Re: [AUCTeX-devel] Re: Quoting problems.

[David Kastrup]

Ralf Angeli <address@hidden> writes:

> * David Kastrup (2006-07-25) writes:
>
>>>> Why is there no setting "nil"?
>>>
>>> Which semantics would such a setting have?
>>
>> No quote fontification.
>>
>>> `font-latex-quotes' currently tells font-latex which type of
>>> guillemets is used.  Consequently a setting of nil would mean to
>>> disable fontification of guillemets and leave fontification of other
>>> quotes activated.  However, this seems kind of useless.
>>
>> Why?  If documents don't use guillemets and the fontification gets
>> garbled, why not have a way to turn it off?
>
> Quote fontification can potentially go wrong on other quotation mark
> types, too.
>
>> Why is the variable called font-latex-quotes if it really only
>> works on guillemets?
>
> I didn't choose the name.

That does not mean we need to stay with it if it is inaccurate.  Also
the doc string says something else.

>>> Uhh, I thought multi-char macros can only consist of address@hidden
>>
>> They do.
>>
>>> There is quite some code in AUCTeX relying on this assumption.
>>
>> This is like \chapter*: the real macro is called \<, but if it is
>> called with a suffix of < following it, the \WithSuffix definition
>> gets used.
>>
>> suffix.sty is a fun little style file...
>
> The following code seems to work quite fine even without suffix.sty:
>
> \documentclass{article}
> \def\<<{foo}
> \begin{document}
> \<<
> \end{document}
>
> That makes me a bit nervous.

Oh, but it doesn't work since it destroys the meaning of \<
altogether.  After this definition, _every_ \< has to be followed by
<.  The \WithSuffix definition retains the usual meaning of \<, in
contrast.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum