6:25 a.m.
Uwe Brauer writes:
>> Regarding Re: mark-word (forward and backward) many
words????; "Stephen J. Turnbull" <stephen(a)xemacs.org> adds:
> (defun mark-word-backward (&optional count)
> (interactive "p")
> (mark-something 'mark-word-backward
> (lambda (x)
> (and (mark)
> ;; this is tricky: if we get here, then
> ;; mark-something has already done (goto-char (mark))
> (= (mark) (point))
> (setq x (- x)))
> (backward-word x))
> count))
Cool, that works, thanks!
You're welcome!
BTW, I didn't do this at first because your requirements weren't clear
to me, and the code is mildly complex because of the direction
reversal handling. The key phrase was "I can't predict what repeating
the command will mark" which clued me in that the direction reversal
handling was important, then a bit of experiment led to the code above
(it took about five tries, actually).
Why not include it in the next release of simple.el?
Well, I'll think about it. But there are two problems. First, we'd
need to do the same for other backward motion commands. More
important, it's just not right. The right fix is to make
mark-something work as expected with commands whose default direction
is backwards.
Also, Shift-Control-Left does the right thing by default, so it's not
really necessary to make mark-word-backward work by default to get
this level of convenience.
What bothers me more is that mark-something commands don't mark the
whole object, they just mark forward or backward. I guess I'd want an
"outward" direction for mark-something, too. This may complicate the
issue quite a bit.
Posted to the tracker as issue732 (feature request).
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
9:28 p.m.
New subject: mark-word (forward and backward) many words????
> Regarding Re: mark-word (forward and backward) many words????;
"Stephen J. Turnbull" <stephen(a)xemacs.org> adds:
>
> Cool, that works, thanks!
You're welcome!
> Why not include it in the next release of simple.el?
Well, I'll think about it. But there are two problems.
First, we'd need to do the same for other backward
motion commands. More important, it's just not right.
The right fix is to make mark-something work as
expected with commands whose default direction is
backwards.
Well I actually don't understand your code, especially the
lambda function, could you wrap that into another functions
body call it mark-something-backwards and use it in other
function? But again I really don't understand the code.
Also, Shift-Control-Left does the right thing by
default, so it's not really necessary to make
mark-word-backward work by default to get this level of
convenience.
This key combination is not working on my Kubuntu 8.10. By
that I mean when I press
do C-h k Shift-Control-Left xemacs only responds:
C-left is bound to other-window (so it seems to ignore the
shift key.) However in a message buffer xemacs tells me:
C-Sh-left runs `mml-secure-encrypt-pgp odd odd,
So what is this is this key supposed to be?
Oh in xemacs -vanilla Shift-Control-Left does a
mark-backward word, but still tells me:
C-left is bound to other-windows very odd, seems to be a
feature of the X system which is used by XEmacs, at least
partially.
Uwe
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
9:37 p.m.
New subject: C-S-left (was: mark-word (forward and backward) many words????)
> Regarding Re: mark-word (forward and backward) many words????;
"Stephen J. Turnbull" <stephen(a)xemacs.org> adds:
Also, Shift-Control-Left does the right thing by
default, so it's not really necessary to make
mark-word-backward work by default to get this level of
convenience.
Ok I see what was going on (my fault: in my setting Control
left was bound to other window) the default binding is
backward-word, now in XEmacs when you press also the shift
key the mark is set!
This *does not* work in GNU emacs, I think this an example
of a key difference between GNU emacs and XEmacs.
Andreas asked about this a while ago.
Ute
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
9:13 a.m.
New subject: C-S-left
Am 23.10.2010 22:37, schrieb Uwe Brauer:
>> Regarding Re: mark-word (forward and backward) many
words????; "Stephen J. Turnbull"<stephen(a)xemacs.org> adds:
> Also, Shift-Control-Left does the right thing by
> default, so it's not really necessary to make
> mark-word-backward work by default to get this level of
> convenience.
Ok I see what was going on (my fault: in my setting Control
left was bound to other window) the default binding is
backward-word, now in XEmacs when you press also the shift
key the mark is set!
This *does not* work in GNU emacs, I think this an example
of a key difference between GNU emacs and XEmacs.
Andreas asked about this a while ago.
Ute
Hi,
to get the word under cursor marked always completely, you need still
something like
(unless (looking-at "\\>") (forward-word 1))
code below works with
GNU Emacs 23.1.1 (i586-suse-linux-gnu, GTK+ Version 2.20.1) of 2010-07-05
ie mark-something must get known too, but works afterwards.
Andreas
--
https://code.launchpad.net/~a-roehler/python-mode/python-mode-components
https://code.launchpad.net/s-x-emacs-werkstatt/
::::::::::::::
(defun mark-word-backward (&optional count)
(interactive "p")
(unless (looking-at "\\>") (forward-word 1))
(mark-something 'mark-word-backward
(lambda (x)
(and (mark)
;; this is tricky: if we get here, then
;; mark-something has already done (goto-char
(mark))
(= (mark) (point))
(setq x (- x)))
(backward-word x))
count))
(defun mark-something (mark-fn movement-fn arg)
"internal function used by mark-sexp, mark-word, etc."
(let (newmark (pushp t))
(save-excursion
(if (and (eq last-command mark-fn) (mark))
;; Extend the previous state in the same direction:
(progn
(if (< (mark) (point)) (setq arg (- arg)))
(goto-char (mark))
(setq pushp nil)))
(funcall movement-fn arg)
(setq newmark (point)))
(if pushp
(push-mark newmark nil t)
;; Do not mess with the mark stack, but merely adjust the
previous state:
(set-mark newmark)
;; (activate-region)
)))
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
5143
days inactive
5144
days old
3 comments
3 participants
participants (3)
-
Andreas Röhler -
Stephen J. Turnbull -
Uwe Brauer