[PATCH] (Or "question"...) isearch-kill-found

[PATCH] (Or "question"...) isearch-kill-found

[Deniz Dogan]

[-- Attachment #1: Type: text/plain, Size: 386 bytes --]

Recently, in a discussion on #emacs, we looked for a command which
kills the text that has been found using isearch. None of us knew if
there existed anything like that and I coludn't find it.

So attached is a patch which adds isearch-kill-found and binds it to
M-C-m (not too keen on the choice of binding though).

Any thoughts or comments? Does this already exist?

-- 
Deniz Dogan

[-- Attachment #2: isearch-kill-found.patch --]
[-- Type: application/octet-stream, Size: 780 bytes --]

=== modified file 'lisp/isearch.el'
--- lisp/isearch.el	2010-05-20 22:33:09 +0000
+++ lisp/isearch.el	2010-05-30 12:27:58 +0000
@@ -462,6 +462,7 @@
     (define-key map "\M-\C-w" 'isearch-del-char)
     (define-key map "\M-\C-y" 'isearch-yank-char)
     (define-key map    "\C-y" 'isearch-yank-line)
+    (define-key map "\M-\C-m" 'isearch-kill-found)
 
     (define-key map (char-to-string help-char) isearch-help-map)
     (define-key map [help] isearch-help-map)
@@ -1497,6 +1498,12 @@
   (interactive)
   (isearch-yank-string (x-get-selection)))
 
+(defun isearch-kill-found ()
+  "Kills the region that isearch has found."
+  (interactive)
+  (isearch-exit)
+  (kill-region isearch-other-end (point)))
+
 
 (defun isearch-mouse-2 (click)
   "Handle mouse-2 in Isearch mode.

Re: [PATCH] (Or "question"...) isearch-kill-found

[Johan Bockgård]

Deniz Dogan <deniz.a.m.dogan@gmail.com> writes:

> Recently, in a discussion on #emacs, we looked for a command which
> kills the text that has been found using isearch. None of us knew if
> there existed anything like that and I coludn't find it.
>
> So attached is a patch which adds isearch-kill-found and binds it to
> M-C-m (not too keen on the choice of binding though).
>
> Any thoughts or comments? Does this already exist?

A command to simply exit isearch and set the region around the match
would be more generally useful. The key M-RET would make more sense
there too.

http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00492.html

Re: [PATCH] (Or "question"...) isearch-kill-found

[Deniz Dogan]

2010/5/30 Johan Bockgård <bojohan@gnu.org>:
> Deniz Dogan <deniz.a.m.dogan@gmail.com> writes:
>
>> Recently, in a discussion on #emacs, we looked for a command which
>> kills the text that has been found using isearch. None of us knew if
>> there existed anything like that and I coludn't find it.
>>
>> So attached is a patch which adds isearch-kill-found and binds it to
>> M-C-m (not too keen on the choice of binding though).
>>
>> Any thoughts or comments? Does this already exist?
>
> A command to simply exit isearch and set the region around the match
> would be more generally useful. The key M-RET would make more sense
> there too.
>
> http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00492.html
>

I agree, that makes sense.

-- 
Deniz Dogan

RE: [PATCH] (Or "question"...) isearch-kill-found

[Drew Adams]

> >> Recently, in a discussion on #emacs, we looked for a command which
> >> kills the text that has been found using isearch. None of 
> >> us knew if there existed anything like that and I coludn't find it.
> >>
> >> So attached is a patch which adds isearch-kill-found and 
> >> binds it to M-C-m (not too keen on the choice of binding though).
> >>
> >> Any thoughts or comments? Does this already exist?
> >
> > A command to simply exit isearch and set the region around the match
> > would be more generally useful. The key M-RET would make more sense
> > there too.
> >
> > http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00492.html
> 
> I agree, that makes sense.

(FYI, that link is to a side post that is not representative of the thread's
proposal.)

FWIW, as the person who proposed this feature, here is the history:

I mentioned this possibility in help-gnu-emacs on 2006-07-21:
http://lists.gnu.org/archive/html/help-gnu-emacs/2006-07/msg00370.html
Andreas Roehler contributed some ideas in that thread.
I added the feature to my isearch+.el code on 2006-07-29.

I proposed the feature to emacs-devel on 2007-07-10:
http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00490.html

The bottom line was Richard's decision:

 "I don't think it is needed.  You can type C-@ C-r C-r to put
 the region around the match that was just found."

If you want to try it, to see what you think:
http://www.emacswiki.org/emacs/isearch%2b.el

`C-SPC' during isearch toggles the effect of the option
`isearchp-set-region-flag', which sets the default behavior.

Re: [PATCH] (Or "question"...) isearch-kill-found

[Christoph]

On 05/30/2010 09:25 AM, Drew Adams wrote:

> I proposed the feature to emacs-devel on 2007-07-10:
> http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00490.html
>
> The bottom line was Richard's decision:
>
>   "I don't think it is needed.  You can type C-@ C-r C-r to put
>   the region around the match that was just found."
>    
In that same thread Davis Herring had an idea that sounds interesting 
and useful to me:
http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00492.html

>  Thinking about this, I realize
>  that I would get almost all the utility of this idea with just one command
>  added to isearch which meant "exit and move point to start of match", or
>  (closer to your idea) a command which meant "exit and drop mark at start
>  of match" in place of your toggle.  WDOT?


I think a function "exit and drop mark at start of match" would be very 
useful.

Christoph

Re: [PATCH] (Or "question"...) isearch-kill-found

[Lennart Borgman]

On Mon, May 31, 2010 at 12:19 AM, Christoph <cschol2112@googlemail.com> wrote:
>
> In that same thread Davis Herring had an idea that sounds interesting and
> useful to me:
> http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00492.html
>
>>  Thinking about this, I realize
>>  that I would get almost all the utility of this idea with just one
>> command
>>  added to isearch which meant "exit and move point to start of match", or
>>  (closer to your idea) a command which meant "exit and drop mark at start
>>  of match" in place of your toggle.  WDOT?
>
>
> I think a function "exit and drop mark at start of match" would be very
> useful.


Would it disturb if instead mark was always set to start of match?

RE: [PATCH] (Or "question"...) isearch-kill-found

[Drew Adams]

> > I think a function "exit and drop mark at start of match" 
> > would be very useful.
> 
> Would it disturb if instead mark was always set to start of match?

Always? Why?

Why not let users toggle the behavior using `C-SPC', as the proposal suggests?

The implementation I use gives users the choice (1) of the default behavior (an
option) and (2) on the fly (a toggle).

Re: [PATCH] (Or "question"...) isearch-kill-found

[Michael Welsh Duggan]

Lennart Borgman <lennart.borgman@gmail.com> writes:

> On Mon, May 31, 2010 at 12:19 AM, Christoph <cschol2112@googlemail.com> wrote:
>>
>> In that same thread Davis Herring had an idea that sounds interesting and
>> useful to me:
>> http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00492.html
>>
>>>  Thinking about this, I realize
>>>  that I would get almost all the utility of this idea with just one
>>> command
>>>  added to isearch which meant "exit and move point to start of match", or
>>>  (closer to your idea) a command which meant "exit and drop mark at start
>>>  of match" in place of your toggle.  WDOT?
>>
>>
>> I think a function "exit and drop mark at start of match" would be very
>> useful.
>
>
> Would it disturb if instead mark was always set to start of match?

It would.  I am often in a situation where I am at the beginning of a
construct, use C-s to find the end of the construct, and then C-w or M-w
to cut or copy it.  I use this behavior frequently, especially when
writing macros.

-- 
Michael Welsh Duggan
(md5i@md5i.com)

Re: [PATCH] (Or "question"...) isearch-kill-found

[Lennart Borgman]

On Mon, May 31, 2010 at 5:30 AM, Michael Welsh Duggan <md5i@md5i.com> wrote:
> Lennart Borgman <lennart.borgman@gmail.com> writes:
>
>> On Mon, May 31, 2010 at 12:19 AM, Christoph <cschol2112@googlemail.com> wrote:
>>>
>>> In that same thread Davis Herring had an idea that sounds interesting and
>>> useful to me:
>>> http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00492.html
>>>
>>>>  Thinking about this, I realize
>>>>  that I would get almost all the utility of this idea with just one
>>>> command
>>>>  added to isearch which meant "exit and move point to start of match", or
>>>>  (closer to your idea) a command which meant "exit and drop mark at start
>>>>  of match" in place of your toggle.  WDOT?
>>>
>>>
>>> I think a function "exit and drop mark at start of match" would be very
>>> useful.
>>
>>
>> Would it disturb if instead mark was always set to start of match?
>
> It would.  I am often in a situation where I am at the beginning of a
> construct, use C-s to find the end of the construct, and then C-w or M-w
> to cut or copy it.  I use this behavior frequently, especially when
> writing macros.


Thanks Michael, I see. And learned a bit more about using the mark.

RE: [PATCH] (Or "question"...) isearch-kill-found

[Drew Adams]

> > I proposed the feature to emacs-devel on 2007-07-10:
> > http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00490.html
> >
> > The bottom line was Richard's decision:
> >
> >   "I don't think it is needed.  You can type C-@ C-r C-r to put
> >   the region around the match that was just found."
>  
> In that same thread Davis Herring had an idea that sounds interesting 
> and useful to me:
> http://lists.gnu.org/archive/html/emacs-devel/2007-07/msg00492.html
> 
> >  Thinking about this, I realize
> >  that I would get almost all the utility of this idea with 
> >  just one command added to isearch which meant "exit and move
> >  point to start of match", or (closer to your idea) a command
> >  which meant "exit and drop mark at start
> >  of match" in place of your toggle.  WDOT?
> 
> I think a function "exit and drop mark at start of match" 
> would be very useful.

Hm.  That interesting and useful idea that "Davis Herring had" I implemented in
2006, a year before the proposal.  And I pointed the list (hence Davis too) to
that code.

I call that command `set-region-around-search-target'.[*]

(defun set-region-around-search-target ()
  "Set the region around the last search or query-replace target."
  (interactive)
  (case last-command
    ((isearch-forward isearch-backward
      isearch-forward-regexp isearch-backward-regexp)
     (push-mark isearch-other-end t 'activate))
    (t (push-mark (match-beginning 0) t 'activate)))
  (setq deactivate-mark nil))

It does exactly what he described: "drop mark at start of match".  Unlike
`isearchp-set-region' (the proposal), it is not conditional on the user option
or on `transient-mark-mode', and thus not sensitive to the toggle key, `C-SPC'.

I do not bind that command by default, but you can bind it in `isearch-map'.  Or
you can add it, in place of `isearchp-set-region', to `isearch-mode-end-hook'.

It is in fact the _same_ as `isearchp-set-region' (the proposal) in this
context, except that it is unconditional instead of respecting the toggle key.

IOW, if `transient-mark-mode' is on and the option is on, then the proposed
behavior is exactly what Davis described: just put mark at the match start.  If
t-m-mode is off and you still want the start to be marked, then use
`set-region-around-search-target' instead.

Here is what I proposed (I use it on `isearch-mode-end-hook'):

(defun isearchp-set-region ()
  "Set region around search target, if `isearchp-set-region-flag'.
Used only for Transient Mark mode."
  (when (and isearchp-set-region-flag transient-mark-mode)
    (push-mark isearch-other-end t 'activate)))

You might notice, BTW, that Davis started by saying that he does not use either
transient-mark mode or delete-selection mode.  Yet t-m mode is the context the
feature was designed for - it is active only in t-m mode.  (Transient-mark mode
is now the default.)

---
[*] There is perhaps a better test to use here than `last-command'.  Maybe
isearch has a state variable that can be checked for this - dunno.

Re: [PATCH] (Or "question"...) isearch-kill-found

[Juri Linkov]

> Recently, in a discussion on #emacs, we looked for a command which
> kills the text that has been found using isearch. None of us knew if
> there existed anything like that and I coludn't find it.
>
> So attached is a patch which adds isearch-kill-found and binds it to
> M-C-m (not too keen on the choice of binding though).
>
> Any thoughts or comments? Does this already exist?

A command to delete the current match without exiting isearch
would be more useful.  It will allow to continue searching for the
next matches and delete them too:

(define-key isearch-mode-map [delete] 'isearch-delete-current-match)

(defun isearch-delete-current-match ()
  "Kills the region that isearch has found."
  (interactive)
  (kill-region isearch-other-end (point)))

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

Re: [PATCH] (Or "question"...) isearch-kill-found

[Chong Yidong]

Juri Linkov <juri@jurta.org> writes:

> A command to delete the current match without exiting isearch
> would be more useful.  It will allow to continue searching for the
> next matches and delete them too:
>
> (define-key isearch-mode-map [delete] 'isearch-delete-current-match)
>
> (defun isearch-delete-current-match ()
>   "Kills the region that isearch has found."
>   (interactive)
>   (kill-region isearch-other-end (point)))

I think this is a good idea, since [delete] isn't currently used.  Maybe
add C-d also.

Re: [PATCH] (Or "question"...) isearch-kill-found

[Lennart Borgman]

On Mon, May 31, 2010 at 4:24 PM, Chong Yidong <cyd@stupidchicken.com> wrote:
> Juri Linkov <juri@jurta.org> writes:
>
>> A command to delete the current match without exiting isearch
>> would be more useful.  It will allow to continue searching for the
>> next matches and delete them too:
>>
>> (define-key isearch-mode-map [delete] 'isearch-delete-current-match)
>>
>> (defun isearch-delete-current-match ()
>>   "Kills the region that isearch has found."
>>   (interactive)
>>   (kill-region isearch-other-end (point)))
>
> I think this is a good idea, since [delete] isn't currently used.  Maybe
> add C-d also.


Can't you easily do this by for example switching to query-replace
from within isearch?

I think being able to mark the match at exit would be more useful.
(This could also easily be used to delete the matched region.) And
perhaps exiting with a submatch marked instead could be done with
numerical prefixes.

Re: [PATCH] (Or "question"...) isearch-kill-found

[Stefan Monnier]

>> A command to delete the current match without exiting isearch
>> would be more useful.  It will allow to continue searching for the
>> next matches and delete them too:
>> (define-key isearch-mode-map [delete] 'isearch-delete-current-match)
>> (defun isearch-delete-current-match ()
>>   "Kills the region that isearch has found."
>>   (interactive)
>>   (kill-region isearch-other-end (point)))

> I think this is a good idea, since [delete] isn't currently used.  Maybe
> add C-d also.

Every new key we use tends to collide with its current use of "exit
isearch and run the command globally bound to that key".  I.e. people
who currently expect C-d/delete to exit isearch and delete the next char
will be badly surprised.

Now, I haven't heard a proposal yet that doesn't suffer from this kind
of problem, but I think that using a new key should only be done if
there's a good chance that it's very rarely used (C-d might not be very
frequently used to exit isearch, but there's no reason to think it's
rarely used either) *and* if the new functionality is
sufficiently important.  In my experience deleting the matched text is
not that frequent a need.  Yes, I also do it every once in a while, but
I also often want to delete something slightly different.  Or I want to
do something else with it.

So I'd much rather see an "exit and set the mark at isearch-other-end"
command, after which you can do C-w if you want to delete that text (I
could also imagine a "set the mark isearch-other-end but don't exit",
but then deleting the text would require RET C-w instead).

Richard thinks that C-SPC C-r C-r works as well, but for regexp-patterns
it won't always match the same text, and I must say that C-SPC C-r C-r
RET C-w is pretty long winded.  Still, if we want a new
isearch-set-mark-and-exit, we'll need to find a good keybinding for it.
I'm not sure C-SPC is a good choice (OT1H it's pretty common to use
C-SPC to exit isearch, so many uses will be affected, OTOH it will
behave in a similar way, just pushing the mark at the other end of the
search).  Of course, we could provide the command without binding it,
like we've done with isearch-toggle-case-fold, isearch-toggle-regexp and
a few others.


        Stefan

Re: [PATCH] (Or "question"...) isearch-kill-found

[Lennart Borgman]

On Mon, May 31, 2010 at 5:54 PM, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:
>
> So I'd much rather see an "exit and set the mark at isearch-other-end"
> command, after which you can do C-w if you want to delete that text


How about reusing C-M-h? Let it be "exit isearch and put mark at
current point and move point to match beginning".

Typing C-M-h once again if needed is not too hard and would normally
fit despite that point has moved.

RE: [PATCH] (Or "question"...) isearch-kill-found

[Drew Adams]

> Every new key we use tends to collide with its current use of "exit
> isearch and run the command globally bound to that key".  I.e. people
> who currently expect C-d/delete to exit isearch and delete 
> the next char will be badly surprised.
> 
> Now, I haven't heard a proposal yet that doesn't suffer from this kind
> of problem...
>
> I'm not sure C-SPC is a good choice (OT1H it's pretty common to use
> C-SPC to exit isearch, so many uses will be affected,

What you're thinking of: `C-s KEY C-w', where KEY exits isearch and sets the
mark at the search-match start.  You use `C-w' only if you want to delete the
selected text.  In `delete-selection-mode', you could use `C-s KEY abcd' to
replace the text with `abcd'.  Or you could extend the active region (`M-f'
etc.).

If you do the search+kill several times in succession: `C-s KEY C-w C-s C-s KEY
C-w...'.  Likewise, for replacements.

What you say about the limitation regarding the choice of KEY is correct.  But a
single-touch chord or key such as `C-SPC' is not a big deal - you just hit it
twice to get the normal effect (almost - there will be an extra mark on the
ring).  Or hit `RET' and then the key: `RET C-SPC'.

The alternative that I proposed does not affect any keys that exit isearch - you
exit it normally, any way you want.  What it does do is leave the mark-setting
on or off until you toggle it, with `C-SPC' during isearch.

IOW, once you've hit `C-SPC' to turn marking on: `C-s RET C-w' or `C-s RET
abcd', where RET could be any other key that exits isearch.  If you do the
search+kill several times in succession: `C-s RET C-w C-s C-s RET C-w...'.  This
is identical to your case above, with RET (or any exiting key) in place of KEY.

So that's the choice I see:

1. a KEY (e.g. C-SPC) to exit and mark

versus

2. a KEY (e.g. C-SPC) to toggle marking

#1 has the disadvantage that KEY can no longer be used to exit normally.  But
that just means you have to hit KEY twice to get the normal behavior.  (Well,
ALMOST the normal behavior, because the first hit adds a position to the mark
ring.)

#2 has the disadvantage that you need to turn it off if you want to do this only
as a one-off action.  But that just means you need to hit KEY a second time.

The choice I think depends on how often you (a) want to exit normally using KEY
versus (b) how often you want to do this only as a one-off.

For #1, each time you would exit normally using KEY you need to hit KEY twice.
For #2, each time you would set mark as a one-off you need to hit KEY twice.

For #2, if you never use this marking feature, then nothing changes for you.
For #1, if you never use KEY to exit normally, then nothing changes for you.

If you use this feature sometimes, but rarely, then change "never" and "nothing"
to "rarely" and "almost nothing".  For people who use this rarely or never, #2
is preferable.

Re: [PATCH] (Or "question"...) isearch-kill-found

[Stefan Monnier]

> The alternative that I proposed does not affect any keys that exit
> isearch - you exit it normally, any way you want.  What it does do is
> leave the mark-setting on or off until you toggle it, with `C-SPC'
> during isearch.

Eh.... it *does* affect a key that exits isearch: C-SPC.

Of course I also dislike it because it adds a state.  As for simply
providing a variable to change the behavior (without a key-binding to
change it), that's another solution, indeed.  Maybe not a bad one.


        Stefan

RE: [PATCH] (Or "question"...) isearch-kill-found

[Drew Adams]

> > The alternative that I proposed does not affect any keys that exit
> > isearch - you exit it normally, any way you want.  What it does do
> > is leave the mark-setting on or off until you toggle it, with
> > `C-SPC' during isearch.
> 
> Eh.... it *does* affect a key that exits isearch: C-SPC.

Yes. I should have said only that you need not exit specially to get the marking
behavior.

> Of course I also dislike it because it adds a state.  As for simply
> providing a variable to change the behavior (without a key-binding to
> change it), that's another solution, indeed.  Maybe not a bad one.

Why is not having a key to toggle the variable better?
It is the variable that "adds a state", not the toggle key.

Re: [PATCH] (Or "question"...) isearch-kill-found

[Stefan Monnier]

>> Of course I also dislike it because it adds a state.  As for simply
>> providing a variable to change the behavior (without a key-binding to
>> change it), that's another solution, indeed.  Maybe not a bad one.

> Why is not having a key to toggle the variable better?

Because keys are in very short supply.

> It is the variable that "adds a state", not the toggle key.

No: if the variable never changes, then it doesn't contribute to state.


        Stefan

RE: [PATCH] (Or "question"...) isearch-kill-found

[Drew Adams]

> >> Of course I also dislike it because it adds a state.  As for simply
> >> providing a variable to change the behavior (without a 
> >> key-binding to change it), that's another solution, indeed.
> >> Maybe not a bad one.
> 
> > Why is not having a key to toggle the variable better?
> 
> Because keys are in very short supply.

Not really, for a keymap such as isearch.

> > It is the variable that "adds a state", not the toggle key.
> 
> No: if the variable never changes, then it doesn't contribute 
> to state.

;-)

If a variable (in the usual programming sense, not in the math sense) never
changes, then it's not very variable. Why have a variable in that case?

Re: [PATCH] (Or "question"...) isearch-kill-found

[Stefan Monnier]

>> >> Of course I also dislike it because it adds a state.  As for simply
>> >> providing a variable to change the behavior (without a 
>> >> key-binding to change it), that's another solution, indeed.
>> >> Maybe not a bad one.
>> > Why is not having a key to toggle the variable better?
>> Because keys are in very short supply.
> Not really, for a keymap such as isearch.

Huh?  Please suggest free key-bindings, then.  Since keys not bound in
isearch-mode-map lead to exiting isearch and running the corresponding
command in the global map, isearch-mode-map is basically just as full as
the global map.

>> > It is the variable that "adds a state", not the toggle key.
>> No: if the variable never changes, then it doesn't contribute to state.
> ;-)
> If a variable (in the usual programming sense, not in the math sense) never
> changes, then it's not very variable. Why have a variable in that case?

Because it can be customized.


        Stefan

Re: [PATCH] (Or "question"...) isearch-kill-found

[Lennart Borgman]

On Tue, Jun 1, 2010 at 2:56 PM, Stefan Monnier <monnier@iro.umontreal.ca> wrote:
>>> >> Of course I also dislike it because it adds a state.  As for simply
>>> >> providing a variable to change the behavior (without a
>>> >> key-binding to change it), that's another solution, indeed.
>>> >> Maybe not a bad one.
>>> > Why is not having a key to toggle the variable better?
>>> Because keys are in very short supply.
>> Not really, for a keymap such as isearch.
>
> Huh?  Please suggest free key-bindings, then.  Since keys not bound in
> isearch-mode-map lead to exiting isearch and running the corresponding
> command in the global map, isearch-mode-map is basically just as full as
> the global map.


I made another suggestion, reuse C-M-h.

It will not interfere with the global command in a bad manner I
believe. You just have to do it twice in case you want the global
binding. I believe it will be easy getting used to it - as long as the
region is visible...

Re: [PATCH] (Or "question"...) isearch-kill-found

[Stefan Monnier]

>>>> > Why is not having a key to toggle the variable better?
>>>> Because keys are in very short supply.
>>> Not really, for a keymap such as isearch.
>> Huh?  Please suggest free key-bindings, then.  Since keys not bound in
>> isearch-mode-map lead to exiting isearch and running the corresponding
>> command in the global map, isearch-mode-map is basically just as full as
>> the global map.
> I made another suggestion, reuse C-M-h.
> It will not interfere with the global command in a bad manner I
> believe.  You just have to do it twice in case you want the global
> binding.  I believe it will be easy getting used to it - as long as the
> region is visible...

Your choice of C-M-h, which has no particular mnemonic or logical
connection to the functionality you suggest, together with the
argumentation about how its impact would be bearable, is a vivid proof
of my argument: keys are in very short supply.


        Stefan

Re: [PATCH] (Or "question"...) isearch-kill-found

[Lennart Borgman]

On Tue, Jun 1, 2010 at 9:03 PM, Stefan Monnier <monnier@iro.umontreal.ca> wrote:
>>>>> > Why is not having a key to toggle the variable better?
>>>>> Because keys are in very short supply.
>>>> Not really, for a keymap such as isearch.
>>> Huh?  Please suggest free key-bindings, then.  Since keys not bound in
>>> isearch-mode-map lead to exiting isearch and running the corresponding
>>> command in the global map, isearch-mode-map is basically just as full as
>>> the global map.
>> I made another suggestion, reuse C-M-h.
>> It will not interfere with the global command in a bad manner I
>> believe.  You just have to do it twice in case you want the global
>> binding.  I believe it will be easy getting used to it - as long as the
>> region is visible...
>
> Your choice of C-M-h, which has no particular mnemonic or logical
> connection to the functionality you suggest,

Maybe my mnemonics and fuzzy logic are a bit wider then... ;-)

I suggested a key binding that marks something (it is normally
mark-defun) since we (at least I) wanted something to mark the matched
region.

> together with the
> argumentation about how its impact would be bearable, is a vivid proof
> of my argument: keys are in very short supply.

This conclusion is maybe a bit circular since it is the assumption I
started with ;-)

Re: [PATCH] (Or "question"...) isearch-kill-found

[Richard Stallman]

    > Because keys are in very short supply.

    Not really, for a keymap such as isearch.

The characters that have a special meaning in isearch
are started to interfere with the default meaning
that they exit the search and then do their normal job.

Re: [PATCH] (Or "question"...) isearch-kill-found

[Juri Linkov]

> So I'd much rather see an "exit and set the mark at isearch-other-end"
> command, after which you can do C-w if you want to delete that text (I
> could also imagine a "set the mark isearch-other-end but don't exit",
> but then deleting the text would require RET C-w instead).

For "exit and set the mark" we could add a prefix argument to RET
(`isearch-exit') to do that.  I remember some other suggestions
for the prefix argument of `isearch-exit' like keeping lazy-highlighted
matches, so maybe it should provide a set of different prefix arguments.

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