Caps-mode vs. Overwrite mode

Caps-mode vs. Overwrite mode

[Dan Espen]

Hi,

I'm using caps-mode.el from here:

https://github.com/emacsmirror/caps-mode/blob/master/caps-mode.el

When I toggle caps-mode on, overwrite mode stops working.
That is, inserted characters insert instead of overwriting.

Any hints?

-- 
Dan Espen

Re: Caps-mode vs. Overwrite mode

[Stefan Monnier]

> When I toggle caps-mode on, overwrite mode stops working.
> That is, inserted characters insert instead of overwriting.
> Any hints?

The caps-mode.el that's in GNU ELPA shouldn't suffer from this
problem, AFAIK.


        Stefan

Re: Caps-mode vs. Overwrite mode

[Dan Espen]

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

>> When I toggle caps-mode on, overwrite mode stops working.
>> That is, inserted characters insert instead of overwriting.
>> Any hints?
>
> The caps-mode.el that's in GNU ELPA shouldn't suffer from this
> problem, AFAIK.

My first search only turns up caps-lock.el?

Works better than caps-mode, but does type lower case
when the shift key is used.  I can live with that I think,
might even learn to like it.

So, good enough, thanks!


-- 
Dan Espen

Re: Caps-mode vs. Overwrite mode

[Eli Zaretskii]

> From: Dan Espen <despen@verizon.net>
> Date: Fri, 03 Apr 2015 10:47:03 -0400
> 
> Stefan Monnier <monnier@iro.umontreal.ca> writes:
> 
> >> When I toggle caps-mode on, overwrite mode stops working.
> >> That is, inserted characters insert instead of overwriting.
> >> Any hints?
> >
> > The caps-mode.el that's in GNU ELPA shouldn't suffer from this
> > problem, AFAIK.
> 
> My first search only turns up caps-lock.el?
> 
> Works better than caps-mode, but does type lower case
> when the shift key is used.  I can live with that I think,
> might even learn to like it.

How do you even get caps-mode to work?  If I type any self-inserting
character after turning on caps-mode, I get an error:

  insert-char: Symbol's value as variable is void: last-command-char

Re: Caps-mode vs. Overwrite mode

[Dan Espen]

Dan Espen <despen@verizon.net> writes:

> Stefan Monnier <monnier@iro.umontreal.ca> writes:
>
>>> When I toggle caps-mode on, overwrite mode stops working.
>>> That is, inserted characters insert instead of overwriting.
>>> Any hints?
>>
>> The caps-mode.el that's in GNU ELPA shouldn't suffer from this
>> problem, AFAIK.
>
> My first search only turns up caps-lock.el?
>
> Works better than caps-mode, but does type lower case
> when the shift key is used.  I can live with that I think,
> might even learn to like it.
>
> So, good enough, thanks!

Spoke too soon.
caps-lock mode isn't buffer local.

I want to turn this on only in selected buffers.

It looks like I have to toggle caps-lock mode off to enter
lower case commands in the mini-buffer.

I'm using this for editing HLASM, PL/I, etc, languages that
are traditionally done all upper case.

-- 
Dan Espen

Re: Caps-mode vs. Overwrite mode

[Dan Espen]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Dan Espen <despen@verizon.net>
>> Date: Fri, 03 Apr 2015 10:47:03 -0400
>> 
>> Stefan Monnier <monnier@iro.umontreal.ca> writes:
>> 
>> >> When I toggle caps-mode on, overwrite mode stops working.
>> >> That is, inserted characters insert instead of overwriting.
>> >> Any hints?
>> >
>> > The caps-mode.el that's in GNU ELPA shouldn't suffer from this
>> > problem, AFAIK.
>> 
>> My first search only turns up caps-lock.el?
>> 
>> Works better than caps-mode, but does type lower case
>> when the shift key is used.  I can live with that I think,
>> might even learn to like it.
>
> How do you even get caps-mode to work?  If I type any self-inserting
> character after turning on caps-mode, I get an error:
>
>   insert-char: Symbol's value as variable is void: last-command-char

Good question.
My copy has the same author and version number, but it uses
last-command-event:

(defun caps-mode-self-insert-command (&optional n)
  "Like `self-insert-command', but uppercase the the typed character."
  (interactive "p")
  (message "insert")
  (insert-char (upcase last-command-event) n))


Not sure if I made that change or I got my copy somewhere else.


-- 
Dan Espen

Re: Caps-mode vs. Overwrite mode

[Eli Zaretskii]

> Date: Fri, 03 Apr 2015 18:10:43 +0300
> From: Eli Zaretskii <eliz@gnu.org>
> 
> > From: Dan Espen <despen@verizon.net>
> > Date: Fri, 03 Apr 2015 10:47:03 -0400
> > 
> > Stefan Monnier <monnier@iro.umontreal.ca> writes:
> > 
> > >> When I toggle caps-mode on, overwrite mode stops working.
> > >> That is, inserted characters insert instead of overwriting.
> > >> Any hints?
> > >
> > > The caps-mode.el that's in GNU ELPA shouldn't suffer from this
> > > problem, AFAIK.
> > 
> > My first search only turns up caps-lock.el?
> > 
> > Works better than caps-mode, but does type lower case
> > when the shift key is used.  I can live with that I think,
> > might even learn to like it.
> 
> How do you even get caps-mode to work?  If I type any self-inserting
> character after turning on caps-mode, I get an error:
> 
>   insert-char: Symbol's value as variable is void: last-command-char

OK, I see now: caps-mode should be fixed to use last-command-event
instead of (the now deleted) last-command-char.

And the problem with overwrite mode is that caps-mode doesn't pay
attention to overwrite-mode, it always inserts a character, even if
overwrite-mode is on.

So it's a bug in caps-mode: its caps-mode-self-insert-command should
be smarter, and pay attention to overwrite-mode.

Re: Caps-mode vs. Overwrite mode

[Stefan Monnier]

>> The caps-mode.el that's in GNU ELPA shouldn't suffer from this
>> problem, AFAIK.
> My first search only turns up caps-lock.el?

Ah, right, that's the one.  Sorry for the typo.

> Spoke too soon.
> caps-lock mode isn't buffer local.

Indeed.

You can make it buffer-local with the patch below.

If someone wants to submit a patch which adds buffer-local variant, that
could be included in the GNU ELPA package.


        Stefan


diff --git a/packages/caps-lock/caps-lock.el b/packages/caps-lock/caps-lock.el
index 24729a4..5787ef8 100644
--- a/packages/caps-lock/caps-lock.el
+++ b/packages/caps-lock/caps-lock.el
@@ -29,10 +29,10 @@
 ;;;###autoload
 (define-minor-mode caps-lock-mode
   "Make self-inserting keys invert the capitalization."
-  :global t
+  ;; :global t
   (if caps-lock-mode
-      (add-hook 'pre-command-hook #'caps-lock--pch)
-    (remove-hook 'pre-command-hook #'caps-lock--pch)))
+      (add-hook 'pre-command-hook #'caps-lock--pch nil t)
+    (remove-hook 'pre-command-hook #'caps-lock--pch t)))
 
 (defun caps-lock--pch ()
   (when (and (characterp last-command-event)

Re: Caps-mode vs. Overwrite mode

[Dan Espen]

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

>>> The caps-mode.el that's in GNU ELPA shouldn't suffer from this
>>> problem, AFAIK.
>> My first search only turns up caps-lock.el?
>
> Ah, right, that's the one.  Sorry for the typo.
>
>> Spoke too soon.
>> caps-lock mode isn't buffer local.
>
> Indeed.
>
> You can make it buffer-local with the patch below.
>
> If someone wants to submit a patch which adds buffer-local variant, that
> could be included in the GNU ELPA package.
>
>
>         Stefan
>
>
> diff --git a/packages/caps-lock/caps-lock.el b/packages/caps-lock/caps-lock.el
> index 24729a4..5787ef8 100644
> --- a/packages/caps-lock/caps-lock.el
> +++ b/packages/caps-lock/caps-lock.el
> @@ -29,10 +29,10 @@
>  ;;;###autoload
>  (define-minor-mode caps-lock-mode
>    "Make self-inserting keys invert the capitalization."
> -  :global t
> +  ;; :global t

This change gives me an error.

Debugger entered--Lisp error: (void-variable caps-lock-mode)
  (if caps-lock-mode (add-hook (quote pre-command-hook) (function
  caps-lock--pch)
  nil t) (remove-hook (quote pre-command-hook) (function caps-lock--pch) t))
  (defvar caps-lock-mode 

Without that change, the buffer local issue is fixed.

-- 
Dan Espen

Re: Caps-mode vs. Overwrite mode

[Stefan Monnier]

>> -  :global t
>> +  ;; :global t

Oh, right, sorry it should be ":global nil".


        Stefan

Re: Caps-mode vs. Overwrite mode

[Dan Espen]

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

>>> -  :global t
>>> +  ;; :global t
>
> Oh, right, sorry it should be ":global nil".

Works.  Thanks again.

-- 
Dan Espen