Is there any way to accept only one char? This issue is mainly targeted at East Asian users.

[gaoqiangks]

For Chinese users, accept_word is far too broad in scope. This is because Chinese does not separate words with spaces—an English "word" is more akin to a single Chinese character (char). As a result, accept_word for Chinese users essentially equates to "accept input until the next punctuation mark".
I wonder if there is any way to accept only a single char? Alternatively, is there any way to customize the behavior of accept_word—for example, by defining a custom hook that, when detecting the current virtual text is Chinese, makes accept_word return just the next char instead of reading until the next space or punctuation mark?

[AntoineGS]

Yeah I can detect if the next char is a hanzi/kanji and make the accept word keybind accept just the one character. Can you open a issue for this?

[gaoqiangks]

Yeah I can detect if the next char is a hanzi/kanji and make the accept word keybind accept just the one character. Can you open a issue for this?

I've figured out how to handle it, thanks a lot!