calc-undo keybinding

calc-undo keybinding

[joakim]

I like to bind the undo command to c-z.
calc-ext.el rebinds all standard undo keys to calc-undo.

Now, I can of course create my prefered binding in the mode hook, but is
that really the best way? None of the keymap remap commands seemed
suitable, but I'm happily corrected.

-- 
Joakim Verona

Re: calc-undo keybinding

[Stefan Monnier]

> I like to bind the undo command to c-z.

OK.

> calc-ext.el rebinds all standard undo keys to calc-undo.

OK.

> Now, I can of course create my prefered binding in the mode hook, but is
> that really the best way? None of the keymap remap commands seemed
> suitable, but I'm happily corrected.

I missed the part where you describe the problem.  Is it that
calc-ext.el rebinds all standard undo keys to calc-undo but doesn't
rebind your C-z binding?


        Stefan

Re: calc-undo keybinding

[joakim]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> I like to bind the undo command to c-z.
>
> OK.
>
>> calc-ext.el rebinds all standard undo keys to calc-undo.
>
> OK.
>
>> Now, I can of course create my prefered binding in the mode hook, but is
>> that really the best way? None of the keymap remap commands seemed
>> suitable, but I'm happily corrected.
>
> I missed the part where you describe the problem.  Is it that
> calc-ext.el rebinds all standard undo keys to calc-undo but doesn't
> rebind your C-z binding?

Yes. I can fix it in the mode hook, but is this really the way it should
work? It just feels like calc should hook into the existing undo
function.

>
>         Stefan

-- 
Joakim Verona

Re: calc-undo keybinding

[Stefan Monnier]

>> I missed the part where you describe the problem.  Is it that
>> calc-ext.el rebinds all standard undo keys to calc-undo but doesn't
>> rebind your C-z binding?

> Yes. I can fix it in the mode hook, but is this really the way it should
> work? It just feels like calc should hook into the existing undo
> function.

I don't know enough about calc-undo's behavior to know whether it could
work by hooking into Emacs's undo.  Clearly, it was written long before
Emacs's undo provided hooks.

But I guess a change like the patch below would be in order and should
fix the problem for you.


        Stefan


=== modified file 'lisp/calc/calc-ext.el'
--- lisp/calc/calc-ext.el	2011-01-16 02:21:30 +0000
+++ lisp/calc/calc-ext.el	2011-01-19 23:12:20 +0000
@@ -136,8 +136,7 @@
   (define-key calc-mode-map "\C-w" 'calc-kill-region)
   (define-key calc-mode-map "\M-w" 'calc-copy-region-as-kill)
   (define-key calc-mode-map "\M-\C-w" 'kill-ring-save)
-  (define-key calc-mode-map "\C-_" 'calc-undo)
-  (define-key calc-mode-map "\C-xu" 'calc-undo)
+  (substitute-key-definition 'undo 'calc-undo calc-mode-map global-map)
   (define-key calc-mode-map "\M-\C-m" 'calc-last-args)
 
   (define-key calc-mode-map "a" nil)

Re: calc-undo keybinding

[Štěpán Němec]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> I missed the part where you describe the problem.  Is it that
>>> calc-ext.el rebinds all standard undo keys to calc-undo but doesn't
>>> rebind your C-z binding?
>
>> Yes. I can fix it in the mode hook, but is this really the way it should
>> work? It just feels like calc should hook into the existing undo
>> function.
>
> I don't know enough about calc-undo's behavior to know whether it could
> work by hooking into Emacs's undo.  Clearly, it was written long before
> Emacs's undo provided hooks.
>
> But I guess a change like the patch below would be in order and should
> fix the problem for you.
>
>
>         Stefan
>
>
> === modified file 'lisp/calc/calc-ext.el'
> --- lisp/calc/calc-ext.el	2011-01-16 02:21:30 +0000
> +++ lisp/calc/calc-ext.el	2011-01-19 23:12:20 +0000
> @@ -136,8 +136,7 @@
>    (define-key calc-mode-map "\C-w" 'calc-kill-region)
>    (define-key calc-mode-map "\M-w" 'calc-copy-region-as-kill)
>    (define-key calc-mode-map "\M-\C-w" 'kill-ring-save)
> -  (define-key calc-mode-map "\C-_" 'calc-undo)
> -  (define-key calc-mode-map "\C-xu" 'calc-undo)
> +  (substitute-key-definition 'undo 'calc-undo calc-mode-map global-map)
>    (define-key calc-mode-map "\M-\C-m" 'calc-last-args)
>  
>    (define-key calc-mode-map "a" nil)

Why not just use command remapping, i.e.

  (define-key calc-mode-map [remap undo] 'calc-undo)

? (It won't work on XEmacs AFAIK, but that's not a problem for you I'm
guessing.)

  Štěpán

Re: calc-undo keybinding

[Stefan Monnier]

>> --- lisp/calc/calc-ext.el	2011-01-16 02:21:30 +0000
>> +++ lisp/calc/calc-ext.el	2011-01-19 23:12:20 +0000
>> @@ -136,8 +136,7 @@
>> (define-key calc-mode-map "\C-w" 'calc-kill-region)
>> (define-key calc-mode-map "\M-w" 'calc-copy-region-as-kill)
>> (define-key calc-mode-map "\M-\C-w" 'kill-ring-save)
>> -  (define-key calc-mode-map "\C-_" 'calc-undo)
>> -  (define-key calc-mode-map "\C-xu" 'calc-undo)
>> +  (substitute-key-definition 'undo 'calc-undo calc-mode-map global-map)
>> (define-key calc-mode-map "\M-\C-m" 'calc-last-args)
>> 
>> (define-key calc-mode-map "a" nil)

> Why not just use command remapping, i.e.

>   (define-key calc-mode-map [remap undo] 'calc-undo)

> ? (It won't work on XEmacs AFAIK, but that's not a problem for you I'm
> guessing.)

That's an alternative, yes.  Not knowing enough about how what calc-undo
does, I can't judge whether that's preferable (which mostly depends on
whether it's OK to render the real `undo' completely unreachable (other
than via M-x undo) in those buffers).  I'll let the Calc maintainer
figure that one out.


        Stefan

Re: calc-undo keybinding

[Jay Belanger]

Stefan Monnier <monnier@iro.umontreal.ca> writes:
...
>>> +  (substitute-key-definition 'undo 'calc-undo calc-mode-map global-map)
...
>> Why not just use command remapping, i.e.
>
>>   (define-key calc-mode-map [remap undo] 'calc-undo)
...
> That's an alternative, yes.  Not knowing enough about how what calc-undo
> does, I can't judge whether that's preferable (which mostly depends on
> whether it's OK to render the real `undo' completely unreachable (other
> than via M-x undo) in those buffers).

The real `undo' is pretty useless in a Calc buffer; making it
unreachable isn't a problem.  I'm not sure what different effects the
above two remapping commands have, but it sounds like the second one
will work fine.  I'll make the change when I get home.

Jay