Re: .emacs key maps are overwritten by php-mode

[Lennart Borgman]

Sure you do not understand when I give you the wrong answer ;-)

Sorry, my mistake. I thought you where using 'C-c d'. That is a sequence 
reserved for users (see below). If you really want to overwrite a 
binding in a major mode you can do that in the mode hook. Mode hooks are 
run as the last thing when a mode is setup. For php-mode the hook should 
be called php-mode-hook, they are always named like that. Use add-hook 
to add a function that set up the key binding for you. Something like

  (add-hook 'php-mode-hook
          (lambda()
            (local-set-key ....)
           )

Here you should use local-set-key instead of global-set-key. This way 
you change the major modes keymap.


Stefan Horomnea wrote:
> Hello,
>
> I think I don't understand. I have just Control+D, and not Control-C 
> and a Letter.
> If I only find a way to load my .emacs file or some other file with my 
> keybindings after the last mode has loaded and overwrite those ?
>
> Stefan
>
> Lennart Borgman wrote:
> > Stefan Horomnea wrote:
> > > Hello,
> > > I write something in my .emacs file like this:
> > > (global-set-key [(control ?d)] 'kill-ring-save)
> > >
> > > This works fine, in other modes (text, html, etc.). But when I open 
> > > a php file, my Ctrl+d instead of kill-ring-save does 
> > > c-electric-delete-forward .
> > > And I have several commands in my .emacs file, that are broken by 
> > > the loading of the php-mode. What's the catch ?
> > >
> > > Thank you.
> > >
> > > Stefan
> > There is a convention that major modes should use, see "(elisp) Key 
> > Binding Conventions" (maybe this is only find in current CVS Emacs?):
> >
> >   * Please do not define `C-c LETTER' as a key in Lisp programs.
> >     Sequences consisting of `C-c' and a letter (either upper or lower
> >     case) are reserved for users; they are the *only* sequences
> >     reserved for users, so do not block them.
> >
> > So the php-mode you use does not follow this convention. Perhaps it 
> > would be good to contact the author?