User-Agent: Thunderbird 2.0.0.16 (Windows/20080708)
MIME-Version: 1.0
References: <48D252EC.7030802@morningstar2.co.uk>
            <B54C1B2A-FC44-418C-BC7F-AD5F8DCBDD04@gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Message-ID:  <48D274C5.40504@morningstar2.co.uk>
Date:         Thu, 18 Sep 2008 16:33:25 +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: \scantokens, \everyeof, \endlinechar
To: LATEX-L@LISTSERV.UNI-HEIDELBERG.DE
In-Reply-To:  <B54C1B2A-FC44-418C-BC7F-AD5F8DCBDD04@gmail.com>
Precedence: list
Status: R

Will Robertson wrote:

> Quite the opposite! \scantokens can be really helpful.
> I might suggest a name like
>   \token_rescan:n
> or maybe
>   \tlist_rescan:n
> Opinions?

Based on things like \expandafter = \exp_after:NN, I'd imagine
\token_scan:n or \token_rescan:n make most sense. (I think \tlist...
incorrectly describes the function).

> The other two I'm not so sure where they should go. I'd *probably*
> recommend defining your own wrappers around the :D versions and then we
> can work out how to proceed from there.

I think that, if you want to use these you probably want access to the
raw (e)-TeX primitives.  However, the expl3 documentation make it pretty
clear that outside of the LaTeX3 team no-one should use anything ending
:D.  So there needs to be a non-D name for each one.  For \everyeof, I
guess this is (somewhat) related to I/O, so perhaps:

\l_ior_eof_toks

(based on \ior_eof:NTF, etc.).

\endlinechar is more tricky, clearly an _int, perhaps in tokens or chars?

\l_token_endline_int
\l_char_endline_int

seem possible.

I did think if there was a way to wrap both up along with \scantokens to
avoid the entire issue of naming the other parts.  But I don't think
that is possible.
-- 
Joseph Wright