Emacs completion and handling accented characters with orderless

Feb 26, 2026| emacs

I like using the orderless completion package for Emacs because it allows me to specify different parts of a completion candidate than any order I want. Because I'm learning French, I want commands like consult-line (which uses minibuffer completion) and completion-at-point (which uses in-buffer completion) to also match candidates where the words might have accented characters. For example, instead of having to type "utilisé" with the accented é, I want to type "utilise" and have it match both "utilise" and "utilisé".

(defvar my-orderless-accent-replacements
  '(("a" . "[aàáâãäå]")
    ("e" . "[eèéêë]")
    ("i" . "[iìíîï]")
    ("o" . "[oòóôõöœ]")
    ("u" . "[uùúûü]")
    ("c" . "[cç]")
    ("n" . "[nñ]"))) ; in case anyone needs ñ for Spanish

(defun my-orderless-accent-dispatch (pattern &rest _)
  (seq-reduce
   (lambda (prev val)
     (replace-regexp-in-string (car val) (cdr val) prev))
   my-orderless-accent-replacements
   pattern))

(use-package orderless
  :custom
  (completion-styles '(orderless basic))
  (completion-category-overrides '((file (styles basic partial-completion))))
  (orderless-style-dispatchers '(my-orderless-accent-dispatch orderless-affix-dispatch)))

Figure 1: Screenshot of consult-line showing matching against accented characters

Figure 2: Screenshot of completion-at-point matching "fev" with "février"

This is an entry for Emacs Carnival February 2026: Completion.

This is part of my Emacs configuration.

View Org source for this post

You can comment on Mastodon or e-mail me at sacha@sachachua.com.