Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!dali.cs.montana.edu!milton!charlie
From: charlie@milton.u.washington.edu (Charles Geyer)
Newsgroups: comp.text
Subject: Re: Why is there no seperator in TeX/LaTeX
Message-ID: <6433@milton.u.washington.edu>
Date: 17 Aug 90 01:55:42 GMT
References: <1990Aug16.214657.16764@sctc.com>
Reply-To: charlie@stat.washington.edu (Charlie Geyer)
Distribution: comp
Organization: UW Statistics, Seattle
Lines: 64
Summary:
Expires:
Sender:
Followup-To:


In article <1990Aug16.214657.16764@sctc.com> herndon@sctc.com (William R.
Herndon) writes:

>    I am curious as to why, in the design of TeX/LaTeX, no character 
>    exists that can be used as a seperator in the names of macros.

Because that's the way it is.  The reason might have been to make $\alpha-z$
make sense.  Or it might not.

>    After reading the TeXBook section on character codes and classes, I 
>    thought that the hypen would be the character to use.  After all it
>    can be used in labels ( using the \label \ref commands ), and is not
>    called out as a reserved character.

But \label and \ref are LaTeX, not plain TeX.  Why would you expect TeX
command sequences to obey the rules that are allowed for LaTeX labels?

>    Well, I was wrong.  After defining about 500 macros, all of which had
>    hyphens in them to help improve the readability of the names, I discovered
>    much to my dismay that TeX obviously does accord some special meaning to
>    the hyphen character and won't allow definitions with the hyphen in 
>    them.  

It didn't occur to test one of these macros before you typed 500 of them?

>    So now what am I to do.  There appearantly exists no character that can
>    be used as a seperator, and I need one for clarity.  I guess that I've 
>    learned my lesson and will no longer take anything that the TeXBook
>    states at face value again.

The TeXbook is the most accurate computer manual I know.  Generally far more
accurate that commercial software manuals.

From page 7 of the TeXbook.

  Control sequences come in two flavors.  The first kind, like
  {\tt\char`\\input}, is called a {\it control word;} it consists of
  an escape character followed by one or more {\it letters,} followed by
  a space of by something besides a letter.  (\TeX\ has to know where the
  control sequence ends, so you must put a space after a control word if
  the next character is a letter.  For example, if you type
  `{\tt\char`\\inputMS}', \TeX\ will naturally interpret this as a
  control word with seven letters.)  In case you're wondering what a
  ``letter'' is, the answer is that \TeX\ normally regards the 52 symbols
  {\tt A...Z} and {\tt a...z} as letters.  The digits
  {\tt 0...9} are {\it not\/} considered to be letters, so they don't
  appear in control sequences of the first kind.

  A control sequence of the other kind, like {\tt\char`\\'}, is called
  a {\it control symbol;}  it consists of the escape character followed
  by a single nonletter.

That's it.  A backslash and one or more upper or lower case letters or a
backslash and a single character.  Nothing else can be a control sequence.
How could it be stated any clearer?

>    Arrrggggghhhhhhh!!!!!

Indeed.

-- 
charlie@stat.washington.edu (Charlie Geyer)
Department of Statistics, University of Washington