On Mon, Jan 1, 2024 at 10:04 AM Carlo Marcelo Arenas Belón <carenas@xxxxxxxxx> wrote: > Since acabd2048e (grep: correctly identify utf-8 characters with > \{b,w} in -P, 2023-01-08), PCRE2's UCP mode has been enabled when > UTF content was expected and therefore muktibyte characters are > included as part of all character classes (when defined). s/muktibyte/multibyte/