% -*-latex-*-
% Document name: /u/sy/beebe/emacs/filehdr.ltx
% Creator: Nelson H. F. Beebe [beebe@magna.math.utah.edu]
% Creation Date: Wed Nov  6 20:51:19 1991

%%% ====================================================================
%%%  @LaTeX-file{
%%%     author          = "Nelson H. F. Beebe",
%%%     version         = "1.28",
%%%     date            = "06 March 1996",
%%%     time            = "13:34:24 MST",
%%%     filename        = "filehdr.ltx",
%%%     address         = "Center for Scientific Computing
%%%                        Department of Mathematics
%%%                        University of Utah
%%%                        Salt Lake City, UT 84112
%%%                        USA",
%%%     telephone       = "+1 801 581 5254",
%%%     FAX             = "+1 801 581 4148",
%%%     URL             = "http://www.math.utah.edu/~beebe",
%%%     checksum        = "10578 2582 11023 83232",
%%%     email           = "beebe@math.utah.edu (Internet)",
%%%     codetable       = "ISO/ASCII",
%%%     keywords        = "file header, checksum",
%%%     supported       = "yes",
%%%     docstring       = "This LaTeXinfo document describes
%%%                        filehdr.el, a GNU Emacs support package for
%%%                        the creation and maintenance of standard
%%%                        file headers, such as this one.  It may be
%%%                        processed by LaTeX to produce a typeset
%%%                        document, or by M-x latexinfo-format-buffer
%%%                        in GNU Emacs to produce an info file for
%%%                        on-line documentation.
%%%
%%%                        The checksum field above contains a CRC-16
%%%                        checksum as the first value, followed by the
%%%                        equivalent of the standard UNIX wc (word
%%%                        count) utility output of lines, words, and
%%%                        characters.  This is produced by Robert
%%%                        Solovay's checksum utility.",
%%%  }
%%% ====================================================================

\documentstyle[latexinfo,rcs]{book}

% Comment out to suppress page footers with RCS version info:
\RCSID{$Id: filehdr.ltx,v 1.9 1993/08/30 19:41:05 beebe Exp beebe $}

\renewcommand{\i}[1]{{\em #1}}  % change from \sl to \em
\let\emph=\i

\pagestyle{empty}

\newindex{cp} \c concepts
\newindex{fn} \c functions
\newindex{pg} \c programs
\newindex{tp} \c persons
\newindex{vr} \c variables

\typeout{May need LaTeXinfo submode for latex.el
for dots, emph, et al if Mike Clarkson hasn't
fixed the problem yet.}\c

\c \finalout

\title{
        Standard File Headers
}

\author{
        Nelson H. F. Beebe\\
        Center for Scientific Computing\\
        Department of Mathematics\\
        University of Utah\\
        Salt Lake City, UT 84112\\
        USA\\
        Tel: +1 801 581 5254\\
        FAX: +1 801 581 4148\\
        E-mail: {\tt beebe@math.utah.edu}
}

\date{06 March 1996 \\
Version 1.28}

\c =================================================

\begin{iftex}
\c We need to allow hyphenation at - characters in
\c \tt fonts.
\global\hyphenchar\nintt = `\-
\global\hyphenchar\tentt = `\-
\global\hyphenchar\elvtt = `\-
\global\hyphenchar\twltt = `\-
\end{iftex}

\begin{document}

\bibliographystyle{plain}

\maketitle

\c =================================================

\clearpage
\vspace*{0pt plus 1filll}
Copyright \copyright{} 1991 Free Software Foundation, Inc.

This file documents version 1.28 of the standard
file header support package for GNU Emacs, version
18 or later.\refill

Permission is granted to make and distribute
verbatim copies of this manual provided the
copyright notice and this permission notice are
preserved on all copies.\refill

Permission is granted to process this file through
\TeX{} and print the results, provided the printed
document carries copying permission notice
identical to this one except for the removal of
this paragraph (this paragraph not being relevant
to the printed manual).\refill

Permission is granted to copy and distribute
modified versions of this manual under the
conditions for verbatim copying, provided that the
entire resulting derived work is distributed under
the terms of a permission notice identical to this
one.\refill

Permission is granted to copy and distribute
translations of this manual into another language,
under the above conditions for modified versions,
except that this permission notice may be stated
in a translation approved by the Foundation.\refill

\clearpage
\pagestyle{headings}
\pagenumbering{roman}
\tableofcontents

% \listoftables

\clearpage
\pagenumbering{arabic}

\c =================================================
\c Anything before the \setfilename will not appear
\c in the .info file
\setfilename{filehdr.info}

\node Top, Licensing information, Variable Index, (dir)

\begin{ifinfo}
This file documents version 1.28 of the standard
file header support package for GNU Emacs, version 18
or later.\refill

Copyright (C) 1991 Free Software Foundation, Inc.\refill

Permission is granted to make and distribute
verbatim copies of this manual provided the
copyright notice and this permission notice are
preserved on all copies.\refill

Permission is granted to copy and distribute
modified versions of this manual under the
conditions for verbatim copying, provided that the
entire resulting derived work is distributed under
the terms of a permission notice identical to this
one.\refill

Permission is granted to copy and distribute
translations of this manual into another language,
under the above conditions for modified versions,
except that this permission notice may be stated
in a translation approved by the Foundation.\refill
\end{ifinfo}

\begin{menu}
* Licensing information::     Conditions for use
* Author and version::        Who wrote this
* Background::                The origins of all of this
* What's in a header?::
* Putting it all together::
* Outline of file headers::
* Attribute descriptions::    Details of attributes
* GNU Emacs editing support:: Letting Emacs do the work for you
* Simple customization::      Changing things the easy way
* Advanced customization::    Changing things the hard way
* Bug reporting::             How to report bugs, comments, etc.
* Bibliography::              Literature references
* Concept Index::             General topic index
* Function Index::            Lisp functions
* Person Index::              Individuals cited in this text
* Program Index::             Operating system programs
* Variable Index::            Lisp variables
\end{menu}

\c =================================================

\c ===> \node node-name, next, previous, up <===
\node Licensing information, Author and version, Top, Top

\c =============================================
\c There appears to be a bug in LaTeXinfo 1.3.5
\c \unnumberedsec{xxx} is dropped from the
\c info file.  We cannot use \section*{} in
\c LaTeX mode because tables of contents entries
\c are then lost.
\c =============================================

\begin{ifinfo}
\chapter*{Licensing information}
\end{ifinfo}
\begin{tex}
\unnumbered{Licensing information}
\end{tex}
\cindex{licensing information}
The program currently being distributed that
relates to standard file headers is contained in
the file \file{filehdr.el}.  It consists of
numerous support functions for to the creation and
maintenance of file headers.  This program is
\dfn{free}; this means that everyone is free to
use it and free to redistribute it on a free
basis.\refill

Specifically, we want to make sure that you have
the right to give away copies of the programs that
relate to \file{filehdr.el}, that you receive
source code or else can get it if you want it,
that you can change these programs or use pieces
of them in new free programs, and that you know
you can do these things.\refill

To make sure that everyone has such rights, we
have to forbid you to deprive anyone else of these
rights.  For example, if you distribute copies of
the file \file{filehdr.el}, you must give the
recipients all the rights that you have.  You must
make sure that they, too, receive or can get the
source code.  And you must tell them their
rights.\refill

Also, for our own protection, we must make certain
that everyone finds out that there is no warranty
for the programs that relate to \file{filehdr.el}.
If these programs are modified by someone else and
passed on, we want their recipients to know that
what they have is not what we distributed, so that
any problems introduced by others will not reflect
on our reputation.\refill

The precise conditions of the licenses for the
programs currently being distributed that relate
to \file{filehdr.el} are found in the General Public
Licenses that accompany them.  The programs that
are part of GNU Emacs are covered by the GNU Emacs
copying terms
(\pxref{License, , , emacs, The GNU Emacs Manual}),
and other programs are covered by licenses that
are contained in their source files.\refill

\node Author and version, Background, Licensing information, Top

\c We only include this section in the info file,
\c because the printed manual has a title page
\c with author information.
\begin{ifinfo}
The GNU Emacs Lisp file \file{filehdr.el}
was developed by\refill
\begin{verbatim}
Nelson H. F. Beebe
Center for Scientific Computing
Department of Mathematics
University of Utah
Salt Lake City, UT 84112
USA
Tel: +1 801 581 5254
FAX: +1 801 581 4148
E-mail: beebe@math.utah.edu
URL: http://www.math.utah.edu/~beebe
\end{verbatim}

\noindent
in the summer of 1991, and contributed to the Free
Software Foundation.\refill

This documentation corresponds to version 1.28 of
\file{filehdr.el}.  This should match the code
version that is stored in the Emacs variable
\code{file-header-code-version}.\c
\vindex{file-header-code-version}
The author information is similarly stored in the variable
\code{file-header-code-author}.\refill\c
\vindex{file-header-code-author}
\end{ifinfo}

\node Background, What's in a header?, Author and version, Top

\chapter{Background}

With the rapid spread of the global Internet,
which by 1991 reaches more than a half-million
computers\c
\cpsubindex{Internet}{size of}
all around the world \cite{Lottor:CACM-34-11-21},\c
\tindex{Lotter, Mark}
the opportunities for free exchange of software
and textual data are greatly enhanced.\refill

While this brings exciting new capabilities to
many people, not just those involved in academic
research, it is hampered by several factors.\refill

First, not all network file exchange\c
\cindex{network file exchange}
is error-free.  Electronic mail\c
\cpsubindex{electronic mail}{corruption problems}
systems in particular are notorious for corrupting
information, either by truncation of lines or
message bodies, or by transliteration or other
altering of certain characters.  These problems
are most severe for mail exchanges \emph{between}
major networks, such as between the Internet and
Usenet\c
\cindex{Usenet}
or Bitnet.\refill\c
\cindex{Bitnet}

Second, no standards yet exist for describing the
contents of files.  While this is an area of
research at some academic institutions, the wide
variety of operating systems in use, and the growing
numbers of computers (approaching 100 million
on a world-wide basis in 1991), suggest that such
standards may never exist, any more than products
on the commercial market, from soup to saltines,
have standard labels.\refill

Third, without a record of origin of software and
data, it is impossible for users to verify that
they have up-to-date copies, or to contribute
improvements and additions back to the original
authors.\refill

Fourth, without a standard means of encoding
information in file headers, there is no hope of
automating the process of collecting information
from file headers to produce enhanced file archive
summaries, catalogs, and the like.\refill

During the author's 1991--92 tenure as President of the
\TeX{} Users Group,\c
\cindex{TeX Users Group}
efforts were undertaken to improve the quality and
quantity of electronic distribution of
\TeX{}-related software and data.  While this work
had a narrow focus, it has quite general
ramifications, and the GNU Emacs support code here
is quite general, and capable of handling almost
any type of computer-readable textual
material.\refill

It does \emph{not}, however, address the issue of
exchange of binary (non-textual) data; that has a
number of difficulties associated with it, the two
most severe being rigid formats intolerant of
extension, and machine-specific encoding and byte
order.\refill

During a visit to Heidelberg University\c
\cindex{Heidelberg University}
in June
1990, the author spent a pleasant brain-storming
session that lasted until 3am with a dozen
colleagues (who names, alas, were unrecorded) from
Heidelberg, Mainz, Darmstadt, and Goettingen.\refill

We discussed many things that evening, but one
topic in particular led to this work: an informal
proposal for standard file headers that could
address all of the problems noted above.\refill

\node What's in a header?, Putting it all together, Background, Top

\chapter{What's in a header?}\c
\cpsubindex{file header}{contents}

The Bib\TeX{}\c
\pindex{bibtex}
system for support of bibliographic data bases was
developed by Oren Patashnik\c
\tindex{Patashnik, Oren}
at Stanford University, based on earlier work by
Brian Reid\c
\tindex{Reid, Brian}
at Carnegie Mellon University on the
Scribe document formatting system\c
\cindex{Scribe document formatting system}
\cite{Unilogic:SDP84}.  Bib\TeX{} is described in
Leslie Lamport's\c
\tindex{Lamport, Leslie}
book \cite{Lamport:LDP85} on
\LaTeX{}.\c
\cindex{LaTeX}
It is based on the notion that bibliographic items
can be divided into distinct \emph{classes}:
articles, books, reports, theses, and so
on.\refill

Each class of documents has certain features in
common.  For example, journal articles have
authors, titles, volume numbers, often issue
numbers, page numbers, and dates of publications.
Theses and reports would have the name of an
institution attached.\refill

The number of classes of documents is not fixed;
indeed, it may change with time, or between
cultures and languages.  Thus, a bibliographic
system must be \emph{extensible}.   Bib\TeX{}
provides this critical feature by an
implementation in a programming language that
knows how to parse the general structure of a
bibliographic data base entry, without particular
knowledge of the classes, or attributes of
classes.  That information is instead encoded in a
\emph{style file}, which is written in a much more
compact form that is specialized for its job, and
is presumably easier for users to change than
Bib\TeX{} itself is.\refill

The style file can specify which attributes are
required to be present in a class (e.g.\ a Ph.D.
thesis must have an institution), and which
attributes are optional (a book may or may not
have an International Standard Book Number,\c
\cindex{International Standard Book Number (ISBN)}
ISBN).\refill

Some styles may not require all attributes in a
particular class, so Bib\TeX{} simply
\emph{ignores} attributes not required by the
current style, checking them only cursorily for
proper syntax.\refill

In addition, the style file can specify how individual
bibliographic entries extracted by Bib\TeX{} from
data base files are to be formatted.  In a
typesetting application, this flexibility is
important, because there are a great many
bibliography formatting styles, and each journal
or publisher often has rather strict (and
arbitrary) rules that authors must adhere to.\refill

How does this relate to the question of file
headers?\refill

Clearly, the notion of classes and attributes
applies to all computer files as well.  The class
is the file type, such as Lisp file, Pascal code
file, and national census data file.  The
attributes are things like author(s), author's
address, date of last modification, file name,
revision history, character set name, and so on.\refill

In many operating systems, file naming conventions
have been adopted by which the name encodes
information about the class to which the file
belongs.  For example, if the file name ends in
\file{.c}, it is assumed to contain code written
in the C programming language.  Unfortunately,
few file systems are general enough to permit the
creators of computer files to encode additional
header information that might be more detailed.\refill

Since this additional information cannot be
standardly encoded in the file system, it must be
supplied in some way inside the files themselves.
This is not universally possible, particularly
with binary files.\refill

However, textual data tends to be much more
portable between computer systems, and all
reasonable programming languages and text
processing systems make some provision for
\emph{comments},\c
\cindex{comment}
that is, explanatory material inserted into the
file which is otherwise ignored by the program
which processes the file.\refill

Such comments are generally identified by a unique
start symbol, followed by the comment text, and a
unique end symbol.\refill

The start symbol is usually a particular special
character, or special short character sequence,
not otherwise required in the language in which
the file is encoded.  Sometimes the start symbol
must begin in a certain column of the line, such
as Fortran's \code{C} or \code{*} in column 1, or
is implicitly present at a certain column
(assembly languages for older computers often
decreed something like ``a comment starts in
column 32 of the input record'').\refill

The end symbol is frequently an end-of-line
condition, which need not be an actual character.
This convention is simple, but limits comments to
single lines.  If a comment end symbol other than
end-of-line is chosen, the comment body may span
multiple lines.  Thus, the PL/1 and C programming
languages delimit comments by \code{/*} and
\code{*/}, and Pascal by \code{(*} and \code{*)},
or by paired braces.  Some programming languages
even permit comments to be properly \emph{nested},
so that one can comment out a block of code that
itself contains comments.\refill

Ideally, a comment syntax should be simple, yet
permit \emph{any} processor-representable
characters to appear in the comment text, so as
not to hinder freedom of expression.\refill

In any event, with most programming languages, we
should be able to encode file header information
as comments in such a way that expression is not
restricted, yet both humans and suitable computer
programs can recognize the presence of the file
header.\refill

\node Putting it all together, Outline of file headers, What's in a header?, Top

\chapter{Putting it all together}

The preceding sections have outlined the notions
of \emph{classes}, \emph{attributes}, and
\emph{comment embedding}.  What we want to do is
to borrow the syntax used by Bib\TeX{}\c
\pindex{bibtex}
for bibliography data base files, and encode the
file header as comment body text in whatever
syntax the programming language allows, but to do
so in such a way that it can be readily recognized
by both humans and computer programs.\refill

Thus, in a Fortran file, for which comments run
from a \code{C} in column 1 to end of line, our
file header might look something like this:\refill
\begin{verbatim}
C     @Class{
C       attribute1 = "value1",
C       attribute2 = {value2},
C       attribute3 = {value3 with {extra braces}},
C       attribute4 = {value4 with "quotation marks"},
C       attribute5 = "value5 with ""quotation marks""",
C       ...
C     }
\end{verbatim}

The key to programmatic recognition of the header
is the syntax \emph{name followed by an opening
brace, zero or more attribute assignments, and a
closing brace}.  The attribute value fields can be
enclosed in quotation marks, or balanced
braces, as shown above.\refill

In the event that braces otherwise have special
significance (such as in one of Pascal's comment
forms), other distinct paired delimiters could be
used; in the ASCII character set, this means
parentheses, square brackets, or angle
brackets.\refill

The order of attributes is significant only in the
event of duplications; in such a case, the
\emph{last} value assignment is the one to be
used.  Conventions for the order of attributes
will make file headers easier to read,
however.\refill

Readers familiar with Bib\TeX{}\c
\cindex{citation tag}
will note the absence of a \emph{tag} following
the opening brace.  In the bibliography data base
application, the tag serves as a unique citation
key that can be placed in other documents to
uniquely identify the bibliographic reference.  In
the current application for file headers, we have
no need of such a tag.\refill

For languages in which comments continue from a
start symbol to end of line, it will be useful,
though not essential, to make the comment section
containing the file header more visible.  This can
be done in a variety of ways, such as by doubling
or tripling the comment start symbol, or putting a
distinctive character sequence, like several
asterisks or an arrow, \code{==>}, after it.  The
essential point is that if each line begins with a
comment start symbol, \emph{that same prefix must
be used on every line of the header}.  Not only
does this enhance visibility, it makes it possible
for a relatively simple computer program to
identify the first line of the header and
recognize the comment syntax automatically, and
then collect the remainder of the header by
discarding identical comment prefixes\c
\cpsubindex{comment}{prefix stripping}
from succeeding lines until a complete header has
been collected.\refill

\node Outline of file headers, Attribute descriptions, Putting it all together, Top

\chapter{Outline of file headers}\c
\cpsubindex{file header}{outline}\c
\cindex{outline of file header}

This chapter briefly describes what the file
headers contain: class names, attribute names, and
attribute values.  Each is treated in a separate
section.  Detailed descriptions of attributes will
be found in the next chapter (\pxref{Attribute
descriptions}).\refill

\begin{menu}
* Class names::
* Attribute names::
* Attribute values::
\end{menu}

\node Class names, Attribute names, Outline of file headers, Top

\section{Class names}\c
\cindex{class name}

What should the class name in a file header be?
We want it to be indicative of the file contents,
even to a reader unfamiliar with the computer
system from which it originated.  Here are some
desirable criteria:\refill
\begin{itemize}

\item
The class name should \emph{not} be restricted by
the length constraints of many file systems, and
it should not use abbreviations, because they are
often unintelligible to readers unfamiliar with
the originating computer system, or with the
language in which the header is written.\refill

\item
It must also be possible to generate the class
name automatically from knowledge of what the file
name is, at least for those many classes of files
that are distinguished by particular phrases in
their file names.\refill

\item
Class names must be standard across different
operating systems, so that when files are moved
between such systems, they can be readily
associated with the correct class.\refill

\item
Class names must be recognizable by a simple
computer program, and thus must conform to an
agreed-upon syntax.\refill
\end{itemize}

I therefore propose that class names consist of an
optional at-sign, \code{@}, immediately followed
by an initial letter, optionally followed by
letters, digits, and hyphens, followed by the
phrase \code{-file}.\refill

Letter case may be mixed for readability,
\emph{but is not otherwise significant}:
\code{@LATEX-FILE} and \code{@LaTeX-file}
represent the same file class.\refill

This style of naming is common to many programming
languages.   Hyphens between words improve
readability, while avoiding ambiguities introduced
when spaces are allowed to be part of
names.\refill

\node Attribute names, Attribute values, Class names, Top

\section{Attribute names}\c
\cindex{attribute names}

What file header attributes do we need?  Here are
several that are desirable:
abstract,
author(s),
checksum,
code table,
date,
documentation,
filename,
keywords (for later indexing and cross-referencing),
postal, electronic mail, and WorldWide Web addresses,
and
version.\refill

Attribute names have the same syntax as class
names, except that an at-sign, \code{@}, is never
present.  New attribute names can be added as
needed, with the understanding that the file
header processing software will ignore attributes
that it has not been programmed to deal
with.\refill

\node Attribute values, Outline of file headers, Attribute names, Top

\section{Attribute values}\c
\cindex{attribute values}

What about attribute values?  These are for the
most part arbitrary text strings, usually
delimited by quotation marks.  In the event that
quotation marks are needed in the text itself,
braces (or parentheses, square brackets, or angle
brackets) may be used instead, provided that they
are properly nested.  The value text should
\emph{not} presuppose the existence of any
particular text formatting system;\c
\cpsubindex{attribute value}{no formatting system}
in particular,
it should be understandable to a human reader when
it is displayed in the 95 printable characters of
the ASCII character set.\refill

Attribute values may span multiple lines, and in
most cases, newlines can be treated like spaces.\c
\cpsubindex{attribute value}{newlines in}
However, file header processing software
\emph{must} distinguish between spaces and
newlines, and in some cases, such as for address
values, newlines will be preserved in the
output.\refill

Since file headers are encoded inside language
comments, each line will often begin with a
comment start symbol and white space chosen to
provide neat formatting of the header to enhance
its readability.  Thus, after stripping the
comment start symbol, leading white space (blanks
and horizontal tabs) may be ignored.\refill\c
\cpsubindex{attribute value}{leading white space in}

File header processing software \emph{may} choose
to eliminate common prefix strings consisting of a
comment start symbol and following white space
from successive lines of a single value, but
preserve additional indentation space.  Thus, the
input\refill
\begin{verbatim}
;;; name = "Blah blah blah blah blah blah
;;;         blah blah blah blah blah blah.
;;;
;;;              Blah blah blah blah blah.
;;;         blah blah blah blah blah blah.
;;;
;;;              Blah blah blah blah blah."
\end{verbatim}

\noindent
could produce the value string\refill
\begin{verbatim}
Blah blah blah blah blah blah
blah blah blah blah blah blah.

     Blah blah blah blah blah.
blah blah blah blah blah blah.

     Blah blah blah blah blah.
\end{verbatim}

\noindent
if common prefixes are stripped, or\refill
\begin{verbatim}
Blah blah blah blah blah blah
blah blah blah blah blah blah.

Blah blah blah blah blah.
blah blah blah blah blah blah.

Blah blah blah blah blah.
\end{verbatim}

\noindent
if all leading white space is discarded.\refill

Bib\TeX{}\c
\pindex{bibtex}
adopts that convention that braced
groups inside a value string are protected from
certain actions, such as letter case conversion,
or sorting.  In particular, a single quotation
mark may be enclosed in braces to prevent its
recognition as a value string terminator, assuming
the string was started by a quotation mark.  Since
Bib\TeX{} expects that its output will be
processed by the \TeX{} typesetting system, where
braces serve as grouping commands, and are not
normally themselves printable, this is a
reasonable choice: the value string \code{"A
quotation mark, \{"\}, must be braced"} will be
reduced by \TeX{} to \code{A quotation mark, ",
must be braced}.\refill

In the context of general file headers, this
convention is not reasonable, because the value
strings will not in general be processed by
\TeX{}, but instead, will be treated as verbatim
strings.\refill

Similarly, although the C programming language has
character escape conventions to permit encoding of
non-printable characters in printable form, such
as \code{\back n} for newline and \code{\back t}
for horizontal tab, such usages are undesirable in
the context of general file headers that must
serve for many different programming languages and
file types.\refill

Several programming languages adopt the convention
that a quote inside a quoted string is represented
by an adjacent pair of quotes.  This convention is
easy to understand, requires no additional escape
characters, and permits unrestricted
representation of all printable characters, and of
course, white space (blanks and horizontal tabs).
We adopt this convention for attribute value
strings, but note that since balanced braces
(parentheses, square brackets, angle brackets) can
also be used to delimit value strings, the need
for such doubling will be rare.\refill\c
\cpsubindex{attribute value}{quote characters in}

\node Attribute descriptions, GNU Emacs editing support, Attribute values, Top

\chapter{Attribute descriptions}\c
\cindex{attribute descriptions}

In this chapter, we go into the details of each of
the currently-defined attributes in a standard
file header.  Attributes are treated in
alphabetical order in the following sections; they
need not occur in that order in file headers.\refill

\begin{menu}
* abstract::
* address::
* author::
* checksum::
* codetable::
* date::
* docstring::
* email::
* FAX::
* filename::
* keywords::
* supported::
* telephone::
* time::
* URL::
* version::
* multiple values::
\end{menu}

\node abstract, address, Attribute descriptions, Attribute descriptions

\section{Abstract}\c
\cpsubindex{attribute}{abstract}

The \code{abstract}\c
\cindex{abstract}
attribute can supply a short abstract string to
complement the longer \code{docstring} entry.
This should normally be limited to a single
paragraph.\refill

For example, large research institutes often
prepare an annual publication list with abstracts
of documents prepared by staff members.  With care
in the preparation of the file headers, and
suitable software support, much of that annual
report could be extracted automatically from the
file headers.\refill

\node address, author, abstract, Attribute descriptions

\section{Address}\c
\cpsubindex{attribute}{address}

The \code{address}\c
\cindex{address}
attribute should have a postal
address.  Be sure to include a country in your
address; your file may be shared with users all
around the world.\refill

Here is an example from the file header for this
document:\refill
\begin{verbatim}
%%%     address         = "Center for Scientific Computing
%%%                        Department of Mathematics
%%%                        University of Utah
%%%                        Salt Lake City, UT 84112
%%%                        USA",
\end{verbatim}

\node author, checksum, address, Attribute descriptions

\section{Author}\c
\cpsubindex{attribute}{author}

The \code{author}\c
\cindex{author}
attribute should give the full
name of the author, in the order as it is
conventionally spoken.  In much of the Western
world, the family name goes last.\refill

If there are multiple authors, separate them by
the word \code{and},\c
\cindex{and}
rather than by commas.  The
reason for this is that Bib\TeX{}\c
\cindex{name parsing}
has special algorithms that use this convention to
allow parsing of names in some foreign languages,
as well as names with qualifiers, like \code{Jr.},
and those algorithms could be adapted by other
programs that process file headers.  Even simple
programs could separate the names by splitting at
the word \code{and}.\refill

Here is the \code{author}\c
\cindex{author}
attribute from this
document's file header:\refill
\begin{verbatim}
%%%     author          = "Nelson H. F. Beebe",
\end{verbatim}

\node checksum, codetable, author, Attribute descriptions

\section{Checksum}\c
\cpsubindex{attribute}{checksum}

The background chapter
(\pxref{Background})
noted that it is important to be
able to verify the correctness of files that are
moved between different computing systems.  The
way that this is traditionally handled is to
compute a number which depends in some clever way
on all of the characters in the file, and which
will change, with high probability, if any
character in the file is changed.
Such a number is called a \emph{checksum}.\refill\c
\cindex{checksum}

Good algorithms for computing checksums are not
obvious.  One possibility is to count up the
number of characters, words, and lines; in the
UNIX world, this is easily done with the
\code{wc}\c
\pindex{wc}\c
\cindex{word count}
program.  Another possibility is to just add up
the numerical values of all the characters and use
the resulting sum as the checksum.  Both of these
would change if characters were added or removed,
but they would not change under transposition of
characters, words, or lines.\refill

Consequently, a lot of research has been done on
algorithms for finding checksums, and some have
even achieved international standardization.  One
of these standard algorithms is known as a CRC-16
checksum.  CRC stands for \emph{cyclic redundancy
checksum},\c
\cindex{cyclic redundancy checksum}\c
\cpsubindex{checksum}{cyclic redundancy}
and the redundancy of following it with
the word \emph{checksum} is accepted practice.
The CRC-16 checksum\c
\cpsubindex{checksum}{CRC-16}
is capable of detecting error bursts up to 16
bits, and 99 percent of bursts greater than 16
bits in length.  The checksum number is
represented as a 16-bit unsigned number,
encompassing the range 0 \dots{} 65535.  Thus,
there is roughly one chance in 65535 of an error
not being detected, that is, of two different
files having the same checksum.\refill

Of course, no human should have to compute a
checksum; that is a job for a computer program.
The GNU Emacs support software described in this
document handles the job for you.\refill

We cannot use just any checksum program, however,
for several reasons:\refill
\begin{itemize}

\item
The checksum program must itself be portable and
freely available, because verification of the
checksum may be required on any machine that the
file is transported to.\refill

\item
File formats change from system to system.  On
some file systems, text files are represented by
fixed-length records.  On others, variable
length records include a count of the number of
characters in each line.  On still others, lines
end with character terminator sequences like CR,
LF, or CR LF.\refill

\item
The file must contain the checksum, but somehow,
the checksum itself must not be counted when the
checksum is computed.  Otherwise, we could never
achieve self-consistency: each insertion of a new
checksum would change the checksum.\refill

\item
Because of the varying line representations in
file systems, trailing blanks should not be
included in the checksum.  Such blanks waste
space, and should never be significant; they can
be lost when text is refilled in a line-wrapping
editor, or during electronic mail transmission.
It is a good idea to get rid of them; the Emacs
file header maintenance functions described
elsewhere (\pxref{GNU Emacs editing support})
do this for you automatically.\refill

\item
Horizontal tabs\c
\cindex{tab character}
look like spaces on the computer
display, but are really separate characters.
They are often subject to translation to spaces by
electronic mail systems.  For most text files, you
can safely replace them by blanks, which is easy
to do in Emacs: just mark the whole buffer with
\kbd{C-x h}, and then type \kbd{M-x untabify}.\refill

UNIX \code{Makefile}s\c
\pindex{Makefile}
and \code{troff}\c
\pindex{troff}
files are
notable exceptions to this; tabs are
\emph{significant} and cannot be replaced without
destroying the meaning of those files.  That is
why the GNU Emacs file header maintenance
functions never touch tabs.\refill
\end{itemize}

These considerations make it clear than existing
software for computing checksums just will not do.
I raised these points in an editorial challenge
\cite{Beebe:TB11-4-485-487}\c
\tindex{Beebe, Nelson H. F.}
in the \TeX{} Users Group\c
\cindex{TeX Users Group}
journal, TUGboat,\c
\cindex{TUGboat}
and in the spring of 1991 received a clever
solution from Robert Solovay\c
\tindex{Solovay, Robert}
at the University of California,
Berkeley.\refill

Solovay's program, called simply \code{checksum},\c
\pindex{checksum}
is written in a literate programming
\cindex{literate programming}
language called CWEB.\c
\pindex{CWEB}
The output is C code that conforms
to the 1989 ANSI/ISO C Standard.  In computing the
checksum, it ignores line terminators, and any
previous checksum, and since it has been placed in
the public domain, it solves all of the problems
noted above.  Besides a CRC-16 checksum,\c
\cpsubindex{checksum}{CRC-16}
it also produces counts of characters, words, and
lines.  In the event that \code{checksum}\c
\pindex{checksum}
has not yet been installed, this information can
be compared against the output of the UNIX
\code{wc}\c
\pindex{wc}\c
\cindex{word count}
utility.  \code{wc} is simple enough that it can
easily be reimplemented on any system.\refill

\code{checksum}\c
\pindex{checksum}
also has an option to verify the
correctness of the checksum in a file;\c
\cpsubindex{checksum}{validation of}
you could use this to check for corruption after
transferring a file with standard file headers to
your system.\refill

Although \code{checksum}\c
\pindex{checksum}
can be run manually, the
GNU Emacs support code does it for you, producing
an entry in the file header that looks something
like this:\refill
\begin{verbatim}
%%%     checksum        = "25868 849 3980 28305",
\end{verbatim}

\noindent
The four numbers are the CRC-16 checksum,\c
\cpsubindex{checksum}{CRC-16}
line count,\c
\cindex{line count}
word count,\c
\cindex{word count}
and character count.\c
\cindex{character count}
You must remember that the character count will
change if the file is stored with different line
terminator conventions; the other numbers will
remain constant.\refill

\node codetable, date, checksum, Attribute descriptions

\section{Codetable}\c
\cpsubindex{attribute}{codetable}

In the computing world of the 1990s, two major
character sets are in wide use: EBCDIC\c
\cindex{EBCDIC character set}\c
\cpsubindex{character set}{EBCDIC}
on IBM mainframes and their clones, and
ISO/ASCII\c
\cindex{ASCII character set}\c
\cindex{ISO character set}\c
\cpsubindex{character set}{ISO}\c
\cpsubindex{character set}{ASCII}
on everything else.  EBCDIC is an 8-bit character
set, offering characters in the range 0 \dots{}
255, while ISO/ASCII is a 7-bit character set,
with characters in the range 0 \dots{} 127.  On
most machines, ISO/ASCII text is stored in 8-bit
characters.\refill

In turns of numbers of computers, ISO/ASCII is by
far the most common, since it is the character set
used by all personal computers and
workstations.\refill

Unfortunately, a 128-character set with 95
printable characters and 33 control characters is
inadequate for most non-English languages.
Many European languages require accented
characters or additional letters, and
Chinese,\c
\cindex{Chinese characters}
Japanese,\c
\cindex{Japanese characters}
and Korean\c
\cindex{Korean characters}
have thousands of pictographic
characters.\refill\c
\cpsubindex{character set}{pictographic}

Consequently, computer vendors have dealt with
this by offering ISO `code pages'\c
\cpsubindex{character set}{code pages}
--- variations
in the encoding of characters 128 \dots{} 255, and
sometimes even in the encoding of punctuation
characters in the range 0 \dots{} 127.\refill

Standards bodies are actively working on the
development of a new character set that will
support all, or almost all, of the world's present
and past languages.  One of these efforts is a
16-bit character set called Unicode,\c
\cpsubindex{character set}{Unicode}
and another
is a 32-bit character set called ISO 10646.\c
\cpsubindex{character set}{ISO 10646}
Efforts are now underway to merge these efforts
into a character set called ISO 10646M (M for
merged).\refill\c
\cpsubindex{character set}{ISO 10646M}

Given the speed at which committees work, and the
enormous impact on millions of computers, and
people, of a change in text encoding, it seems
unlikely that the impact of these efforts will be
felt for another decade.\refill

The code page problem, however, does have to be
dealt with.  The standard file headers provide for
this with an attribute entry like\refill
\begin{verbatim}
%%%     codetable       = "ISO/ASCII",
\end{verbatim}

\noindent
If the file is encoded in, say code page
ISO-8859-3, then the header could say that:\refill
\begin{verbatim}
%%%     codetable       = "ISO-8859-3",
\end{verbatim}

\noindent
Of course, if an ASCII file were transferred to a
system with EBCDIC, the file would not be
immediately readable until the character values
were translated to EBCDIC.  The checksum described
in the preceding section would be incorrect, but
at least the fact that the file header stated that
the code was originally ISO/ASCII would explain
any translation peculiarities that cropped up
later.\refill

The attribute name \code{codetable}\c
\cindex{codetable}
was chosen
over \code{codepage} because the latter notion is
restricted to variants of ISO/ASCII.\refill

\node date, docstring, codetable, Attribute descriptions

\section{Date}\c
\cpsubindex{attribute}{date}

Computer files should always carry a date-and-time\c
\cindex{time stamp}\c
\cindex{date stamp}
stamp to record time of the last modification.
Some file systems even store date-and-time stamps
for last read, last write, last backup, and so
on.\refill

Unfortunately, many computers do not have a
reliable time standard, and if they lack a network
connection, have no way of maintaining a correct
one.  Date-and-time stamps are recorded in the
file system, rather than the file itself, and are
usually lost when the file is transferred to
another system.  That is regrettable, but it is a
fact of life we still have to tolerate.\refill

Consequently, a standard file header should carry
a date and time.  The editing support described
here supplies it in the form\refill
\begin{verbatim}
%%%     date            = "07 Oct 1991",
\end{verbatim}

Dates and times are expressed in a variety of
formats that depend on the country and culture.\c
\cpsubindex{date}{cultural dependence}\c
\cpsubindex{time}{cultural dependence}
Some software can deal with a considerable variety
of formats, ranging from ``last Wednesday'' to
``1991.11.06:12.34.17''.  The important point is
that the encoding \emph{must be unambiguous}.  In
particular, forms like \code{12/06/91} should be
avoided: does it mean the 12th day of the 6th
month, or the 6th day of the 12th month?  The year
should \emph{not} be abbreviated to two digits;
the new millenium is not far away.\refill

\node docstring, email, date, Attribute descriptions

\section{Docstring}\c
\cpsubindex{attribute}{docstring}

For the purposes of cataloging files, and
recognizing their contents, it is helpful to have
a few paragraphs of description.  This is provided
for by the \code{docstring}\c
\cindex{docstring}
attribute, which might
look like this:\refill
\begin{verbatim}
%%%     docstring       = "This LaTeXinfo document describes
%%%                        filehdr.el, a GNU Emacs support
%%%                        package for the creation and
%%%                        maintenance of standard file
%%%                        headers, such as this one.  It
%%%                        may be processed by LaTeX to
%%%                        produce a typeset document, or by
%%%                        M-x latexinfo-format-buffer in
%%%                        GNU Emacs to produce an info file
%%%                        for on-line documentation.
%%%
%%%                        The checksum field above contains
%%%                        a CRC-16 checksum as the first
%%%                        value, followed by the equivalent
%%%                        of the standard UNIX wc (word
%%%                        count) utility output of lines,
%%%                        words, and characters.  This is
%%%                        produced by Robert Solovay's
%%%                        checksum utility.",
\end{verbatim}

This documentation need not be a user's manual for
the file, unless the necessary information can be
communicated in a few paragraphs of no more than a
couple of thousand characters.  Think of it
instead as an extended abstract.\refill\c
\cindex{abstract}\c
\cpsubindex{documentation string}{as abstract}

Someday, we may have tools that will extract
documentation strings from standard file headers
and turn them into catalogs.\refill

\node email, FAX, docstring, Attribute descriptions

\section{Email}\c
\cpsubindex{attribute}{email}

People who exchange computer files now often have
network access, and the worldwide Internet is
growing rapidly.  It will not be long before
network connections are as commonplace, and
important, as telephone connections now are.  Most
networks support electronic mail, and the trend is
to develop uniform addressing schemes that will
work the world over.  Thus, an electronic mail\c
\cindex{electronic mail}
address, when available, is as important as a
postal address\c
\cindex{postal address}
for the author(s).\refill

Here is an example:\refill
\begin{verbatim}
%%%     email           = "beebe@math.utah.edu (Internet)",
\end{verbatim}

\noindent
Since there are several networks in existence,
with different naming conventions, it is helpful
to identify the network as in this example.\refill

In the event that there are multiple authors,
electronic mail addresses should be given in the
same order, separated by the word \code{and}, just
the way the author attribute value is coded.  Of
course, not all of the authors might have such an
address, so additional qualification, such as by a
parenthesized set of initials, could follow each
address.  Use your ingenuity, but in such a way
that someone you've never met will still
understand what you mean.\refill

\node FAX, filename, email, Attribute descriptions

\section{FAX}\c
\cpsubindex{attribute}{fax}

The \code{FAX}\c
\cindex{FAX}
attribute should be formatted just
like the \code{telephone}\c
\cindex{telephone}
entry.  Here is an
example:\refill
\begin{verbatim}
%%%     FAX             = "+1 801 581 4148",
\end{verbatim}

FAX machines are now very commonly used in
business throughout the world, so if you have such
a facility, it is a good idea to include it in the
file header.\refill

\node filename, keywords, FAX, Attribute descriptions

\section{Filename}\c
\cpsubindex{attribute}{filename}

Different computing systems have different file
naming conventions; in particular, there are
significant variations in the naming of files.
Some systems, like the Apple Macintosh, permit
arbitrary strings of characters, including blanks.
Others, like MS DOS on the IBM PC and clones,
limit names to two parts, a base name and an
extension, or type, with the two separated by a
period (dot, full stop).\refill

File headers should therefore carry an indication
of the original name of the file, and if the file
is expected to be referenced by other files, then
it is \emph{imperative} that the name chosen be
representable on a wide variety of, and preferably
all, computing systems.  Today, this in practice
means the 8-character base name and 3-character
file extension of MS DOS, which runs in tens of
millions of personal computers.\c
\cpsubindex{filename}{portable subset}
There are still a few survivors of older operating
systems with more stringent requirements on file
names, but they are obsolete and rapidly
disappearing.\refill

The filename should be case \emph{insensitive},\c
\cpsubindex{filename}{case insensitivity}
and in the header, spelled in lower-case letters.
It should start with a letter, and use only
letters, digits, and perhaps, hyphens (minus
signs) in the rest of the name, with no more than
a single period in the name.\refill\c
\cpsubindex{filename}{characters allowed in}

This document's file header contains the attribute
entry\refill
\begin{verbatim}
%%%     filename        = "filehdr.ltx",
\end{verbatim}

\noindent
\code{filehdr} is an abbreviation for ``file
header'', and \code{ltx} for ``\LaTeX{}'',\c
\cindex{LaTeX}
the name of the document formatting system that
typesets this document.\refill

\node keywords, supported, filename, Attribute descriptions

\section{Keywords}\c
\cpsubindex{attribute}{keywords}

Large archives always pose a search problem for
human users, and it has long been traditional to
try to classify members of the archives by
\emph{keywords} that might come to mind when
someone is searching for the file.  Some journals
have standard sets of keywords to classify
articles by, and include them near the abstract of
each paper.\refill

With standard file headers, the range of possible
keywords is enormous, and authors will just have
to be diligent about finding good sets of
descriptive keywords.  They should appear in the
attribute value as phrases separated by commas, as
for this document:\refill
\begin{verbatim}
%%%     keywords        = "file header, checksum",
\end{verbatim}

\node supported, telephone, keywords, Attribute descriptions

\section{Supported}\c
\cpsubindex{attribute}{supported}

All computer files reach a stage of stagnation,
where for various reasons, their authors no longer
maintain them.  Nevertheless, it is helpful to
know whether the author of a given file is
interested in hearing of problems or comments, and
the file header can say so by an entry like this
one:\refill
\begin{verbatim}
%%%     supported       = "yes",
\end{verbatim}

If it says \code{yes}, this does not provide any
guarantee that any problems reported will be
fixed, but just that the author's intentions are
good, and reasonable efforts will be made to do
so.  Some authors even care so much about their
work that they offer monetary rewards for reports
of bugs and errors.\refill

If it says \code{no}, then you are on your own,
because the author never wants to hear from you on
the subject of this particular file.\refill

Other attribute values can be readily imagined,
like \code{only for money, cash in advance}, but a
simple \code{yes} or \code{no} is probably
adequate for most people.\refill

\node telephone, time, supported, Attribute descriptions

\section{Telephone}\c
\cpsubindex{attribute}{telephone}

The \code{telephone}\c
\cindex{telephone}
attribute should include the
area code with telephone number.  If there are
multiple values, separate them by commas.  Here is
an example from the file header of this document:\refill
\begin{verbatim}
%%%     telephone       = "+1 801 581 5254",
\end{verbatim}
Use the international form of the number,
including the country and city\slash area code.

\node time, version, telephone, Attribute descriptions

\section{Time}\c
\cpsubindex{attribute}{time}

The \code{time}\c
\cindex{time}
attribute should be of the form
\code{hh:mm:ss}, or if a time zone abbreviation
(say, \code{GMT}) can be found, \code{hh:mm:ss
GMT}.  It is recorded separately from the
\code{date}\c
\cindex{date}
to ease the parsing job of software
that processes file headers.\refill

Here is a typical example:\refill
\begin{verbatim}
%%%     time            = "18:02:38 MST",
\end{verbatim}

\node version, multiple values, time, Attribute descriptions

\section{URL}\c
\cpsubindex{attribute}{URL}

Since its introduction in the early 1990s, the
WorldWide Web has spread rapidly, so that most
public interest in the Internet is associated with
it, and so that most Internet sites that
previously had electronic mail, ftp, and telnet
services, now also have a WorldWide Web presence.\refill

The Uniform Resource Locator, or URL, is therefore
a suitable addition to the standard file headers;
the one in this file looks like this:\refill
\begin{verbatim}
%%%     URL             = "http://www.math.utah.edu/~beebe",
\end{verbatim}

Since most sites have found it convenient to name
a particular machine with the prefix ``www.'',
from an electronic mail address one can often
guess what the corresponding URL should be.
Nevertheless, the host with that name is often
different from the login host, so the Emacs code
in \file{filehdr.el} may not successfully identify
it automatically. Thus, you can provide an
overriding private definition like this in your
\file{.emacs} startup file:\refill
\begin{verbatim}
(setq file-header-user-URL "http://www.math.utah.edu/~beebe")
\end{verbatim}
\vindex{file-header-user-URL}


\section{Version}\c
\cpsubindex{attribute}{version}

Computer files created by humans almost inevitably
go through many revisions, whether they are
programs to control a satellite, or just the words
of a promotion for the latest soap product.\refill

Computer vendors have long dealt with this by
attaching \emph{version numbers}\c
\cindex{version number}
to software releases.  These consist of two or
three numbers with some separator character, such
as a period (full stop, dot).  The first number is
called the \emph{major version number}; it gets
changed only at long intervals, usually years,
when really significant changes have been
incorporated.  A second number is a \emph{minor
version number} which is incremented as smaller
changes and bug fixes are incorporated.  Sometimes
a third number is appended, which is an \emph{edit
number}; it gets incremented every time any change
at all is made to the file.\refill

In careful software production, a change log\c
\cindex{change log}
is kept to record the reasons for every change;
this is particularly important when commercial
interests or legal issues are at stake.  [Military
organizations the world over are famous for their
paperwork trails; perhaps that is what helps to
keep them busy during times of peace.]\refill

For smaller files, you can probably get by with
just a major version number and an edit number;
for larger projects, three or more are
recommended.\refill

Here is what one version of this document had in
its standard file header:\refill
\begin{verbatim}
%%%     version         = "1.01",
\end{verbatim}

Version numbers are particularly useful when
reporting problems to the author of a file; they
allow rapid verification of whether the author and
end user are even talking about the same
thing.\refill

\node multiple values, Attribute descriptions, version, Attribute descriptions

\section{Multiple values}\c
\cpsubindex{attribute}{multiple values}

Keywords like \code{author}\c
\cindex{author}
and \code{address}\c
\cindex{address}
may be inadequate for files prepared by more than
one person.  If several authors share a common
address, then using the keyword \code{and},\c
\cindex{and}
to separate names in the \code{author}\c
\cindex{author}
field is unambiguous.  However, if the postal
address, electronic mail address, telephone
number, and FAX number vary, it is advisable to
clarify the header by attaching a hyphen and a
numeric suffix to the attribute name.  Here is an
example:\refill
\begin{verbatim}
%%%     author-1        = "Marie Claire LeBrun",
%%%     author-2        = "Hans Peter Brun",
%%%     author-3        = "Jill Brown",
%%%     address-1       = "...",
%%%     address-2       = "...",
%%%     address-3       = "...",
%%%     email-1         = "...",
%%%     email-2         = "...",
%%%     email-3         = "...",
%%%     telephone-1     = "...",
%%%     telephone-2     = "...",
%%%     telephone-3     = "...",
%%%     FAX             = "...",
\end{verbatim}

File-header parsing software must be prepared to
handle numeric suffixes like this for any keyword.
If a keyword doesn't have such a suffix, as the
\code{FAX}\c
\cindex{FAX}
keyword in this example, then it should be assumed
to apply to all authors.\refill

\node GNU Emacs editing support, Simple customization, Attribute descriptions, Top

\chapter{GNU Emacs editing support}\c
\cindex{editing support}\c
\cindex{Emacs editing support}\c
\cindex{GNU Emacs editing support}

The preceding chapters have outlined the
background for, and contents of, standard file
headers.  Here we show how to generate them with
very little effort.\refill

The GNU Emacs file
\file{filehdr.el} contains the following
user-callable functions:\refill
\begin{verbatim}
make-file-header
show-file-header-variables
test-file-header
update-checksum
update-date
update-date-and-minor-version
update-file-header-and-save
update-major-version
update-minor-version
update-simple-checksum
\end{verbatim}

There are several other functions in that file,
but they are for internal use only, and will not
be further documented here.\refill

When you want to add a new file header to an
existing file, you just type \kbd{M-x
make-file-header}; this produces something like
this at the top of your file:\refill
\begin{verbatim}
%%% ====================================================================
%%%  @LaTeX-file{
%%%     author          = "Nelson H. F. Beebe",
%%%     version         = "1.28",
%%%     date            = "06 March 1996",
%%%     time            = "13:14:03 MST",
%%%     filename        = "filehdr.ltx",
%%%     address         = "Center for Scientific Computing
%%%                        Department of Mathematics
%%%                        University of Utah
%%%                        Salt Lake City, UT 84112
%%%                        USA",
%%%     telephone       = "+1 801 581 5254",
%%%     FAX             = "+1 801 581 4148",
%%%     URL             = "http://www.math.utah.edu/~beebe",
%%%     checksum        = "53883 2543 10843 81774",
%%%     email           = "beebe@math.utah.edu (Internet)",
%%%     codetable       = "ISO/ASCII",
%%%     keywords        = "file header, checksum",
%%%     supported       = "yes",
%%%     docstring       = "This LaTeXinfo document describes
%%%                        filehdr.el, a GNU Emacs support package for
%%%                        the creation and maintenance of standard
%%%                        file headers, such as this one.  It may be
%%%                        processed by LaTeX to produce a typeset
%%%                        document, or by M-x latexinfo-format-buffer
%%%                        in GNU Emacs to produce an info file for
%%%                        on-line documentation.
%%%
%%%                        The checksum field above contains a CRC-16
%%%                        checksum as the first value, followed by the
%%%                        equivalent of the standard UNIX wc (word
%%%                        count) utility output of lines, words, and
%%%                        characters.  This is produced by Robert
%%%                        Solovay's checksum utility.",
%%%  }
%%% ====================================================================
\end{verbatim}

\noindent
Where does it get all of this information?  Well,
the file name, date and time stamps, author name,
electronic mail address, and date are all
determined automatically from calls to various
system services.  For example, on UNIX, the author
name comes from the file \file{/etc/passwd}; on
VAX VMS, it will come from the file
\file{SYS$MANAGER:SYSUAF.DAT}.\refill

The comment syntax was determined from the file
extension, and we'll say more about it
later.\refill

The only information above that Emacs cannot
determine is your postal address,\c
\cpsubindex{postal address}{defining}
and telephone\c
\cpsubindex{telephone number}{defining}
and FAX numbers, and possibly,  your WorldWide Web
URL.\c
\cpsubindex{FAX number}{defining}
These only have to be supplied once, usually in
your GNU Emacs startup file, \file{.emacs}.  This
is most easily done with Lisp code that looks
something like this:\refill
\begin{verbatim}
(setq file-header-user-address  ; for M-x make-file-header
  "Center for Scientific Computing
Department of Mathematics
University of Utah
Salt Lake City, UT 84112
USA")

(setq file-header-user-telephone "+1 801 581 5254")

(setq file-header-user-FAX "+1 801 581 4148")

(setq file-header-user-URL "http://www.math.utah.edu/~beebe")
\end{verbatim}
\vindex{file-header-user-address}\c
\vindex{file-header-user-telephone}\c
\vindex{file-header-user-FAX}
\vindex{file-header-user-URL}

Once this is installed in the \file{.emacs} file,
GNU Emacs will find it every time it starts up.\refill

If the electronic-mail address constructed from
the Emacs
\code{user-login-name}\c
\findex{user-login-name}
and
\code{system-name}\c
\findex{system-name}
functions is not suitable, you can provide an
alternative one like this:\refill
\begin{verbatim}
(setq file-header-user-email "beebe@math.utah.edu")
\end{verbatim}
\vindex{file-header-user-email}\c

\noindent
In any of the following situations, you
should set \code{file-header-user-email}\c
\vindex{file-header-user-email}
in your startup \file{.emacs} file.\refill
\begin{itemize}

\item
You work on multiple machines, but prefer to have
only one public electronic-mail address.\refill

\item
At some sites, \code{system-name}\c
\findex{system-name}
does not return a fully-qualified Internet host
name, so the default address constructed by
\code{file-header-email}\c
\findex{file-header-email}
is unusable outside your local
installation.\refill

\item
Your site is not on the Internet, but you can
receive electronic mail via some other
network.\refill
\end{itemize}

The version number is left empty; you can manually
insert an appropriate one, perhaps 1.00, or if you
are just starting, 0.00.\refill

The checksum and keywords entries are also left
empty.  There is no point in inserting a checksum
until you are ready to save the file, and the
keywords have to be supplied by a human.\refill

Now suppose you've just edited a file with such a
file header, and you would like to update the
header to reflect the changes, and then save the
file.  All you need to type is \kbd{M-x
update-file-header-and-save}, and with Emacs'
normal command completion, you can probably hit
the tab key after the \kbd{f} in \kbd{file}.\refill

The function \code{update-file-header-and-save}\c
\findex{update-file-header-and-save}
will update the date and time stamps, the minor
version number, the checksum, and save the
file.\refill

If the file is a \LaTeX{}\c
\cpsubindex{LaTeX}{date update}
file, the date update
will also search forward for text that looks
something like\refill
\begin{verbatim}
\\date{29 November 1991 \\
Version 1.01}
\end{verbatim}

\noindent
and change it to the current date and version.
That makes it easy to get the version number and
revision date printed on the title page.\refill

You can do these updates manually if you like by
invoking the functions
\code{update-checksum},\c
\findex{update-checksum}
\code{update-date},\c
\findex{update-date}
\code{update-minor-version},\c
\findex{update-minor-version}
and
\code{update-date-and-minor-version}\c
\findex{update-date-and-minor-version}
explicitly.\refill

Major version numbers are rarely changed, and you
could easily do the job manually.  Nevertheless,
for completeness, \code{update-major-version}\c
\findex{update-major-version}
is supplied to automate the job.\refill

\code{update-checksum}\c
\findex{update-checksum}
will trim trailing
whitespace\c
\cpsubindex{whitespace}{discarding trailing}
(but leave embedded tabs intact), send the buffer
to the \file{checksum} program, and replace it
with the output.  Don't interrupt it while it is
working, or you might lose your file!\refill
\typeout{Check interruption of update-checksum;
maybe use save-for-undo}\c
\typeout{Maybe should have man pages for checksum
in an appendix?}\c

The Emacs interface to \file{checksum} has not yet
been tested on VAX VMS,\c
\cindex{VAX VMS}
so
\code{update-checksum}\c
\findex{update-checksum}
on that system calls
\code{update-simple-checksum}\c
\findex{update-simple-checksum}
instead.  That
function will compute counts of lines, words, and
characters and insert them in the checksum value.
You could use this if for some reason you don't
have \code{checksum}\c
\pindex{checksum}
installed yet.
\code{checksum} should be available from the same
place you got \file{filehdr.el}; eventually it
will be on dozens of \TeX{} archive machines
around the world.\refill

\node Simple customization, Advanced customization, GNU Emacs editing support, Top

\chapter{Simple customization}\c
\cpsubindex{customization}{simple}

The GNU Emacs Lisp code in \file{filehdr.el} has
been written to make it easy to customize without
your having to become a Lisp programmer.  Of
course, Lisp is so much fun that you might want to
do that anyway!\refill

The code contains several large tables stored in Lisp
variables:\refill
\begin{verbatim}
file-header-standard-at-sign-special-cases
file-header-standard-comment-prefixes
file-header-standard-entries
file-header-standard-paired-comment-delimiter-languages
file-header-standard-suffix-and-type
\end{verbatim}

\noindent
These are not intended to be modified by users, as
the phrase \code{-standard-} in their names
indicates.\refill

Each of them is a list of lists; the innermost
lists contain two or three character strings.
Sublists are ordered alphabetically for human
readability; the code does not care what order
they appear in.\refill

The first of them,
\code{file-header-standard-at-sign-special-cases},\c
\vindex{file-header-standard-at-sign-special-cases}
is used to handle those few exceptional file
classes that do not permit at-signs, \code{@}, to
be used in comments without special handling.
Here is the current value of this variable:\refill
\begin{verbatim}
(
    ("BibTeX"               " at ")
    ("C-Web"                "@@")
    ("Web"                  "@@")
    ("Web-change"           "@@")
)
\end{verbatim}

\noindent
This means that when a header for a file in class
\samp{BibTeX} is created, at-signs should be
replaced by the string \samp{ at }.  For the other
classes, at-signs must be doubled.\refill

The second variable,
\code{file-header-standard-comment-prefixes},\c
\vindex{file-header-standard-comment-prefixes}
has
a very long value, so we show only a portion
here:\refill
\begin{verbatim}
(
    ("Adobe-Font-Metric"    "Comment ")
    ("AmSTeX"               "%%% ")
    ("Awk"                  "### ")
...
    ("Web-change"           "%%% ")
    ("Yacc"                 "")
)
\end{verbatim}

\noindent
This means that in an Adobe Font Metric file,\c
\cindex{Adobe Font Metric file}
comments must begin a line with the string
\samp{Comment }.  For \code{awk}\c
\pindex{awk}
files, a triple
sharp sign and a space will begin all file header
lines.  \code{yacc}\c
\pindex{yacc}
file headers have no comment
prefix at all.\refill

The third variable,
\code{file-header-standard-entries},\c
\vindex{file-header-standard-entries}
contains pairs of entry names and functions to
supply values for them.  It looks something like
this: \refill
\begin{verbatim}
(
 ("author"       file-header-author)
 ("version"      file-header-version)
 ("date"         file-header-date)
 ("time"         file-header-time)
 ("filename"     file-header-filename)
 ("address"      file-header-address)
 ("telephone"    file-header-telephone)
 ("FAX"          file-header-FAX)
 ("URL"          file-header-URL)
 ("checksum"     file-header-checksum)
 ("email"        file-header-email)
 ("codetable"    file-header-codetable)
 ("keywords"     file-header-keywords)
 ("supported"    file-header-supported)
 ("docstring"    file-header-docstring)
)
\end{verbatim}

\noindent
The file header is created by processing these
entry names in order.\refill

The fourth variable, with the name
\code{file-header-standard-paired-comment-delimiter-languages},\c
\vindex{file-header-standard-paired-comment-delimiter-languages}
is a little more complex.  Its classes cover
languages that use distinct starting and ending
comment strings, instead of having comments that
terminate at end of line.  For each class name,
its list entries contain two strings, one for
the comment start, and one for the comment end.
To help make them stand out better, the strings
are often stretched to 72 characters in
length:\refill
\begin{verbatim}
(
    ("C"
     (concat "/*"  (make-string 70 ?\*) "\n")
     (concat (make-string 70 ?\*) "*/\n"))

    ("Font-Property-List"
      (concat "(COMMENT "(make-string 63 ?\*) "\n")
      (concat (make-string 71 ?\*) ")\n"))
...
    ("Scribe"
      "@Begin{Comment}\n"
      "@End{Comment}\n")
...
    ("Yacc"
     (concat " /*" (make-string 69 ?\*) "\n")
     (concat " " (make-string 69 ?\*) "*/\n"))
    )
)
\end{verbatim}

\noindent
To avoid the need for long constant strings in the
code, several of them are generated dynamically
by the Lisp concatenation operator,
\code{concat}.\refill\c
\findex{concat}

Class names in this variable do \emph{not} include
the phrase \code{-file} that appears in the file
header; that suffix is supplied automatically by
the Emacs functions.\refill

The last variable,
\code{file-header-standard-suffix-and-type},\c
\vindex{file-header-standard-suffix-and-type}
is the biggest of them all.  It relates file
extensions to file classes.  This indirection was
chosen because there are often several file
extensions belonging to a single class.  Its value
looks something like this:\refill
\begin{verbatim}
(
    ("1"            "Troff-man")
    ("1l"           "Troff-man")
    ("2"            "Troff-man")
...
    ("afm"          "Adobe-Font-Metric")
...
    ("web"          "Web")
    ("y"            "Yacc")
    ("yacc"         "Yacc")
)
\end{verbatim}

\noindent
Observe that the extensions do \emph{not} include
a leading period.\refill

The list of extensions was constructed by going
through some large UNIX file systems (several
hundred thousand files) to produce a set of unique
file extensions, and then augmenting the list by
hand based on the author's personal experience on
several other operating systems.  The resulting
list has about 150 file extensions, and 85 file
classes.  If a file extension is unrecognized, it
is assigned the class name \code{UNKNOWN}.\refill

Here now is how you can customize the behavior of
\code{make-file-header}.\c
\findex{make-file-header}
For each Lisp variable with
the phrase \code{-standard-}, there is a
corresponding one with the phrase \code{-extra-}
instead.  These new variables are intended for
user customization; you can initialize them in
your startup \file{.emacs} file, and they will
automatically be added to the standard ones at run
time.\refill

Here is a set of sample customizations:\refill\c
\cpsubindex{customization}{examples}
\begin{verbatim}
(setq file-header-extra-at-sign-special-cases
      '(
        ("Foo-Bar"      " <<<AT>>> ")
        ))

(setq file-header-extra-comment-prefixes
      '(
        ("Foo-Bar"      "!FB!")
        ))

(setq file-header-extra-entries
      '(
        ("copyright"    file-header-copyright)
        ))

(setq file-header-extra-suffix-and-type
      '(
        ("foobar"       "Foo-Bar")
       ))

(setq file-header-extra-paired-comment-delimiter-languages
      '(
        ("Foo-Bar"
         (concat "/#"  (make-string 70 ?\#) "\n")
         (concat (make-string 70 ?\#) "#/\n"))
        ))
\end{verbatim}
\vindex{file-header-extra-at-sign-special-cases}\c
\vindex{file-header-extra-comment-prefixes}\c
\vindex{file-header-extra-entries}\c
\vindex{file-header-extra-suffix-and-type}\c
\vindex{file-header-extra-paired-comment-delimiter-languages}\c

\noindent
These would define a new file class \code{Foo-Bar}
attached to files with extension \code{.foobar},
for which comments are delimited by \code{/# \dots{} #/},
and by \code{!} to end-of-line.  The file header body
lines would all begin with \code{!FB!}.\refill

The Lisp form \code{(setq var value)}\c
\findex{setq}
assigns \code{value} to the variable \code{var}; most
other programming languages would write this as
\code{var = value}.\refill

The extra values set in these variables are
appended to the end of the standard ones, so they
can augment, \emph{but not replace}, the standard
values.  This design choice was made intentionally
to encourage \emph{standardization} of the file
headers.  If you need to do something differently,
you'll have to learn some Lisp, and look in the
next chapter.\refill

You can test your additions by visiting
files with the new extensions, and then running
\kbd{M-x make-file-header}.\refill\c
\findex{make-file-header}

You can test the entire collection of code by
typing \kbd{M-x test-file-header}.\c
\findex{test-file-header}
This takes a while, but is thorough: it will
create file headers in a temporary editor buffer
for every file extension defined in the two lists
\code{file-header-standard-suffix-and-type}\c
\vindex{file-header-standard-suffix-and-type}
and
\code{file-header-extra-suffix-and-type}.\refill\c
\vindex{file-header-extra-suffix-and-type}

To see the settings of the variables named
\code{file-header-standard-xxx}\c
\vindex{file-header-standard-xxx}
and
\code{file-header-extra-xxx},\c
\vindex{file-header-extra-xxx}
type
\kbd{M-x show-file-header-variables}.\c
\findex{show-file-header-variables}
The results will appear in a temporary
buffer.\refill

Prior to version 19 (released in early summer of
1993), GNU Emacs did not provide the time zone,\c
\cindex{time zone}
but on UNIX systems, it can
be obtained from the output of the \code{date}\c
\pindex{date}
command.  Since this takes a few seconds to run as
a subprocess, the result is saved in a global
variable, \code{file-header-timezone-string}.\c
\vindex{file-header-timezone-string}
Subsequent file headers will be produced much more
rapidly.  With Version 19 or later, this delay is
eliminated.
\refill

If you find the delay on the first use
objectionable, you can set the time zone in your
\file{.emacs} file:\refill
\begin{verbatim}
(setq file-header-timezone-string "MST")
\end{verbatim}

\noindent
This practice is not recommended, since you'll
have to change it twice a year, or if you work in
a different time zone.\refill

\node Advanced customization, Bug reporting, Simple customization, Top

\chapter{Advanced customization}\c
\cpsubindex{customization}{advanced}

What do you do if you want to insert additional
fields in all new file headers?  You have to do
some Lisp programming to add to the
functions in \file{filehdr.el}.  \emph{Under no
circumstances should you modify \file{filehdr.el}
itself!}  That is the sole prerogative of its
original author.  You can freely copy code from
it, but put that code in a file with a different
name.\refill

If you are a real Lisp wizard, you can just read
the code in \file{filehdr.el}, and write whatever
new code you want.  On the other hand, if you were
such a wizard, you'd probably ``read the code
instead of this documentation.''\refill

The most likely function you'll want to modify is
\code{make-file-header}.\c
\findex{make-file-header}
Here is what its body
looks like:\refill
\begin{verbatim}
    (file-header-comment-block-begin)
    (file-header-entry)
    (mapcar '(lambda (entry)
               (file-header-key (car entry) (nth 1 entry)))
            (append file-header-standard-entries
                    file-header-extra-entries))
    (file-header-exit)
    (file-header-comment-block-end)
\end{verbatim}
\vindex{file-header-standard-entries}\c
\vindex{file-header-extra-entries}\c
\findex{file-header-comment-block-begin}\c
\findex{file-header-entry}\c
\findex{lambda}\c
\findex{mapcar}\c
\findex{append}\c
\findex{file-header-key}\c
\findex{car}\c
\findex{nth}\c
\findex{file-header-exit}\c
\findex{file-header-comment-block-end}\c

\noindent
Each of these lines is a Lisp function call; the
function name is the first one in each
parenthesized list.  Each function supplies part
of the standard file header.\refill

The first and last function calls provide a full
line comment start and end, if the file class
requires it.\refill

The \code{file-header-entry}\c
\findex{file-header-entry}
and \code{file-header-exit}\c
\findex{file-header-exit}
functions supply the class name tag and the final
closing brace.  That is, they generate something
like this:\refill
\begin{verbatim}
%%%  @LaTeX-file{
%%%  }
\end{verbatim}

\noindent
The individual file attributes are then supplied
by calls to the generic function
\code{file-header-key},\c
\findex{file-header-key}
which is given the
attribute name as its first argument, and the name
of a function to call to generate a string for the
attribute's initial value.  The returned string
may span multiple lines; it will be neatly
formatted and properly indented by a service
function called inside
\code{file-header-key}.\refill

The Lisp \code{mapcar}\c
\findex{mapcar}
function called in the body
of \code{make-file-header}\c
\findex{make-file-header}
applies its second argument, here an anonymous
\code{lambda}\c
\findex{lambda}
function, to each element of the list supplied as
its third argument.  The keywords that are
inserted are determined by the entries in the
lists \code{file-header-standard-entries}\c
\vindex{file-header-standard-entries}
and
\code{file-header-extra-entries},\c
\vindex{file-header-extra-entries}
which are
appended into one big list.\refill

Here is a simple example of one of these
initial value-returning functions:\refill
\begin{verbatim}
(defun file-header-codetable ()
  "Return as a string the default codetable value."
  "ISO/ASCII"
  )
\end{verbatim}

If you want to add a new file header attribute
entry, you need to add an entry to
\code{file-header-extra-entries},\c
\vindex{file-header-extra-entries}
and write a
function to return an appropriate initial
value.\refill

This is best illustrated by a real example---the
addition of a copyright attribute\c
\cpsubindex{attribute}{copyright}
in the file header.\refill

First we insert the lines\refill
\begin{verbatim}
(setq file-header-extra-entries
      '(
        ("copyright" file-header-copyright)
        ))
\end{verbatim}

\noindent
in the \file{.emacs} file.\refill

Next, we write the function to return the initial
value:\refill
\begin{verbatim}
(defun file-header-copyright ()
  "Return as a string the default copyright value."
  "None.  This file is PUBLIC DOMAIN."
)
\end{verbatim}

That is all there is to it.  To test the new code,
you can compile it inside Emacs in Emacs-Lisp
editing mode by typing \kbd{ESC C-x} with the
cursor inside the function, and then run it by
name from the minibuffer:
\kbd{ESC ESC (file-header-copyright)}.\refill

When you run \code{make-file-header},\c
\findex{make-file-header}
it should now produce an attribute entry
like\refill
\begin{verbatim}
%%%     copyright     = "None.  This file is PUBLIC DOMAIN.",
\end{verbatim}

When everything is working, save the new Emacs
Lisp file, and run \kbd{M-x byte-compile-file} on
it.  You can then load it interactively with
\kbd{M-x load-file}, or better, automatically at
Emacs start-up time by adding the line\refill
\begin{verbatim}
(load "myfilhdr" t t nil)
\end{verbatim}

\noindent
assuming you called the modified file \file{myfilhdr.el}.\refill

If the code in \file{myfilhdr.el} is short, you
can keep it in your \file{.emacs} instead, and
altogether avoid the need for a separate file and
the byte compilation and \code{load}\c
\findex{load}
command.
Compilation is only useful for speeding up the
loading of large files of Emacs Lisp code.\refill

You probably will not have to do any more than
this, unless you add a new attribute that must be
updated each time the function
\code{update-file-header-and-save}\c
\findex{update-file-header-and-save}
is invoked.  In
such a case, you'll have to study its body, and
the functions it calls, to make the necessary
modifications.\refill

\node Bug reporting, Bibliography, Advanced customization, Top

\chapter{Bug reporting}\c
\cindex{bug reporting}

Bug reports, and comments, are actively solicited.
Electronic mail to the author is most convenient,
but postal mail, preferably accompanied by
machine-readable material on Apple Macintosh or
IBM PC floppy disks, are also acceptable.  Shorter
communications via FAX are also possible.  Here
are the necessary addresses and telephone numbers:\refill

\begin{verbatim}
Nelson H. F. Beebe
Center for Scientific Computing
Department of Mathematics
University of Utah
Salt Lake City, UT 84112
USA
Tel: +1 801 581 5254
FAX: +1 801 581 4148
Email: beebe@math.utah.edu
URL: http://www.math.utah.edu/~beebe
\end{verbatim}

\node Bibliography, Concept Index, Bug Reporting, Top

\begin{ifinfo}
\begin{description}
\item[Beebe:TB11-4-485-487]
Nelson H. F. Beebe.
{{From the President}}.
\emph{TUGboat}, 11(4):485--487, November
1990.\refill

\item[Lamport:LDP85]
Leslie Lamport.
\emph{{\LaTeX}---A Document Preparation
System---User's Guide and Reference Manual}.
Ad{\-d}i{\-s}on-Wes{\-l}ey, 1985.\refill

\item[Lottor:CACM-34-11-21]
Mark Lottor.
Internet domain system.
\emph{Communications of the Association for Computing Machinery},
34(11):21--22, November 1991.\refill
This letter reports that the ZONE program at the
Network Information Systems Center at SRI
International in July 1991 found approximately
535,000 Internet hosts in 16,000 domains. The 10
largest domains were EDU (educational)--206,000,
COM (commercial)--144,000, GOV
(government)---36,000, MIL (military) 26,000, AU
(Australia)--22,000, DE (Germany)---21,000, CA
(Canada)--19,000, ORG (organizations)---15,000, SE
(Sweden)---12,000, and CH
(Switzerland)---10,000.\refill

\item[Unilogic:SDP84]
Unilogic, Ltd.
\emph{Scribe Document Production System User
Manual}, April 1984.\refill
\end{description}
\end{ifinfo}

\begin{iftex}
\bibliography{filehdr}
\end{iftex}

\onecolumn

\node Concept index, Function index, Bibliography, Top

\begin{ifinfo}
\chapter*{Concept index}
\end{ifinfo}
\begin{iftex}
\unnumbered{Concept index}
\end{iftex}
\cindex{concept index}
\cpsubindex{index}{concept}
\printindex{cp}

\onecolumn
\node Function index, Person index, Concept index, Top

\begin{ifinfo}
\chapter*{Function index}
\end{ifinfo}
\begin{iftex}
\unnumbered{Function index}
\end{iftex}
\cindex{function index}
\cpsubindex{index}{function}
\printindex{fn}

\onecolumn
\node Person index, Program index, Function index, Top

\begin{ifinfo}
\chapter*{Person index}
\end{ifinfo}
\begin{iftex}
\unnumbered{Person index}
\end{iftex}
\cindex{person index}
\cpsubindex{index}{person}
\printindex{tp}

\onecolumn
\node Program index, Variable index, Person index, Top

\begin{ifinfo}
\chapter*{Program index}
\end{ifinfo}
\begin{iftex}
\unnumbered{Program index}
\end{iftex}
\cindex{program index}
\cpsubindex{index}{program}
\printindex{pg}

\onecolumn
\node Variable index, Top, Program index, Top

\begin{ifinfo}
\chapter*{Variable index}
\end{ifinfo}
\begin{iftex}
\unnumbered{Variable index}
\end{iftex}
\cindex{variable index}
\cpsubindex{index}{variable}
\printindex{vr}

\end{document}
%%% This is for GNU Emacs file-specific customization:
%%% Local Variables:
%%% fill-column: 50
%%% End: