Re: @#$! new behavior of C-b & C-f in i-search

Re: @#$! new behavior of C-b & C-f in i-search

[Karl Fogel]

Oh well.  I was kind of liking the new behavior.  But it sounds like
we have different usage patterns.

I don't know whether it's best to leave the change in the code or not.
I did want to chime in to say that it's not been a negative experience
for everyone, though.

-Karl

Miles Bader <miles@lsi.nec.co.jp> writes:
> I originally liked the idea of `C-b' & `C-f' in i-search doing special
> stuff, but ever since this was installed, it's been driving me batty --
> I apparently use those keystrokes to exit i-search _way_ more than I
> thought.
> 
> The behavior is especially insidious because following scenario happens
> really often:
> 
> I'll do an i-search, try to exit it by hitting C-b a few times expecting
> the cursor to move, and when it _doesn't_ move, I'll realize `Oh, it's
> the new behavior' and try to exit i-search by hitting RET -- but that
> doesn't work, because hitting C-b has emptied the search string (without
> moving the search location), so RET drops me into the non-incremental
> Search: prompt, and when I then hit C-g to try to _really_ exit, it
> aborts the search, and warps me back to where I began!
> 
> At that point I can't even just redo the search by hitting `C-s C-s',
> because the aborted i-search didn't save it.
> 
> Grrrrrrrr....
> 
> -Miles
> -- 
> `...the Soviet Union was sliding in to an economic collapse so comprehensive
>  that in the end its factories produced not goods but bads: finished products
>  less valuable than the raw materials they were made from.'  [The Economist]
> 
> 
> _______________________________________________
> Emacs-devel mailing list
> Emacs-devel@gnu.org
> http://mail.gnu.org/mailman/listinfo/emacs-devel

@#$! new behavior of C-b & C-f in i-search

[Miles Bader]

I originally liked the idea of `C-b' & `C-f' in i-search doing special
stuff, but ever since this was installed, it's been driving me batty --
I apparently use those keystrokes to exit i-search _way_ more than I
thought.

The behavior is especially insidious because following scenario happens
really often:

I'll do an i-search, try to exit it by hitting C-b a few times expecting
the cursor to move, and when it _doesn't_ move, I'll realize `Oh, it's
the new behavior' and try to exit i-search by hitting RET -- but that
doesn't work, because hitting C-b has emptied the search string (without
moving the search location), so RET drops me into the non-incremental
Search: prompt, and when I then hit C-g to try to _really_ exit, it
aborts the search, and warps me back to where I began!

At that point I can't even just redo the search by hitting `C-s C-s',
because the aborted i-search didn't save it.

Grrrrrrrr....

-Miles
-- 
`...the Soviet Union was sliding in to an economic collapse so comprehensive
 that in the end its factories produced not goods but bads: finished products
 less valuable than the raw materials they were made from.'  [The Economist]

Re: @#$! new behavior of C-b & C-f in i-search

[Juri Linkov]

Miles Bader <miles@lsi.nec.co.jp> writes:
> Grrrrrrrr....

I noticed this problem too.  One solution is to always retain
one character in the search string.  Another solution is to
exit isearch when C-b is typed with the empty isearch string.

But still apparently C-b and C-f can't be made fully transparent
as motion keys, because they change the search string when users
use them as motion keys in isearch mode and don't expect that
they will change the search history.

So other key bindings are needed.  I already proposed some of them:

    (define-key isearch-mode-map "\C-w"    'isearch-yank-word)
    (define-key isearch-mode-map "\C-\M-w" 'isearch-yank-word-or-char)
    (define-key isearch-mode-map "\C-\M-b" 'isearch-del-char)
    (define-key isearch-mode-map "\C-\M-f" 'isearch-yank-char)

    (define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
    (define-key isearch-mode-map "\M-w"    'isearch-yank-word)
    (define-key isearch-mode-map "\C-\M-w" 'isearch-del-char)
    (define-key isearch-mode-map "\C-\M-y" 'isearch-yank-char)

    (define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
    (define-key isearch-mode-map "\M-w"    'isearch-del-char)
    (define-key isearch-mode-map "\C-\M-w" 'isearch-yank-char)
    (define-key isearch-mode-map "\C-\M-y" 'isearch-yank-word)

The last group seems most preferable since these key are less probable
to be used in isearch mode.

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: @#$! new behavior of C-b & C-f in i-search

[Kai Grossjohann]

Miles Bader <miles@lsi.nec.co.jp> writes:

> I originally liked the idea of `C-b' & `C-f' in i-search doing special
> stuff, but ever since this was installed, it's been driving me batty --
> I apparently use those keystrokes to exit i-search _way_ more than I
> thought.

I have similar feelings.  However, I have decided that I would like
to try it for a while longer before reaching a decision.  Perhaps I
can retrain myself to always exit with RET.

Just another voice,
Kai

Re: @#$! new behavior of C-b & C-f in i-search

[David Kastrup]

Kai Grossjohann <kgrossjo@eu.uu.net> writes:

> Miles Bader <miles@lsi.nec.co.jp> writes:
> 
> > I originally liked the idea of `C-b' & `C-f' in i-search doing
> > special stuff, but ever since this was installed, it's been
> > driving me batty -- I apparently use those keystrokes to exit
> > i-search _way_ more than I thought.
> 
> I have similar feelings.  However, I have decided that I would like
> to try it for a while longer before reaching a decision.  Perhaps I
> can retrain myself to always exit with RET.

No problems up to now.  I guess it depends on habits.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: @#$! new behavior of C-b & C-f in i-search

[Robert J. Chassell]

   I originally liked the idea of `C-b' & `C-f' in i-search doing special
   stuff, but ever since this was installed, it's been driving me batty --
   I apparently use those keystrokes to exit i-search _way_ more than I
   thought.

Same here.  Please change them, perhaps to M-C-b and M-C-f.

The feature itself is exellent.  But I still have trouble with the
keybindings.  I never knew I used `C-b' & `C-f' so often, but I do,
and the habit is very hard to break.

-- 
    Robert J. Chassell                         Rattlesnake Enterprises
    As I slowly update it,                     bob@rattlesnake.com
        I rewrite a "What's New" segment for   http://www.rattlesnake.com

Re: @#$! new behavior of C-b & C-f in i-search

[Richard Stallman]

    I originally liked the idea of `C-b' & `C-f' in i-search doing special
    stuff, but ever since this was installed, it's been driving me batty --
    I apparently use those keystrokes to exit i-search _way_ more than I
    thought.

I am not surprised.

	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
	(define-key isearch-mode-map "\M-w"    'isearch-yank-word)
	(define-key isearch-mode-map "\C-\M-w" 'isearch-del-char)
	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-char)

	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
	(define-key isearch-mode-map "\M-w"    'isearch-del-char)
	(define-key isearch-mode-map "\C-\M-w" 'isearch-yank-char)
	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-word)

C-M-w and C-M-y might be good to use, but I suspect
people use M-w to exit a search.  It makes sense there,
to copy the text from where the search started.

Re: @#$! new behavior of C-b & C-f in i-search

[Karl Eichwalder]

Kai Grossjohann <kgrossjo@eu.uu.net> writes:

> I have similar feelings.  However, I have decided that I would like
> to try it for a while longer before reaching a decision.  Perhaps I
> can retrain myself to always exit with RET.

Agreed.  The current implementation looks clean and I'm sure I can get
used to it the next weeks.


"Robert J. Chassell" <bob@rattlesnake.com> writes:

> Same here.  Please change them, perhaps to M-C-b and M-C-f.

Being force to press three key to move one step forward or backward is
too expensive.  As long as you can press M-f or M-b to break out of the
search the implementation is okay - To break out I mostly use: RET,
C-SPC, M-f or M-b, or - yes - arrow-left or arrow-right.

> The feature itself is exellent.  But I still have trouble with the
> keybindings.  I never knew I used `C-b' & `C-f' so often, but I do,
> and the habit is very hard to break.

yes, but I think it is more important that having found the first
occurance of the (beginning of) string you can press several C-f in a
row to extent the search string; that less error prone than being force
to enter all the different letters: k-i-n-d-e-r-g-a-r-t-e-n.  Yes, I
know, you can complete the search word using C-w; but often, you are not
interested in the complete word and thus you cannot make use of C-w.

-- 
                                                         |      ,__o
                                                         |    _-\_<,
http://www.gnu.franken.de/ke/                            |   (*)/'(*)

Re: @#$! new behavior of C-b & C-f in i-searchi

[Alan Mackenzie]

On Tue, 18 May 2004, Richard Stallman wrote:

>    I originally liked the idea of `C-b' & `C-f' in i-search doing
>    special stuff, but ever since this was installed, it's been driving
>    me batty -- I apparently use those keystrokes to exit i-search _way_
>    more than I thought.

>I am not surprised.

> #1	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
> #2	(define-key isearch-mode-map "\M-w"    'isearch-yank-word)
> #3	(define-key isearch-mode-map "\C-\M-w" 'isearch-del-char)
> #4	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-char)

> #5	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
> #6	(define-key isearch-mode-map "\M-w"    'isearch-del-char)
> #7	(define-key isearch-mode-map "\C-\M-w" 'isearch-yank-char)
> #8	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-word)

>C-M-w and C-M-y might be good to use, but I suspect people use M-w to
>exit a search.  It makes sense there, to copy the text from where the
>search started.

<sticks up hand>.  Yes, I frequently use M-w this way.  I put point at
the beginning of something I want to copy, use C-s, C-w, C-w, ...... to
move over and highlight the something, then M-w to copy in into the kill
ring.  I'd be C-f'ing and C-b'linding myself, if this stopped working.  ;-(

Please be _very_ careful, what bindings get shadowed here.

-- 
Alan Mackenzie (Munich, Germany)

Re: @#$! new behavior of C-b & C-f in i-search

[Juri Linkov]

Richard Stallman <rms@gnu.org> writes:
> 	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
> 	(define-key isearch-mode-map "\M-w"    'isearch-yank-word)
> 	(define-key isearch-mode-map "\C-\M-w" 'isearch-del-char)
> 	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-char)
>
> 	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
> 	(define-key isearch-mode-map "\M-w"    'isearch-del-char)
> 	(define-key isearch-mode-map "\C-\M-w" 'isearch-yank-char)
> 	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-word)
>
> C-M-w and C-M-y might be good to use, but I suspect
> people use M-w to exit a search.  It makes sense there,
> to copy the text from where the search started.

It makes no less sense to use C-w to kill the text from where the
search started, but people already know that C-w has special meaning
in isearch mode.

Anyway, it seems that C-M-w and C-M-y are sufficient, and final
key bindings could be:

	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
	(define-key isearch-mode-map "\C-\M-w" 'isearch-del-char)
	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-char)

C-M-w is even more convenient than C-b because `isearch-del-char' and
`isearch-yank-word-or-char' are on the same `w' key, which will
allow more easily to type C-M-w when C-w grabs too much text.

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: @#$! new behavior of C-b & C-f in i-search

[Miles Bader]

Juri Linkov <juri@jurta.org> writes:
> 	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
> 	(define-key isearch-mode-map "\C-\M-w" 'isearch-del-char)
> 	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-char)
> 
> C-M-w is even more convenient than C-b because `isearch-del-char' and
> `isearch-yank-word-or-char' are on the same `w' key, which will
> allow more easily to type C-M-w when C-w grabs too much text.

Perhaps, but the mnemonics are pretty bizarre -- C-w adds _adds_ a word,
but C-M-w _deletes_ a char?

-Miles
-- 
P.S.  All information contained in the above letter is false,
      for reasons of military security.

Re: @#$! new behavior of C-b & C-f in i-search

[Juri Linkov]

Miles Bader <miles@lsi.nec.co.jp> writes:
> Juri Linkov <juri@jurta.org> writes:
>> 	(define-key isearch-mode-map "\C-w"    'isearch-yank-word-or-char)
>> 	(define-key isearch-mode-map "\C-\M-w" 'isearch-del-char)
>> 	(define-key isearch-mode-map "\C-\M-y" 'isearch-yank-char)
>> 
>> C-M-w is even more convenient than C-b because `isearch-del-char' and
>> `isearch-yank-word-or-char' are on the same `w' key, which will
>> allow more easily to type C-M-w when C-w grabs too much text.
>
> Perhaps, but the mnemonics are pretty bizarre -- C-w adds _adds_ a word,
> but C-M-w _deletes_ a char?

There is not too much choice for good mnemonics in such extremely limited
decision space.

But C-M-w still makes sense because C-w not only adds a word, but also
a character (`isearch-yank-word-or-char').

And binding C-M-y to `isearch-yank-char' will have the same mnemonics
as C-y for `isearch-yank-line' (where `y' supposedly means `yank').

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: @#$! new behavior of C-b & C-f in i-search

[Juri Linkov]

Karl Eichwalder <ke@gnu.franken.de> writes:
> "Robert J. Chassell" <bob@rattlesnake.com> writes:
>> The feature itself is exellent.  But I still have trouble with the
>> keybindings.  I never knew I used `C-b' & `C-f' so often, but I do,
>> and the habit is very hard to break.
>
> yes, but I think it is more important that having found the first
> occurance of the (beginning of) string you can press several C-f in a
> row to extent the search string; that less error prone than being force
> to enter all the different letters: k-i-n-d-e-r-g-a-r-t-e-n.  Yes, I
> know, you can complete the search word using C-w; but often, you are not
> interested in the complete word and thus you cannot make use of C-w.

There is no doubt that functions to modify the search
string character-wise should have dedicated key bindings.
But it became apparent that C-f and C-b are quite intrusive.
Luckily, C-M-w and C-M-y are no less convenient to type,
because they are located on the same keys to which other
yank-related functions are bound: C-w, C-y, M-y.  And they
will cause no harm because C-M-w is a rare command (especially
in isearch mode) and C-M-y is undefined.

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: @#$! new behavior of C-b & C-f in i-search

[David Kastrup]

Juri Linkov <juri@jurta.org> writes:

> Karl Eichwalder <ke@gnu.franken.de> writes:
> > "Robert J. Chassell" <bob@rattlesnake.com> writes:
> >> The feature itself is exellent.  But I still have trouble with the
> >> keybindings.  I never knew I used `C-b' & `C-f' so often, but I do,
> >> and the habit is very hard to break.
> >
> > yes, but I think it is more important that having found the first
> > occurance of the (beginning of) string you can press several C-f in a
> > row to extent the search string; that less error prone than being force
> > to enter all the different letters: k-i-n-d-e-r-g-a-r-t-e-n.  Yes, I
> > know, you can complete the search word using C-w; but often, you are not
> > interested in the complete word and thus you cannot make use of C-w.
> 
> There is no doubt that functions to modify the search
> string character-wise should have dedicated key bindings.
> But it became apparent that C-f and C-b are quite intrusive.
> Luckily, C-M-w and C-M-y are no less convenient to type,

Disagree.  On many keyboards you have no Meta key, and even on those
that do, you need two hands to reach that combination.

I think that the search string modifying is more important feature
than search exiting via C-f and C-b.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: @#$! new behavior of C-b & C-f in i-search

[Kenichi Handa]

In article <x57jv736b9.fsf@lola.goethe.zz>, David Kastrup <dak@gnu.org> writes:
>>  There is no doubt that functions to modify the search
>>  string character-wise should have dedicated key bindings.
>>  But it became apparent that C-f and C-b are quite intrusive.
>>  Luckily, C-M-w and C-M-y are no less convenient to type,

> Disagree.  On many keyboards you have no Meta key, and even on those
> that do, you need two hands to reach that combination.

We anyway use two hands to type C-b, but I agree C-M-w and
C-M-y are very inconvenient keys.

> I think that the search string modifying is more important feature
> than search exiting via C-f and C-b.

To me, their importance are almost the same, and I prefer
convient keys to keys of meaningful mnemonic.  So, I'm using
adjoining C-i and C-o for isearch-del-char and
isearch-yank-char recently.  This is quite convenient at
least for me (I have been using C-q C-i to isearch for TAB).
What do you think?
	
And, i(n) and o(ut) may give some meaning to these keys.

---
Ken'ichi HANDA
handa@m17n.org

Re: @#$! new behavior of C-b & C-f in i-search

[David Kastrup]

Kenichi Handa <handa@m17n.org> writes:

> In article <x57jv736b9.fsf@lola.goethe.zz>, David Kastrup <dak@gnu.org> writes:
> >>  There is no doubt that functions to modify the search
> >>  string character-wise should have dedicated key bindings.
> >>  But it became apparent that C-f and C-b are quite intrusive.
> >>  Luckily, C-M-w and C-M-y are no less convenient to type,
> 
> > Disagree.  On many keyboards you have no Meta key, and even on those
> > that do, you need two hands to reach that combination.
> 
> We anyway use two hands to type C-b,

You do?  I usually type C-b with just the left hand, and I guess that
would be the usual way for touch typists.  Or sometimes I do just with
the right, in case I am busy with the left hand eating or something.

> but I agree C-M-w and C-M-y are very inconvenient keys.

> > I think that the search string modifying is more important feature
> > than search exiting via C-f and C-b.
> 
> To me, their importance are almost the same, and I prefer convient
> keys to keys of meaningful mnemonic.  So, I'm using adjoining C-i
> and C-o for isearch-del-char and isearch-yank-char recently.  This
> is quite convenient at least for me (I have been using C-q C-i to
> isearch for TAB).  What do you think?

Well, considering that the right control key on today's keyboards is
rather small if at all existent, those are definitely less convenient
than C-f and C-b.

> And, i(n) and o(ut) may give some meaning to these keys.

But I think it rather common to be searching for TAB characters in
isearch strings, particularly in regexps.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: @#$! new behavior of C-b & C-f in i-search

[Kenichi Handa]

In article <x5u0yb1ii1.fsf@lola.goethe.zz>, David Kastrup <dak@gnu.org> writes:
>>  We anyway use two hands to type C-b,

> You do?  I usually type C-b with just the left hand, and I guess that
> would be the usual way for touch typists.  Or sometimes I do just with
> the right, in case I am busy with the left hand eating or something.

Touch typists don't have to type C-b.  And I think they use
Right Shift to type B (i.e. both hands).

At least with my hand (I think it's average as Japanese, but
perhaps smaller than average Western person), pressing both
Ctrl and b with the left hand is a little bit difficult.

>>  To me, their importance are almost the same, and I prefer convient
>>  keys to keys of meaningful mnemonic.  So, I'm using adjoining C-i
>>  and C-o for isearch-del-char and isearch-yank-char recently.  This
>>  is quite convenient at least for me (I have been using C-q C-i to
>>  isearch for TAB).  What do you think?

> Well, considering that the right control key on today's keyboards is
> rather small if at all existent, those are definitely less convenient
> than C-f and C-b.

I think we don't have to consider that kind of convenience
(i.e. requires two hands or not) to decide a key binding.

>>  And, i(n) and o(ut) may give some meaning to these keys.

> But I think it rather common to be searching for TAB characters in
> isearch strings, particularly in regexps.

For me, not that frequent as using isearch-del-char.  And,
as I wrote, TAB can be easily searched for by typing C-q
C-i.

---
Ken'ichi HANDA
handa@m17n.org

Re: @#$! new behavior of C-b & C-f in i-search

[Kevin Rodgers]

David Kastrup wrote:
 > Juri Linkov <juri@jurta.org> writes:
 >>There is no doubt that functions to modify the search
 >>string character-wise should have dedicated key bindings.
 >>But it became apparent that C-f and C-b are quite intrusive.
 >>Luckily, C-M-w and C-M-y are no less convenient to type,
 >
 > Disagree.  On many keyboards you have no Meta key, and even on those
 > that do, you need two hands to reach that combination.
 >
 > I think that the search string modifying is more important feature
 > than search exiting via C-f and C-b.

"Hi, my name is Kevin and I use C-f and C-b to exit isearch."

Since incremental search is a mode, how about C-c C-f and C-c C-b?

-- 
Kevin Rodgers

Re: @#$! new behavior of C-b & C-f in i-search

[David Kastrup]

Kevin Rodgers <ihs_4664@yahoo.com> writes:

> David Kastrup wrote:
>  > Juri Linkov <juri@jurta.org> writes:
>  >>There is no doubt that functions to modify the search
>  >>string character-wise should have dedicated key bindings.
>  >>But it became apparent that C-f and C-b are quite intrusive.
>  >>Luckily, C-M-w and C-M-y are no less convenient to type,
>  >
>  > Disagree.  On many keyboards you have no Meta key, and even on those
>  > that do, you need two hands to reach that combination.
>  >
>  > I think that the search string modifying is more important feature
>  > than search exiting via C-f and C-b.
> 
> "Hi, my name is Kevin and I use C-f and C-b to exit isearch."
> 
> Since incremental search is a mode, how about C-c C-f and C-c C-b?

While multiple keypresses of that still require much more action than
C-f and C-b, I guess I could live with that.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: @#$! new behavior of C-b & C-f in i-search

[Peter Lee]

>>>> David Kastrup writes:

    David> While multiple keypresses of that still require much more
    David> action than C-f and C-b, I guess I could live with that.

I wouldn't find this useful if it's more than a single keypress.
Otherwise it becomes more efficient to simply M-e and edit the string.
Personally I could become quite fond of C-f and C-b since I usually
use M-b to exit anyway.  My ½ cents :)

Re: @#$! new behavior of C-b & C-f in i-search

[Juri Linkov]

Peter Lee <pete_lee@swbell.net> writes:
>>>>> David Kastrup writes:
>     David> While multiple keypresses of that still require much more
>     David> action than C-f and C-b, I guess I could live with that.
>
> I wouldn't find this useful if it's more than a single keypress.
> Otherwise it becomes more efficient to simply M-e and edit the string.

M-e is a good method to delete characters from the search string,
but it can't append characters from the buffer where isearch is active.

But it's possible to use something like below.  When C-f is typed
in the end of the isearch minibuffer (activated by M-e) it could append
characters from the isearch buffer to the search string in the minibuffer.

(define-key minibuffer-local-isearch-map "\C-f"  'isearch-yank-char-in-minibuffer)
(define-key minibuffer-local-isearch-map [right] 'isearch-yank-char-in-minibuffer)

(defun isearch-yank-char-in-minibuffer ()
  "Pull next character from isearch buffer into end of search string in minibuffer."
  (interactive "p")
  (if (eobp)
      (insert
       (save-excursion
         (set-buffer (cadr (buffer-list)))
         (buffer-substring-no-properties
          (point) (progn (forward-char 1) (point)))))
    (forward-char 1)))

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: @#$! new behavior of C-b & C-f in i-search

[Juri Linkov]

Juri Linkov <juri@jurta.org> writes:
> When C-f is typed in the end of the isearch minibuffer (activated by
> M-e) it could append characters from the isearch buffer to the
> search string in the minibuffer.

Since no one objected, I propose the following patch.  Generally, this
means that to pull characters into the search string instead of typing

C-s C-f C-f ... C-f C-s

to type one key more:

C-s M-e C-f C-f ... C-f C-s

But there is another small difference: the cursor is moved into the
minibuffer.  So separate keybindings for `isearch-yank-char' and
`isearch-del-char' are still needed.  I propose to bind them to
C-M-w and C-M-y, because these keybindings are absolutely harmless
(unlike other proposals) and have good mnemonics being bound to the
same keys as `isearch-yank-word-or-char' and `isearch-yank-line'.
For keyboards where C-M-w and C-M-y are not easy to type, the
alternative is `isearch-yank-char-in-minibuffer' accessible by M-e.

Index: lisp/isearch.el
===================================================================
RCS file: /cvsroot/emacs/emacs/lisp/isearch.el,v
retrieving revision 1.228
diff -u -w -b -r1.228 isearch.el
--- lisp/isearch.el	6 Jun 2004 13:57:39 -0000	1.228
+++ lisp/isearch.el	7 Jun 2004 00:59:38 -0000
@@ -294,6 +294,8 @@
     (define-key map " " 'isearch-whitespace-chars)
     (define-key map [?\S-\ ] 'isearch-whitespace-chars)
 
+    (define-key map "\C-\M-w" 'isearch-del-char)
+    (define-key map "\C-\M-y" 'isearch-yank-char)
     (define-key map "\C-w" 'isearch-yank-word-or-char)
     (define-key map "\C-y" 'isearch-yank-line)
 
@@ -343,11 +345,14 @@
     (define-key map "\M-\t" 'isearch-complete-edit)
     (define-key map "\C-s" 'isearch-forward-exit-minibuffer)
     (define-key map "\C-r" 'isearch-reverse-exit-minibuffer)
+    (define-key map "\C-f"  'isearch-yank-char-in-minibuffer)
+    (define-key map [right] 'isearch-yank-char-in-minibuffer)
     map)
   "Keymap for editing isearch strings in the minibuffer.")
 
@@ -1126,10 +1131,21 @@
 	  (goto-char isearch-other-end))
      (buffer-substring-no-properties (point) (funcall jumpform)))))
 
+(defun isearch-yank-char-in-minibuffer (&optional arg)
+  "Pull next character from isearch buffer into end of search string in minibuffer."
+  (interactive "p")
+  (if (eobp)
+      (insert
+       (save-excursion
+         (set-buffer (cadr (buffer-list)))
+         (buffer-substring-no-properties
+          (point) (progn (forward-char arg) (point)))))
+    (forward-char arg)))
+
 (defun isearch-yank-char (&optional arg)
   "Pull next character from buffer into search string."
   (interactive)
   (isearch-yank-internal (lambda () (forward-char 1) (point))))

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: @#$! new behavior of C-b & C-f in i-search

[Richard Stallman]

    > When C-f is typed in the end of the isearch minibuffer (activated by
    > M-e) it could append characters from the isearch buffer to the
    > search string in the minibuffer.

    Since no one objected, I propose the following patch.  Generally, this
    means that to pull characters into the search string instead of typing

This change is ok with me.

Please write the changes in NEWS and the Emacs Manual before
you install the change in the code.

Re: @#$! new behavior of C-b & C-f in i-search

[Juri Linkov]

Richard Stallman <rms@gnu.org> writes:
>     > When C-f is typed in the end of the isearch minibuffer (activated by
>     > M-e) it could append characters from the isearch buffer to the
>     > search string in the minibuffer.
>
>     Since no one objected, I propose the following patch.  Generally, this
>     means that to pull characters into the search string instead of typing
>
> This change is ok with me.

I have another idea (I'm sorry, it didn't occur to me earlier).

C-M-w could be used as a key switching between word and character
modes, i.e. after typing C-M-w in isearch mode subsequent
C-w will pull a character and DEL will delete a character.
Another C-M-w will switch back to word mode with C-w and DEL
restoring their current meaning.

However, this might be too complicated to be useful.

So unless someone finds this idea worth considering further,
a simpler solution with C-M-y to pull and C-M-w to delete
a character will be installed.

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: @#$! new behavior of C-b & C-f in i-search

[Kim F. Storm]

Juri Linkov <juri@jurta.org> writes:

> C-M-w could be used as a key switching between word and character
> modes, i.e. after typing C-M-w in isearch mode subsequent
> C-w will pull a character and DEL will delete a character.
> Another C-M-w will switch back to word mode with C-w and DEL
> restoring their current meaning.
> 
> However, this might be too complicated to be useful.

I think so.

> 
> So unless someone finds this idea worth considering further,
> a simpler solution with C-M-y to pull and C-M-w to delete
> a character will be installed.

I think that is good enough.  Those of us who want simpler bindings
(that don't interfere with personal habits) can just make them
ourselves.


-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: @#$! new behavior of C-b & C-f in i-search

[Alan Mackenzie]

On Fri, 11 Jun 2004, Juri Linkov wrote:

>Richard Stallman <rms@gnu.org> writes:
>>     > When C-f is typed in the end of the isearch minibuffer (activated by
>>     > M-e) it could append characters from the isearch buffer to the
>>     > search string in the minibuffer.

>>     Since no one objected, I propose the following patch.  Generally, this
>>     means that to pull characters into the search string instead of typing

>> This change is ok with me.

>I have another idea (I'm sorry, it didn't occur to me earlier).

>C-M-w could be used as a key switching between word and character modes,
>i.e. after typing C-M-w in isearch mode subsequent C-w will pull a
>character and DEL will delete a character.  Another C-M-w will switch
>back to word mode with C-w and DEL restoring their current meaning.

Please don't!  This would inevitably lead to certain types of users (of
which I am one) continually getting a word instead of a character, or
vice versa, and getting angry about it.

There are programs in which one has manually to switch between modes to
get work done.  Emacs isn't one of them.  "If you want vi, you know where
to find it."

>However, this might be too complicated to be useful.

>So unless someone finds this idea worth considering further, a simpler
>solution with C-M-y to pull and C-M-w to delete a character will be
>installed.

>Juri Linkov

-- 
Alan Mackenzie (Munich, Germany)
acm@muc.de

Re: @#$! new behavior of C-b & C-f in i-search

[Richard Stallman]

    C-M-w could be used as a key switching between word and character
    modes, i.e. after typing C-M-w in isearch mode subsequent
    C-w will pull a character and DEL will delete a character.
    Another C-M-w will switch back to word mode with C-w and DEL
    restoring their current meaning.

This might be too complex to be useful.  But people could try it out
and report if they use it.