Is there a way to get a character by name (e.g. GREEK SMALL LETTER LAMBDA) instead of by code point (\uX3BB)?


You can use insert-char for interactive usage and

(cdr (assoc-string INPUT (ucs-names) t))

for usage in elisp programs. Thereby, INPUT is the character name string, e.g.,


Please, see the doc strings of insert-char, ucs-names, and assoc-string for more information.


I am using org-entities to input non-ascii characters. Many translations of LaTeX-sequences into utf8 are already implemented there. I added some more that I often need. Then I wrote a command latex-to-utf8 that translates a LaTeX control sequence before point into the corresponding utf8-character. I bound this command to f8.

(defvar latex-to-utf8-scripts)
(setq latex-to-utf8-scripts
      '((?+ ?⁺ ?₊)
    (?- ?⁻ ?₋)
    (?0 ?⁰ ?₀)
    (?1 ?¹ ?₁)
    (?2 ?² ?₂)
    (?3 ?³ ?₃)
    (?4 ?⁴ ?₄)
    (?5 ?⁵ ?₅)
    (?6 ?⁶ ?₆)
    (?7 ?⁷ ?₇)
    (?8 ?⁸ ?₈)
    (?9 ?⁹ ?₉)))

(require 'org-entities)
(mapc (lambda (x) (add-to-list 'org-entities-user x)) '(("bull" "\\bullet" nil "•" "*" "*" "•")
                            ("plusminus" "\\pm" t "±" "+-" "+-" "±")
                            ("sqrt" "\\sqrt" t "√" "sqrt" "sqrt" "√")
                            ("lessless" "\\ll" t "&LessLess;" "<<" "<<" "⪡")
                            ("lesseq" "\\leq" t "&lesseq;" "<=" "<=" "≤")
                            ("greatereq" "\\geq" t "&greatereq;" ">=" ">=" "≥")
                            ("nbsp" "\\medspace" t "&nbsp;" " " " " " ")
                            ("greatergreater" "\\gg" t "&GreaterGreater;" ">>" ">>" "⪢")
                            ("neg" "\\neg" t "&neg;" "/" "/" "¬")
                            ("mapsto" "\\mapsto" t "&mapsto;" "|->"  "|->" "↦")
                            ("times" "\\times" t "&times;" "x" "x" "×")))

(defun latex-to-utf8 ()
  "Replace latex symbol before point by corresponding utf8 symbol."
   ((looking-back "\\\\[[:alpha:]]+" (line-beginning-position))
    (let* ((latex-sym (match-string 0))
       (utf8-sym (loop for org-sym in (append org-entities-user org-entities)
               if (and (listp org-sym) (string-equal latex-sym (nth 1 org-sym))) return (nth 6 org-sym))))
      (if utf8-sym
      (replace-match utf8-sym))))
   ((looking-back "\\([_^]\\)\\([+-]\\|[+-]?[0-9]+\\)" (line-beginning-position))
    (let ((scr_idx (if (string-equal (match-string-no-properties 1) "^") 1 2))
      (digits (string-to-list (match-string 2))))
      (replace-match "")
      (loop for digit in digits do
        (insert (nth scr_idx (assoc digit latex-to-utf8-scripts)))

(global-set-key [f8] 'latex-to-utf8)
| improve this answer | |
  • insert-charis bound to C-x 8 RET, by the way. – asjo Jun 11 '16 at 22:44
  • For usage in elisp programs, I'd recommend just using . After all, with insert-char you can easily insert the λ. BTW, I use insert-char so much I have made a shorter, easier to type keybinding for it. I find C-x 8 RET really awkward to type. – Harald Hanche-Olsen Jun 12 '16 at 7:57
  • Regarding the remark : you should give Tex input method a try. – YoungFrog Jun 14 '16 at 5:09

The question isn't very clear (Is the intended application an interactive command which prompts the user? Is it a non-interactive Elisp program? Do we know the character's name ahead of time? Will it be given as a string? Etc.), but here are some additional ways of mapping character names to code points in Emacs 26.

Emacs 26.1 introduced both the function char-from-name:

(char-from-name "GREEK SMALL LETTER LAMDA")      ; => ?λ
(char-from-name "Greek small letter Lamda" t)    ; => ?λ

and character name escape sequences in literals:

?\N{GREEK SMALL LETTER LAMDA}                    ; => ?λ
?\N{Greek small letter Lamda}                    ; => ?λ
"Lamda: \N{GREEK SMALL LETTER LAMDA}"            ; => "Lamda: λ"
"Lamda: \N{Greek small letter Lamda}"            ; => "Lamda: λ"

These two features are documented under (elisp) Character Codes and (elisp) General Escape Syntax, respectively.

Note that (ucs-names), as mentioned in Tobias' answer, is no longer an alist in Emacs 26, but rather a hash table, so it is now accessed as:

(gethash "GREEK SMALL LETTER LAMDA" (ucs-names)) ; => ?λ

But you're better off using char-from-name instead.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.