もう一本転送します。 -- Hironobu Yamashita ---------- Forwarded message --------- From: Enrico Gregorio <enrico.gregorio@univr.it> Date: 2019年1月16日(水) 4:14 Subject: Re: regexpatch: error on (u)pTeX with Japanese char To: Hironobu Yamashita <h.y.acetaminophen@gmail.com> Now I know better about the issue. Of course, e-TeX is required, but this you already were aware of. The main problem is that `\tl_set_rescan:Nnn needs to check whether the text to be rescanned starts with a space and to do this it looks at the first token. If the token is not a control sequence, the category code is checked, but \catcode doesn’t work with kanji or kana characters (section 10 in the pTeX manual, if I understand it correctly with the help of Google). The team command of Japanese is very scarce (big understatement). Maybe \kcatcode could help in this case, but the details are only in Japanese. If there is a sane way to check whether a character token will give a character code illegal with \catcode, then I guess a further check can be added when the engine is epTeX. Ciao Enrico