Finding the Language of the Current Buffer in Emacs

23 April 2009 - Georgia Tech

I'm currently writing some code that looks at the current buffer and uses it to determine the language the buffer's written in if there is one.

(defun add-import-find-language ()
  (cond
   ;; check major mode
   ((not (string= major-mode "fundamental-mode"))
    (let ((current-mode (symbol-name major-mode)))
      (string-match "^\\([[:alpha:]-]+\\)\\-mode" current-mode)
      (match-string 1 current-mode)))
   ;; check extension
   ((or (buffer-file-name) (buffer-name))
    (let ((bufname (or (buffer-file-name) (buffer-name))))
      (dolist (pair auto-mode-alist)
        (if (string-match (car pair) bufname)
            (let ((current-mode (symbol-name (cdr pair))))
              (string-match "^\\([[:alpha:]-]+\\)\\-mode" current-mode)
              (return (match-string 1 current-mode)))))))
   ;; check buffer-local var language
   (language
    (let ((current-mode (symbol-name language)))
      (string-match "^\\([[:alpha:]-]+\\)" current-mode)
      (return (match-string 0 current-mode)))))
  )

Note: Code highlighting currently disabled due to a bug in Jekyll.

The comments should make most everything pretty clear, but the add-import-find-language function, which will be part of another side project I've been tinkering with, steps through the major-mode, the "file extension" (the caveat there being it checks both buffer-file-name and buffer-name just in case the buffer was created without find-file or similar), and finally checks for a buffer-local variable language to determine the buffer's language, which it returns as a string.

There are no doubt edge-cases where this function fails to do the right thing. If you happen to find one, please file an issue and, if possible, also a link to your own fork of add-import.el with the fix(es).

Related Posts