Re: [Texmacs-dev] Shift modifier, lowercase and uppercase in shortcuts

[Miguel de Benito Delgado]

Checking other shortcuts I see that definitions are case sensitive. We have for 
instance

  ("std z" (undo 0))
  ("std Z" (redo 0))

So, to define Ctrl+Shift+a I need (kbd-map ("C-A" …)). No modification to 
keyPressEvent() is needed and the proper kbd-map is ("C-A" …) not ("C-a" …) 
right? 

If this is the case I don't understand commit 6060.

Best,
--
Miguel de Benito.

On Jan 1, 2013, at 2:53 PM, Miguel de Benito Delgado <address@hidden> wrote:

> Hi all,
> 
>  currently, pressing Ctrl+Shift+a results in the key press "C-A" being sent 
> to TeXmacs instead of "C-S-a". The rewrite is performed inside 
> QTMWidget::keyPressEvent() with
> 
>        if (((char) key) >= 'A' && ((char) key) <= 'Z') {
>          if ((mods & Qt::ShiftModifier) == 0)
>            key= (int) (key + ((int) 'a') - ((int) 'A'));
>        }
>        mods &=~ Qt::ShiftModifier;
> 
> This was part of a patch by Massimiliano and Norbert Nemec (svn 2786).
> 
> I don't understand why this code is so. What bothers me is that the sift 
> modifier is removed unconditionally but the lowercasing is conditional. Are 
> Texmacs shortcuts case sensitive? Should they? I want to define a shortcut 
> "C-S-a", and currently I only can do this with "C-A". I think that the  
> ShiftModifier shouldn't be removed. Changing those 5 lines to:
> 
>        if (((char) key) >= 'A' && ((char) key) <= 'Z')
>            key= (int) (key + ((int) 'a') - ((int) 'A'));
> 
> leaves the modifier and enables us to define shortcuts like "C-S-a". Is there 
> a problem with this?
> 
> Best,
> --
> Miguel