User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:24.0) Gecko/20100101
            Thunderbird/24.6.0
MIME-Version: 1.0
References: <5489.1404208110@cl.cam.ac.uk> <53B2B100.1070907@morningstar2.co.uk>
Content-Type: text/plain; charset=ISO-8859-1
Message-ID:  <53B2B492.7000903@morningstar2.co.uk>
Date:         Tue, 1 Jul 2014 14:16:02 +0100
Reply-To: Mailing list for the LaTeX3 project
              <LATEX-L@LISTSERV.UNI-HEIDELBERG.DE>
Sender: Mailing list for the LaTeX3 project <LATEX-L@LISTSERV.UNI-HEIDELBERG.DE>
From: Joseph Wright <joseph.wright@MORNINGSTAR2.CO.UK>
Subject: Re: Case changing operations
To: LATEX-L@LISTSERV.UNI-HEIDELBERG.DE
In-Reply-To:  <53B2B100.1070907@morningstar2.co.uk>
Precedence: list
Content-Transfer-Encoding: quoted-printable
Envelope-To: <rainer.schoepf@GMX.NET>
Status: R

On 01/07/2014 14:00, Joseph Wright wrote:
>> does anyone know what context do?  is there scope for joint work on a
>> common module?
>=20
> Reading e.g. http://wiki.contextgarden.net/Titles, the ConTeXt approach
> (as I'd guess) has more than one part. The primitives tend to be more
> 'encouraged' in ConTeXt than is likely to be the case for any
> stand-alone LaTeX2, and indeed than is often the case for LaTeX2e. Thus
> they have some examples using \uppercase, which is (in the version of
> ConTeXt I have installed) the primitive unaltered (I did not check for
> any callback here). There is also e.g. \WORD, which in MkIV uses
> LuaTeX's \attribute system to effect the change at the font level. The
> latter is therefore able to e.g. convert "=DF" to "SS". (In my tests I =
did
> not find another example of a 1 -> many situation which is covered by
> the standard fonts.)  Thus they have somewhat different
> requirements/aims. (Testing here suggests that e.g. \Word will pick up
> "." but does not handle for example quotations.)
>=20
> (Note: That page says \WORD will mess up \em, which may be correct for
> MkII but is wrong for MkIV. I've not checked what MkII does here.)
>=20
> (See also http://wiki.contextgarden.net/Command/setcharactercasing for
> the more general case, which works on a per-paragraph basis, again usin=
g
> attributes to achieve this.)

What of course is notable here is that they've not felt the need to
actually convert stored input (easily) into different cases: nether the
primitive nor attribute approach are expandable.
--=20
Joseph Wright