tags:

views:

389

answers:

2

Hi, I am now using Emacs 23 with visual-line-mode turned of for text editing but keep hitting M-q out of habit (thus adding hard-wrapping line endings...). I wonder if there is a way to add a conditional to disable fill-paragraph (or remove the binding to M-q) for modes in which visual-line-mode is turned on, but to re-enable it for those in which I am still using the auto-fill-mode? Thanks!

+4  A: 
(defun maybe-fill-paragraph (&optional justify region)
  "Fill paragraph at or after point (see `fill-paragraph').

Does nothing if `visual-line-mode' is on."
  (interactive (progn
      (barf-if-buffer-read-only)
      (list (if current-prefix-arg 'full) t)))
  (or visual-line-mode
      (fill-paragraph justify region)))

;; Replace M-q with new binding:
(global-set-key "\M-q" 'maybe-fill-paragraph)

Instead of using global-set-key, you can also rebind M-q only in specific modes. (Or, you could change the global binding, and then bind M-q back to fill-paragraph in a specific mode.) Note that many modes are autoloaded, so their keymap may not be defined until the mode is activated. To set a mode-specific binding, I usually use a function like this:

(add-hook 'text-mode-hook
  (defun cjm-fix-text-mode ()
    (define-key text-mode-map "\M-q" 'maybe-fill-paragraph)
    (remove-hook 'text-mode-hook 'cjm-fix-text-mode)))

(The remove-hook isn't strictly necessary, but the function only needs to run once.)

cjm
globally setting `M-q` is evil, because it affects all modes. if you have to do this, use `substitute-key-definition` (or use defadvice or fset instead)
mihi
But that was the whole point. M-q is already a global binding. He wanted it rebound to a function that checked visual-line-mode.
cjm
Hmmm... also a great option... Global setting of M-q may be dangerous but the function is simple enough (and elegant) that its consequences are predictable?
Stephen
ok you won. I thought is was only bound in text-like modes, but I noticed it is bound everywhere, it's `fill-paragraph-function` that's doing the magic in the other modes...
mihi
Thanks for your post and edit! [So I am relatively new at this but wouldn't (add-hook 'text-mode-hook '(lambda () (define-key text-mode-map "\M-q" 'maybe-fill-paragraph))) also do it?]
Stephen
Yes, that will work. I just happen to prefer using a real named function instead of an anonymous one.
cjm
+5  A: 

you can use an advise for this.

For your .emacs:

(defadvice fill-paragraph (around disable-for-visual-line-mode activate)
  (unless visual-line-mode
    ad-do-it))

This will change fill-paragraph to do nothing when visual-line-mode is on. You can also add an error if you prefer that.

mihi
Note that this will completely disable `fill-paragraph` when visual-line-mode is on, even if it's being called from an elisp function. You probably don't want to be that drastic.
cjm
+1 Very pretty
Trey Jackson
I agree, very pretty... but cjm has a point... though I don't know how often fill-paragraph is called from within elisp functions.
Stephen
(I love defadvice though because you can also turn it off without rebooting emacs)
Stephen
**Usage Note:** Advice is useful for altering the behavior of existing calls to an existing function. If you want the new behavior for new calls, or for key bindings, you should define a new function (or a new command) which uses the existing function. (http://www.gnu.org/software/emacs/manual/html_node/elisp/Advising-Functions.html#Advising-Functions)
cjm
Re: "(I love defadvice though because you can also turn it off without rebooting emacs)"You can also bind M-q back to fill-paragraph (either globally or in a specific mode) without restarting Emacs.
cjm
So one point where I saw this might cause conflict was org-mode, where I do use visual-line-mode, but it looks like fill-paragraph is used internally to arrange the table layout. org.el:;; We use our own fill-paragraph function, to make sure that tables;; and fixed-width regions are not wrapped. That function will pass;; through to `fill-paragraph' when appropriate.so I guess I'm a bit queasy about modifying 'fill-paragraph' directly - though it certainly is the sweetest of eye candies!
Stephen