷歌)/CJK
rule: |
lambda x: \
x.split('/')[-1].split('\\')[-1].startswith('NotoSansCJK-') \
or x.split('/')[-1].split('\\')[-1].startswith('NotoSansMonoCJK-') \
or x.split('/')[-1].split('\\')[-1].startswith('NotoSerifCJK-')
language: CJK
- path:
- ./Arphic(文鼎)/简体
- ./DynaFont(华康)/简体
- ./Fontworks/简体
- ./Founder Type(方正)/简体
- ./Hanyi Fonts(汉仪)/简体
- ./Microsoft(微软)/简体
- ./Monotype(蒙纳)/简体
- ./SinoType(华文)/简体
- ./TensenType(腾祥)/简体
- ./其他/简体
language: zh-Hans
- path:
- ./Adobe/繁体
- ./Arphic(文鼎)/繁体
- ./Fontworks/繁体
- ./Founder Type(方正)/繁体
- ./Hanyi Fonts(汉仪)/繁体
- ./Microsoft(微软)/繁体
- ./Monotype(蒙纳)/繁体
- ./SinoType(华文)/繁体
- ./TensenType(腾祥)/繁体
- ./其他/繁体
language: zh-Hant
- path:
- ./Adobe/简繁
- ./Arphic(文鼎)/简繁
- ./DynaFont(华康)/简繁
- ./Founder Type(方正)/简繁
- ./Hanyi Fonts(汉仪)/简繁
- ./Microsoft(微软)/简繁
- ./Monotype(蒙纳)/简繁
- ./Morisawa(森泽)/简繁
- ./SinoType(华文)/简繁
- ./TensenType(腾祥)/简繁
- ./其他/简繁
language: zh
- path:
- ./Adobe/日文
- ./Arphic(文鼎)/日文
- ./DynaFont(华康)/日文
- ./Fontworks/日文
- ./Monotype(蒙纳)/日文
- ./Morisawa(森泽)/日文/MorisawaAOTF/日文
- ./Morisawa(森泽)/日文/MorisawaAPOTF
- ./Morisawa(森泽)/日文/MorisawaGOTF
- ./Morisawa(森泽)/日文/MorisawaUOTF
- ./TensenType(腾祥)/日文
- ./其他/日文
language: ja
- path:
- ./DynaFont(华康)/繁体/港版
language: zh-Hant-HK
- path:
- ./DynaFont(华康)/繁体/台版
language: zh-Hant-TW
- path:
- ./DynaFont(华康)/韩文
- ./Fontworks/韩文
- ./Founder Type(方正)/韩文
- ./Morisawa(森泽)/韩文
- ./其他/韩文
language: ko