Differencing a major mode from its derived modes

Differencing a major mode from its derived modes

[chem]

Hi, 
I would like to add allout mode only to text-mode.  That is, I don't
want allout mode to propagate to all the derived modes, like message
mode, etc.

I tried, using a:
 - if test on the major-mode, and it's not working
 - (if  (get major-mode 'derived-mode-parent)..., not working neither.

It seems that the major-mode on which the derived mode is based is
actually loaded.  Thus, all the hooks are made.  Only then, the
customization of the derived mode are done.  Then, I really don't know
how to isolated the parent from its children.  If someone has a
solution, it would be nice.

Thanks                            
                                       
-- 
Chemtov

Re: Differencing a major mode from its derived modes

[Kai Grossjohann]

chem <chemtov@ifrance.com> writes:

> I would like to add allout mode only to text-mode.  That is, I don't
> want allout mode to propagate to all the derived modes, like message
> mode, etc.

message-mode runs message-mode-hook and then text-mode-hook.
text-mode runs text-mode-hook only.

So you could toggle allout mode from both hooks.  This would turn it
on, then off again, in message-mode, which is what you want.

Kai

Re: Differencing a major mode from its derived modes

[chem]

> message-mode runs message-mode-hook and then text-mode-hook.
> text-mode runs text-mode-hook only.
>
> So you could toggle allout mode from both hooks.  This would turn it
> on, then off again, in message-mode, which is what you want.

Thanks, 
I did it already, but there are a lot of modes deriving from text mode,
and I don't want to modify them all.  That's why I wanted a general
way to specify that a particular hook should only apply to the parent
mode, and not to its children.

PS: I try hard to make gnus and message use the MFT header (I post
from the mailing list), I hope I achieve it.  If not, I'm sorry, and
will try harder.  A typo in my emacs.el hinders me from replying
sooner.
-- 
Chemtov

Re: Differencing a major mode from its derived modes

[Kai Grossjohann]

chem <chemtov@ifrance.com> writes:

> I did it already, but there are a lot of modes deriving from text mode,
> and I don't want to modify them all.  That's why I wanted a general
> way to specify that a particular hook should only apply to the parent
> mode, and not to its children.

I forget whether testing the variable major-mode was already
mentioned.  You could write a function that checks (eq major-mode
'text-mode), and only if this is true, then activate allout mode.

Kai

Re: Differencing a major mode from its derived modes

[chem]

> I forget whether testing the variable major-mode was already
> mentioned.  You could write a function that checks (eq major-mode
> 'text-mode), and only if this is true, then activate allout mode.

yes, i tried that, but it only work if the mode is not derived from
text-mode (eg. LaTeX - from aucTeX), but not if it is (eg. message).
In the latter, code like this one :

(add-hook 'text-mode-hook
          '(lambda ()
              (if (eq major-mode 'text-mode)
                  (flyspell-mode 1)
                (turn-on-auto-fill))))

will run the THEN _and_ the ELSE part of the IF expression, while in
the former, only then ELSE part is evaluated.

Furthermore, something like :

(derived-mode-p 'text-mode)

will return true even in text-mode (while it's hardly derived from
text-mode, isn't it?).

So, I found no easy way to discriminate between parent from derived
children mode.  

Thanks again for the support.

PS: Could you tell me if receive my reply twice : on the mailing list
(or the newsgroup) and in your mailbox.  I'm curious to know if all
(of my gnus and message configurations) is working as intended.

-- 
Chemtov

Re: Differencing a major mode from its derived modes

[Kai Grossjohann]

chem <chemtov@ifrance.com> writes:

>> I forget whether testing the variable major-mode was already
>> mentioned.  You could write a function that checks (eq major-mode
>> 'text-mode), and only if this is true, then activate allout mode.
>
> yes, i tried that, but it only work if the mode is not derived from
> text-mode (eg. LaTeX - from aucTeX), but not if it is (eg. message).
> In the latter, code like this one :
>
> (add-hook 'text-mode-hook
>           '(lambda ()
>               (if (eq major-mode 'text-mode)
>                   (flyspell-mode 1)
>                 (turn-on-auto-fill))))
>
> will run the THEN _and_ the ELSE part of the IF expression, while in
> the former, only then ELSE part is evaluated.

I think that's because text-mode-hook is run twice: once as part of
message-mode, and once more because message-mode is derived from
text-mode.

IMHO this should not happen: text-mode-hook should be run only once
in this case.

I wonder if this is really a bug, and if so, how to fix it.

> Furthermore, something like :
>
> (derived-mode-p 'text-mode)
>
> will return true even in text-mode (while it's hardly derived from
> text-mode, isn't it?).

Hm.

> So, I found no easy way to discriminate between parent from derived
> children mode.  
>
> Thanks again for the support.
>
> PS: Could you tell me if receive my reply twice : on the mailing list
> (or the newsgroup) and in your mailbox.  I'm curious to know if all
> (of my gnus and message configurations) is working as intended.

I got it twice.  I deleted the mail.

But note that the mail reply was addressed to help-gnu-emacs, not to
a newsgroup.  Not sure if that is relevant.

Kai