Re: Incorporate package macrostep into Emacs core

[Gerd Möllmann]

Philip Kaludercic <philipk@posteo.net> writes:

> Gerd Möllmann <gerd.moellmann@gmail.com> writes:
>
>> Philip Kaludercic <philipk@posteo.net> writes:
>>
>>> Gerd Möllmann <gerd.moellmann@gmail.com> writes:
>>>
>>>> Philip Kaludercic <philipk@posteo.net> writes:
>>>>
>>>>>> (add-hook 'c-mode-hook #'macrostep-c-mode-hook)
>>>>>
>>>>> This part is suspicious.  First of all, it looks like one is adding a
>>>>> hook to a hook, but this would unconditionally modify
>>>>> c-mode-hook, which I don't think is reliable.  Can we find some other
>>>>> way to update the variables, depending on the major mode?  E.g. in
>>>>> macrostep.el one could try to intern (format "macrostep-%s-init"
>>>>> major-mode) and check if the symbol is fboundp?
>>>>
>>>> Also, it would be nice to support C++ and Objc, which we have in Emacs
>>>> (c-mode-common-hook).
>>>
>>> From what I understand, this feature is based on cmacexp, which in turn
>>> just uses cpp.  
>>
>> What's the problem using cpp?
>
> I don't know what Objc does, but in the case of C++ my understanding was
> that for example Templates were not expanded by C++.  Of course, if that
> is not the intention, then never mind my comment.

The preprocessor is the same for all three languages. 

If by templates, you mean C++ templates -- these have nothing to do with
the preprocessor. They don't work on the level of text.