🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.text.tex
Article #11842

Re: [Q] behavior of \input

#11842
From: David Carlisle
Date: Wed, 21 Feb 1996 00:00
20 lines
1465 bytes
> Am I right, or have I done something wrong ?

You are right.

The timing at the beginning of a tabular entry is very delicate, TeX
will expand macros (only expand, like in \edef or \write) until it
finds something that isn't expandable. If that is not one of the special
tabular things like \cr (eg from \\) or \crcr (from \end{tabular}) or
\omit (from \multicolumn) then it starts making a new row.

The TeX primitive \input is in fact expandable so that TeX would start
reading the input file without starting a new row, however the LaTeX
\input command does other stuff (such as checking for { } around its
argument) and this means that it is not expandable, So TeX `sees' some
non-expandable token (probably \futurelet if I remember) from the
internal definition of \input and so starts a new row before reading
the input file.

David
Technical Details
Message-ID: <udybpxszua.fsf@top1.ma.man.ac.uk>
Path: arc1-ca.pugleaf.net!unknown!not-for-mail
References: <4genip$ga5@news-sop.inria.fr>

Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: f9e16,93b62a0c6065a799
X-Google-Attributes: gidf9e16,public
X-Google-Thread: 10ee61,93b62a0c6065a799
X-Google-Attributes: gid10ee61,public
From: David Carlisle <carlisle@ma.man.ac.uk>
Subject: Re: [Q] behavior of \input
Date: 1996/02/21
Message-ID: <udybpxszua.fsf@top1.ma.man.ac.uk>
X-Deja-AN: 140389855
sender: carlisle@top1.ma.man.ac.uk
references: <4genip$ga5@news-sop.inria.fr>
to: greg@cajal.inria.fr (Gregoire Malandain)
organization: Manchester University Mathematics Department
newsgroups: comp.text.tex,fr.comp.text.tex