Re: isearch C-o patch (post-freeze resubmission)

[Kenichi Handa]

Karl Fogel <address@hidden> writes:
> I've committed the new `isearch-yank-char' function to isearch.el;
> this feature is similar to C-w and C-y in isearch, except it yanks
> just one char at a time, instead of a word or a whole line.

> It is not bound to any key right now.  I proposed binding it to C-o,
> but Richard thought that while it's worthwhile to make the function
> available, it's not widely useful enough to bind it by default.

> I'd like to convince him otherwise :-), but we need more data here.
> In my experience, the binding is tremendously useful when working with
> input methods that require multiple keystrokes to enter a single char.
> I theorize that Richard never edits in such languages, and therefore
> has never experienced the pain of isearching without yank-char in
> those environments.

> However, the fact that I personally find it useful is not enough -- to
> be worth binding, it should be useful to many people.  

> So could other people who use use complex input systems please give
> this a try and see what you think?

I stand for having some key binding for that command.  It's
very very convenient for such scripts that require
not-simple input method.  Actually, I was using my own
version of the same fanctionality for long by binding it to
C-c (not C-o).  As C-w is for word, I thought C-c for
character is good.  In addition, in my experience, it's very
very rare that I want to invoke some command bound to C-c
plus something without terminating incremental search
explicitly, but it's not that rare that I want to open a
line (by typing C-o) without terminating incremental search.



>    cvs update lisp/isearch.el, load it, and eval this:

>    (defun my-isearch-mode-hook ()
>      ;; Or some other key, doesn't have to be C-o...
>      (define-key isearch-mode-map "\C-o" 'isearch-yank-char))
>    (add-hook 'isearch-mode-hook 'my-isearch-mode-hook)

> Or if you've already been using it, please post saying whether or not
> it makes a difference in your editing.

> -Karl

> _______________________________________________
> Emacs-devel mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/emacs-devel