[texhax] The last character of a string

Uwe Lueck uwe.lueck at web.de
Fri Jan 29 17:11:21 CET 2010

Michael Barr <barr at math.mcgill.ca> schrieb am 29.01.2010 02:17:23:
> Yes, inside TeX.  Since it pertains only to a few macros, I could search 
> them by hand.  An example is that I might get a paper to edit and it has 
> \subjclass{...}.  If it ends with a period, then I want to set it as is. 
> If it doesn't, I want to add one.  Some authors do the one and some do the 
> other.  Another one is \subsection{...} whose parameter might end in a 
> period, a question mark, or even an exclamation point.  Same issue.  A few 
> authors use \subsubsection.  I am ambivalent about \section since the 
> section headers are separated.  If it were just my own work, I could be 
> consistent, but I am TeX editor for an online journal.

\def\subsection{\let\AfterPunctTest\OrdSubsection \MaybeAddPunct}
%% and the like with:

  \AfterPunctTest{#1\ifPunct \else .\fi}%

HTH -- Uwe. 

P.S.: with optional arguments ...

> On Thu, 28 Jan 2010, Pierre MacKay wrote:
> > On 01/28/2010 03:25 PM, Michael Barr wrote:
> >>  Is there a simple way to find the last character of a string?  Assume the
> >>  string is brace delimited.  What I want to do is add a period unless the
> >>  string already ends in a period, question mark, or exclamation mark, but I
> >>  don't see any way short of going through the string knocking off one
> >>  character at a time.
> >> 
> > I assume that this has to be done within TeX.  Otherwise it is a function of 
> > whatever editor you are using:
> >
> > It is a problem I often find, and I use the emacs "narrow-to-region" command. 
> > Find whatever triggers the opening brace,
> > save the position and then (if there are no intervening paired braces) find 
> > the closing brace. Narrow to region, and then search just before the closing 
> > brace.  It is very fast indeed.
> >
> > The need for this ability to narrow to a specific context is one of the 
> > reasons for providing specific *begin* and *end* macros in many contexts, 
> > rather than insisting that the context appear as a macro parameter.  The 
> > LaTeX adapters of Ibycus have insisted on the macro parameter model, but in 
> > Plain Tex I stick with \GK{} and \RM{} which allow a clean "narrow-to-region" 
> > operation on any passage of any size.  These macros make it possible to run 
> > efficiently through a long article and check the correctness of the Greek 
> > without having to scroll through all the non-Greek text.
> >
> > Pierre MacKay

More information about the texhax mailing list