% \def\fileversion{ver.~1.0f}
% \def\filedate{1994/12/07}
% \def\docdate {1994/12/07}
% \CheckSum{92}
%\iffalse
% Copyright (C) 1992-1994 by Frank Bennett.  All rights reserved.
% This file contains the entire LexiBib package.  The complementary
% LaTeX2e style package LexiTeX is contained in the separate file
% lexitex.doc.
% To get started with unpacking, run this file through DOCSTRIP
% giving `driver' as the extraction option, with the extension
% `ins'.  Then run the resulting lexibib.ins file through LaTeX,
% which will strip the files.  Run lexibib.man through LaTeX
% to print the documented file.  I have tried to make the commented
% code as informative as possible; comments are welcome on the
% commentary as well as the code itself.
% IMPORTANT NOTICE:
% You are not allowed to change this file.  You may however copy
% this file to a file with a different name and then change the
% copy if (a) you do not charge for the modified code, (b) you
% acknowledge LexiBib and its author(s) in the new file, if it
% is distributed to others, and (c) you attach these same
% conditions to the new file.
% You are not allowed to distribute this file alone.  You are not
% allowed to take money for the distribution or use of this file
% (or a changed version) except for a nominal charge for copying
% etc.
% You are allowed to distribute this file under the condition that
% it is distributed with all of its contents, intact.
% For error reports, or offers to help make LexiTeX a more powerful,
% friendlier, and better package, please contact me on
% fbennett@clus1.ulcc.ac.uk
% \DoNotIndex{\@,\@@par,\@beginparpenalty,\@empty}
% \DoNotIndex{\@flushglue,\@gobble,\@input}
% \DoNotIndex{\@makefnmark,\@makeother,\@maketitle}
% \DoNotIndex{\@namedef,\@ne,\@spaces,\@tempa}
% \DoNotIndex{\@tempb,\@tempswafalse,\@tempswatrue}
% \DoNotIndex{\@thanks,\@thefnmark,\@topnum}
% \DoNotIndex{\@@,\@elt,\@forloop,\@fortmp,\@gtempa,\@totalleftmargin}
% \DoNotIndex{\",\/,\@ifundefined,\@nil,\@verbatim,\@vobeyspaces}
% \DoNotIndex{\|,\~,\,\active,\advance,\aftergroup,\begingroup,\bgroup}
% \DoNotIndex{\cal,\csname,\def,\documentstyle,\dospecials,\edef}
% \DoNotIndex{\egroup}
% \DoNotIndex{\else,\endcsname,\endgroup,\endinput,\endtrivlist}
% \DoNotIndex{\expandafter,\fi,\fnsymbol,\futurelet,\gdef,\global}
% \DoNotIndex{\hbox,\hss,\if,\if@inlabel,\if@tempswa,\if@twocolumn}
% \DoNotIndex{\ifcase}
% \DoNotIndex{\ifcat,\iffalse,\ifx,\ignorespaces,\index,\input,\item}
% \DoNotIndex{\jobname,\kern,\leavevmode,\leftskip,\let,\llap,\lower}
% \DoNotIndex{\m@ne,\next,\newpage,\nobreak,\noexpand,\nonfrenchspacing}
% \DoNotIndex{\obeylines,\or,\protect,\raggedleft,\rightskip,\rm,\sc}
% \DoNotIndex{\setbox,\setcounter,\small,\space,\string,\strut}
% \DoNotIndex{\strutbox}
% \DoNotIndex{\thefootnote,\thispagestyle,\topmargin,\trivlist,\tt}
% \DoNotIndex{\twocolumn,\typeout,\vss,\vtop,\xdef,\z@}
% \DoNotIndex{\,,\@bsphack,\@esphack,\@noligs,\@vobeyspaces,\@xverbatim}
% \DoNotIndex{\`,\catcode,\end,\escapechar,\frenchspacing,\glossary}
% \DoNotIndex{\hangindent,\hfil,\hfill,\hskip,\hspace,\ht,\it,\langle}
% \DoNotIndex{\leaders,\long,\makelabel,\marginpar,\markboth,\mathcode}
% \DoNotIndex{\mathsurround,\mbox,\newcount,\newdimen,\newskip}
% \DoNotIndex{\nopagebreak}
% \DoNotIndex{\parfillskip,\parindent,\parskip,\penalty,\raise,\rangle}
% \DoNotIndex{\section,\setlength,\TeX,\topsep,\underline,\unskip,\verb}
% \DoNotIndex{\vskip,\vspace,\widetilde,\\,\%,\@date,\@defpar}
% \DoNotIndex{\[,\{,\},\]}
% \DoNotIndex{\count@,\ifnum,\loop,\today,\uppercase,\uccode}
% \DoNotIndex{\baselineskip,\begin,\tw@}
% \DoNotIndex{\a,\b,\c,\d,\e,\f,\g,\h,\i,\j,\k,\l,\m,\n,\o,\p,\q}
% \DoNotIndex{\r,\s,\t,\u,\v,\w,\x,\y,\z,\A,\B,\C,\D,\E,\F,\G,\H}
% \DoNotIndex{\I,\J,\K,\L,\M,\N,\O,\P,\Q,\R,\S,\T,\U,\V,\W,\X,\Y,\Z}
% \DoNotIndex{\1,\2,\3,\4,\5,\6,\7,\8,\9,\0}
% \DoNotIndex{\!,\#,\$,\&,\',\(,\),\+,\.,\:,\;,\<,\=,\>,\?,\_}
% \DoNotIndex{\discretionary,\immediate,\makeatletter,\makeatother}
% \DoNotIndex{\meaning,\newenvironment,\par,\relax,\renewenvironment}
% \DoNotIndex{\repeat,\scriptsize,\selectfont,\the,\undefined}
% \DoNotIndex{\arabic,\do,\makeindex,\null,\number,\show,\write,\@ehc}
% \DoNotIndex{\@author,\@ehc,\@ifstar,\@sanitize,\@title,\everypar}
% \DoNotIndex{\if@minipage,\if@restonecol,\ifeof,\ifmmode}
% \DoNotIndex{\lccode,\newtoks,\onecolumn,\openin,\p@,\SelfDocumenting}
% \DoNotIndex{\settowidth,\@resetonecoltrue,\@resetonecolfalse,\bf}
% \DoNotIndex{\clearpage,\closein,\lowercase,\@inlabelfalse}
% \DoNotIndex{\selectfont,\mathcode,\newmathalphabet,\rmdefault}
% \DoNotIndex{\bfdefault}
% \setcounter{StandardModuleDepth}{1}
% \MakeShortVerb{\"}
% \title{\LexiBib\\a style file to link\\\BibTeX\ databases with
% the \\\LexiTeX\ style engine.\thanks{%
%    This file has version number \fileversion{} dated
%    \filedate{}.
%    The documentation was last revised on \docdate.
%    The documentation and the code for \LexiBib{} and \LexiTeX{} 
%    are \copyright{} 1992--94 Frank Bennett, Jr.  Distribution
%    and use are freely welcomed, on the sole condition that
%    acknowledgement of the \LexiTeX{} package and of its author
%    be made in any published using these utilities.
%       }}
% \author{Frank G. Bennett, Jr.}
% \maketitle
% \begin{abstract}
% \LexiBib\ is a style for use with \BibTeX\ which permits
% legal and other citations to be stored in \BibTeX\ databases,
% and incorporated in documents in a manner similar to
% the normal operation of
% \BibTeX.  Citations exported using \LexiBib\ are written in
% the syntax required for digestion by \LexiTeX, a citation
% engine that produces in-footnote, cross-referenced citations
% from nickname references to a  single citation source list.
% \LexiTeX\ is currently a \LaTeX\ 2.09 style, which will be
% upgraded to \LaTeX2e form in due course.
% \end{abstract}
% \tableofcontents
% \lexiforcefoots
% \section*{Preface to Version 1.0b}
% This {\tt doc} package contains a \BibTeX\ style named
% {\tt lexibib.bst}.  This style file can be used to
% export citations from a \BibTeX{} database in a form
% that can be used by the \LexiTeX\ citation engine.
% This permits a user to store citation details and
% annotations in a separate database, and incorporate
% those citations into a document using simple nickname
% tags.  When the document is printed, the nickname tags
% will automagically be converted into properly formatted,
% Blue Book-conformant citations.\\
% The {\tt lexibib.bst} style file contains only a modest
% amount of code from the original 1988 release of {\tt
% btxbst.doc}
% by Oren Patashnik.  While the original style is very
% efficient in its operation, its output, formatting and
% logical function are intertwined in a way that makes
% modification of the style rather difficult.  At some cost
% in speed, in {\tt lexibib.bst} I have tried to provide
% a set of modular functions that can be more easily modified
% to suit specialized formatting requirements.\\
% Please note that this is a {\sf beta} release.  Full support
% for databases prepared according the general conventions laid
% down by Oren Patashnik is not yet provided,
% so please don't rely on this style in the way you would rely
% on the standard styles (yet).  Your comments will help to
% speed the development of \LexiBib\ into a reliable clone
% of the standard styles.
% \newpage
% \pagenumbering{arabic}
% \section{User Guide}
% \subsection{Introduction}
% An adage in office management is that you should
% only touch incoming paper once; to respond to it, to file
% it, to forward it, or to destroy it.  A number of 
% commercial citation
% database managers provide a facility for ``filing'' citations
% in a flexible form, the idea being to extend this principle
% to citations as well as paper.  A hanging point for this
% strategy as been in-text context-sensitive citation styles.
% Database managers are at their best in exporting 
% entire bibliographies
% and lists of authorities.  Some packages are
% capable of scanning a document for citation ``tags'', 
% which eliminates the need to separately select bibliography
% items in the database.  Some, too, can replace ``tags''
% in the document with the formatted text of a citation.
% But once the text is replaced, the format of citations
% added in this way is fixed; conversion is a one-way process.
% A more serious problem is that, while the database manager
% can easily identify the {\em tag}, it cannot identify its 
% {\em context\/}---whether it occurs in a footnote, how many
% items were in the preceding footnote, how many articles
% by the same author are cited in the document, and so forth.
% As a result, citation formatting of cross-referenced styles
% is still done by hand.
% One of the most demanding cross-referenced styles is that
% laid down in {\sl A Uniform System of Citation}, or ``the Blue
% Book'', for the citation of legal materials.  This style
% has survived the era of computerization largely because
% most U.S. law journals using the style are edited by highly
% competitive law students.  Staff members contribute their
% editorial time to their journal free of charge, because of
% the value of listing law journal membership on their resume.
% The Blue Book style minimizes the bulk of citation
% text, while conveying sufficient information to the reader
% for the location of cited material.  It is also designed
% to provide all information required for the location of
% {\em primary\/} legal authority from the face of any citation
% to it, without tracing down cross-references.
% \LexiTeX{} and \LexiBib{} constitute the first attempt
% to allow documents requiring {\em cross-referenced},
% {\em in-text\/} citations in the demanding Blue Book style
% to draw all citation details
% from a separately maintained bibliography, with all
% citation formatting taking place automatically at print time.
% \LexiTeX{} is a style for use with the \LaTeX{} typesetting
% system.
% The implications of such a system are particularly interesting
% if it comes to be widely used, say, in the publication of
% law journals.  The efficiency with which
% citations can be reported to citation services would be
% increased, since \LexiBib{}-format database entries
% are in a standard format that can
% be processed electronically.  These same lists could
% be made available as text-searchable databases of authority.
% Network discussions of legal issues could be accompanied by
% growing lists of annotated authority, available for all to
% use.
% Even for personal use, the combination of \LexiTeX{} and
% \LexiBib{}
% Details on the use of \LexiTeX\ are contained in a separate
% manual.  The purpose of this guide is to explain the
% preparation
% and processing of \BibTeX\ databases for use with \LexiBib{}
% and
% \LexiTeX.  If you have any queries, please feel free to contact
% me on {\tt fbennett@clus1.ulcc.ac.uk}.  
% \subsection{Preparing bibliography files}
% \subsection{Processing files using \LexiBib}
% \subsection{\LexiBib\ entry types}
% \subsection{\LexiBib\ entry fields}
%\StopEventually{}
% \section{The style code}
% \subsection{To prospective style hackers}
% The \BibTeX\ style language differs from \TeX\ both in style
% and in syntax.  Whereas \TeX\ tears through a document from
% start to finish, clearing memory as it feeds out pages and
% terminating at the end of the input file, \BibTeX\ reads its
% own {\tt *.bst} file in a linear fashion, but the
% commands used in that file to not treat treat the {\em
% database\/} input file as a linear input stream.\\
% At startup, \BibTeX\ latches onto the
% {\tt *.aux} file at which it has been pointed, and scans it for
% certain strings, which are hardwired into \BibTeX's own code.
% These strings tell the program (a) which citation databases it
% must open, and (b) what citation nicknames it must search those
% databases for.  A series of initializing operations are then
% performed, which define variables and functions much as one
% defines \TeX\ macros in advance of their use.  Finally, 
% the matching items from the database can be read for use by
% \BibTeX, and the defined functions can be invoked on each item.
% The entire list may be read several times if necessary.  In the
% end, special functions which write formatted entries on an
% output file are invoked after which, having reached the end of
% the {\tt *.bst} file, the program terminates.
% The {\tt *.bst} style code below has been divided into logical
% sections which, I hope, will illustrate the logical phases of
% \BibTeX's progress.
% \subsection{About cross-referencing}
% About cross-referencing, Oren Patashnik wrote:
% \begin{quote}
% Now come the cross-referencing functions (these are invoked
% because one entry in the database file(s) cross-references
% another, by giving the other entry's database key in a 
% "crossref"
% field).  This feature allows one or more titled things that are
% part of a larger titled thing to cross-reference the larger
% thing.  These styles allow for five posibilities:
% \begin{enumerate}
% \item an {\bf article} may cross-reference an {\bf article};
% \item a cross-reference to a {\bf book} may be made in an
% entry for:
% \begin{enumerate} 
% \item a {\bf book};
% \item an {\bf inbook}; or
% \item an {\bf incollection}; and
% \end{enumerate}
% \item
% an {\bf inproceedings} may cross-reference a {\bf proceedings}.
% \end{enumerate}
%  Each of
% these is explained in more detail later.
%  An {\bf article} entry type may cross reference another
% {\bf article}
% (this
% is intended for when an entire journal is devoted to a single
% topic--- but since there is no {\bf journal} entry type, the
% journal,
% too, should be classified as an {\bf article} but without the
% author
% and title fields). This will result in two warning messages for
% the journal's entry if it's included in the reference list, but
% such is life. The other cross-referencing functions are
% similar,
% so no "comment version" will be given for them. We use just the
% last names of editors for a cross reference: either "editor",
% "editor1 and editor2", or "editor1 et~al." depending on whether
% there are one, or two, or more than two editors.
% \end{quote}
% \def\option#1{\item\def\lawtemp{#1}\def\lawtemptwo{none}
% \ifx\lawtemp\lawtemptwo
% {\em none}
% \else{\small`{\bf#1}':}\fi}
% \newenvironment{Ventry}[1]%
%  {\begin{list}{}{\renewcommand{\makelabel}[1]{\textsf{##1:}\hfil}%
%  \settowidth{\labelwidth}{\textsf{#1:}}%
%  \setlength{\leftmargin}{\labelwidth}
%  \addtolength{\leftmargin}{1\labelsep}}}%
% {\end{list}}
% \changes{1.0f}{1994/12/08}{Rewrote the `spec' macro used in the
% documentation, which was not happy with either compatibility
% mode or 2e.}
% \changes{1.0f}{1994/12/08}{\LaTeX 2e version released!}
% \def\spec#1#2#3{\begin{Ventry}{Expected}
%   \item[Expected] \begin{enumerate}
%     #1
%                   \end{enumerate}
%   \item[Left] \begin{enumerate}
%     #2
%               \end{enumerate}
%   \item[Toggles] \begin{enumerate}
%     #3
%                   \end{enumerate}
% \end{Ventry}}
% Each function defined using the {\tt FUNCTION} operator
% is accompanied by a brief description of what it is
% intended to do, followed by a box in the following form:
% \spec{\item What the function expects to find on the stack,
% numbered from the top (most recently pushed) to the bottom
% item.}
% {\item What the function will leave on the stack, numbered in
% the
% same fashion.}
% {\option{toggle name} What the effect of any toggles is on the
% behaviour of the function.}
% Happy reading.
% \subsection{Hello!}
% As ever, we start by telling the world who we are.
%    \begin{macrocode}
%<*newbst>
FUNCTION { hello }
{ "     ========================================="
"     package, version 2.0e or later"
"     Use with LateX 2.09 and the LexiTeX style"
"     -----------------------------------------"
"     documentation last updated: 5 December 1994"
"     code last updated: 6 December 1994"
"     version: 1.0f"
"     This is the LexiBib style"
"     ========================================="
stack$
%    \end{macrocode}
% \subsection{Variable initializations}
% \DescribeEnv{ENTRY}
% This command takes three braced arguments, which in order are:
% \begin{enumerate}
% \item A list of field names;
% \item A list of integer variables; and
% \item A list of string variables.
% \end{enumerate}
% The field names created here include some not found in the
% standard \BibTeX\ styles.  The use of both the standard and
% these new
% items is fully documented elsewhere, so no commentary is
% given here.  Suffice to say that each item in the list creates
% a field that
% subsequent code can work on.
%    \begin{macrocode}
ENTRY
%    \end{macrocode}
% \DescribeEnv{field names}
% \mbox{}
%    \begin{macrocode}
{ address
annote
author
booktitle
chapter
edition
editor
howpublished
institution
journal
month
number
organization
pages
publisher
school
series
title
volume
year 
jurisdiction
court
division
divno
casedate
translator
booktranslator
%    \end{macrocode}
% \DescribeEnv{entry integers}
% None are defined.  Note that any integers or strings defined
% with the
% ENTRY command are created for {\em each\/} entry in the
% citation list.
%    \begin{macrocode}
%    \end{macrocode}
% \DescribeEnv{entry strings}
% There are no general string variables, either.  Such variables
% might be defined for use in building key labels for use in
% sorting, or in an end-of-document bibliography style.
%    \begin{macrocode}
%    \end{macrocode}
% \DescribeEnv{STRINGS}
% A few string macros are needed to hold things
% temporarily removed from the stack.
% \BibTeX\ only allows ten such string variables; the named
% variables in this list will slowly be reduced in number
% through redesign of the functions that require them.
% Meanwhile, let's hope we don't find a sudden need for
% more \ldots
%    \begin{macrocode}
STRINGS { s t u theyear themonth theday }
%    \end{macrocode}
% \DescribeEnv{INTEGERS}
% The following list of integers is probably also larger than
% it needs to be.  Again, optimization of stack usage will
% lead to the elimination of some of these variables.
%    \begin{macrocode}
INTEGERS { itemcount date.specials charcount a b
           nameptr namesleft numnames a.logical }
%    \end{macrocode}
% \DescribeEnv{FUNCTIONS}
% These functions provide strings used in the parsing of
% Japanese legal citations.  If the language of such citations
% is not English, these strings may need to be edited.
%    \begin{macrocode}
FUNCTION {dc..} {"District Court"}
FUNCTION {sc..} {"Supreme Court"}
%    \end{macrocode}
% \subsection{Function definitions}
% \subsubsection{Logic and measurement}
% \DescribeEnv{not}
% \DescribeEnv{and}
% \DescribeEnv{or}
% These have been copied verbatim from Oren Patashnik's 1988
% release.
% His commentary on these functions reads: 
% {\em These three functions pop one or two (integer) arguments
% from the stack
% and push a single one, either "0" or "1".
% The {\tt 'skip\$} in the `"and"' and `"or"' functions are used
% because
% the corresponding {\tt if\$} would be idempotent.}
% \spec{\item The result of a logical test
% \item The result of a second logical test}
% {\item A single logical test result}
% {\option{none}}
%    \begin{macrocode}
 FUNCTION {not}
 {   { #0 }
     { #1 }
   if$
 FUNCTION {and}
 {   'skip$
     { pop$ #0 }
   if$
 FUNCTION {or}
 {   { pop$ #1 }
     'skip$
   if$
%    \end{macrocode}
% \DescribeEnv{times.ten}
% One of the date calculation routines requires
% string-to-integer conversion.  \BibTeX\ primitives can only
% convert in this direction for single characters.  Accordingly,
% we need to be able to raise by a power of ten.
% \spec{\item A single integer.}
% {\item A single integer.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {times.ten}
 { duplicate$ duplicate$ duplicate$ duplicate$ duplicate$
   duplicate$ duplicate$ duplicate$ duplicate$
   + + + + + + + + +
%    \end{macrocode}
% \DescribeEnv{character.length}
% \BibTeX\ provides a built-in command for measuring the length
% of a string in ``text units''.  Sometimes it's nice to know
% the length of a string in {\em characters\/};
% this function provides
% that facility.
% \spec{\item A string.}
% {\item An integer giving the number of characters in the
% string.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {character.length} 
    { 'u := 
      #0 'charcount := 
      { u empty$ not } 
        { u #2 global.max$ substring$ 'u := 
          charcount #1 + 'charcount := 
        }while$ 
      charcount 
    } 
%    \end{macrocode}
% \DescribeEnv{first.in.second}
% This function checks for a substring at the beginning or at the
% end of a given string.  This facility is needed, for example,
% to provide an automated means of toggling the formatting of
% the court division string for Japanese cases.
% In the defintion below, {\tt t} is the substring length, and
% {\tt u} is the string length.
% \spec{\item A string toggle.
% \item A string (the string in which to look).
% \item A string (the substring to look for).}
% {\item A single integer ({\tt 0} or {\tt 1}), indicating
% whether
% the substring was found at the specified location in the
% string for search.}
% {\option{end} The function looks at the end of the
% given string.
% \option{start} The function looks at the start of the
% given string.}
%    \begin{macrocode}
FUNCTION {first.in.second}
    { 's :=
      duplicate$ empty$
      { pop$ pop$ #0 }
      { swap$ duplicate$ character.length 'a :=
        swap$ duplicate$ character.length 'b :=
        b a <
        { pop$ pop$ #0 }
        { s "end" =
          { b a - #1 + global.max$ substring$ =
            { #1 }
            { #0
            }if$
          }
          { #1 a substring$ =
            { #1 }
            { #0
            }if$
          }if$
        }if$
      }if$
    }
%    \end{macrocode}
% \DescribeEnv{get.character.type}
% This function returns one of three strings, `"letter"',
% `"numeral"' or `"other"` to indicate the type of the
% character it finds on the stack.  This was written for
% use in the "gather.chars" function, but may find other
% uses as well.
%    \begin{macrocode}
FUNCTION { get.character.type }  
  { duplicate$ empty$
    { pop$ "other" }
    { duplicate$ chr.to.int$  
      duplicate$ #47 > swap$ #58 < and  
      { pop$ "numeral" }  
      { chr.to.int$  
        duplicate$ #64 > swap$ duplicate$ #91 < swap$ 
        duplicate$ #96 > swap$ #123 < and 'a.logical :=
        and a.logical or  
        { "letter" }  
        { "other"  
        }if$  
      }if$  
    }if$
  }  
%    \end{macrocode}
% \subsubsection{Housekeeping}
% These functions are used to issue warning messages and avoid
% errors during processing.
% \DescribeEnv{empty.to.null}
% This is simply the "field.or.null" function from the standard
% styles, renamed so that I could more easily remember what it
% does.  Invoked only by the output routines, it assures that
% empty field markers are replaced with the null string, to
% prevent smash-ups.
% \spec{\item The contents of a field, which may be an empty
% field marker.}
% {\item A string, possibly the null string.}
% {\option{none}}
%    \begin{macrocode}
 FUNCTION {empty.to.null}
 { duplicate$ empty$
     { pop$ "" }
     'skip$
   if$
%    \end{macrocode}
% \DescribeEnv{check}
% This check function isolates the production of warning
% messages from the output routines, to enhance transparency.
% Note that this will not convert an empty field marker to
% the null string; that is left to the output routines.
% \spec{\item A string for
% use in an error message.
% \item The contents of a field.}
% {\item The field contents.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION { check }
{ 't :=
  duplicate$ empty$
  { "empty " t * " in " * cite$ * warning$}
  'skip$
  if$
%    \end{macrocode}
% \DescribeEnv{either.or}
% \changes{1.0c}{94/08/10}{Added the {\tt either.or}
% housekeeping function, for use in {\tt j.cases}.}
% This checks whether one of a two-item pair is empty.
% If at least one is empty, the other is returned to the
% stack.  If both are non-empty, a warning is issued, and
% one item is arbitrarily selected.
% \spec{\item A string.
% \item A string.}
% {\item A single string.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {either.or}
 { duplicate$ empty$
   { pop$ duplicate$ empty$
     { pop$ "" }
     'skip$
     if$
   { swap$ duplicate$ empty$
     { pop$ }
     { "both items in an either.or pair are non-empty in " cite$ *
       warning$
       "  I'm using only ONE of these items (the second passed by the function)."
       warning$
       pop$
     }if$
   }if$
%    \end{macrocode}
% \subsubsection{Output}
% In the interest of transparency, I have tried to make the
% output functions as simple to use and as flexible in their
% operation as possible.  Every output routine expects to find
% three items on the stack, and the name of each routine
% describes how it will react to what it finds there.  
% To keep things tidy, the empty field marker
% is replaced with the null string only by the output
% routines.\\
% The names
% are sometimes coincidentally comical, but I hope informative
% as well.
% \DescribeEnv{must.must.must}
% The simplest
% output routine is the "must.must.must" routine; it
% simply catenates the three strings, replacing any
% empty field markers it finds with the null string.
% \spec{\item A string.
% \item A string.
% \item A string.}
% {\item none}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {must.must.must}
    { empty.to.null 't :=
      empty.to.null swap$ empty.to.null swap$ t
      * * 
      write$
    }
%    \end{macrocode}
% \DescribeEnv{might.ifone.must}
% This output routine is heavily used for conditional
% output. If the deepmost of the three stack items is an empty
% field marker, the only the topmost item is written to the
% output file.  If this is not the case, then the three items are
% written in order.
% \spec{\item A string.
% \item A string.
% \item A string.}
% {\item {\em none}}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {might.ifone.must}
    { empty.to.null 't :=
      swap$ duplicate$ empty$
      { pop$ pop$ t }
      { swap$ empty.to.null t * * }
      if$
      write$
    }
%    \end{macrocode}
% \DescribeEnv{iftwo.might.iftwo}
% This function is typically used to output a formatted item
% in enclosing braces, where the entire item and its braces
% should be suppressed if the item is empty.
% \spec{\item A string.
% \item A string.
% \item A string.}
% {\item none}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {iftwo.might.iftwo}
    { 't :=
      duplicate$ empty$
       { pop$ pop$ }
       { empty.to.null swap$ empty.to.null swap$
         t empty.to.null
         * * write$ }
      if$
    }
%    \end{macrocode}
% \DescribeEnv{ifthree.ifthree.might}
% This outputs a pair of prefixes if the suffix to which they are
% to be attached exists.  Otherwise it outputs a null string.
% \spec{\item A string.
% \item A string.
% \item A string.}
% {\item none}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {must.ifthree.might}
    { 't :=
      duplicate$ empty$
      { pop$ write$ }
      { swap$ empty.to.null
        swap$ empty.to.null
        t * * write$ }
      if$
    }
%    \end{macrocode}
% \subsubsection{Parsing and conversion}
% \DescribeEnv{field.tag.no.combine}
% The "no" here stands for ``number''.
% This function is typically used to simplify the task of
% building a numbered label for something (e.g.\ a technical
% report).  It provides facilities for handling blank
% entries, so that these error-handling routines do not
% need to be drafted from scratch for every such
% situation.
% \spec{\item A string toggle.
% \item An identifier (i.e.\ a report number).
% \item A text prefix for the identifier.
% \item A string label (i.e.\ ``Technical Report'' or somesuch).}
% {\item A single string.}
% {\option{endlabel} Combines the text items in the same order
% which they are found.
% \option{frontlabel} Places the label in front, followed by the
% identifier, followed by the first item pushed.}
%    \begin{macrocode}
FUNCTION {field.tag.no.combine}
  { "endlabel" =
    { duplicate$ empty$
      { pop$ pop$ empty.to.null }
      { empty.to.null 's := empty.to.null
      swap$ empty.to.null swap$ s * *
      }if$
    }
    { duplicate$ empty$
      { pop$ pop$ empty.to.null }
      { empty.to.null 's := empty.to.null
        swap$ empty.to.null s swap$ * * 
      }if$
    }if$
%    \end{macrocode}
% \DescribeEnv{change.letter.case}
% This is a simple front-end to the {\tt change.case\$} built-in
% function, used to change characters in a string from upper to
% lower case.  It slightly enhances the built-in function by
% providing an option that leaves the string alone.
% \spec{\item A single character---either 
% an {\tt n}, a {\tt t}, an {\tt l}, or a {\tt u}.
% \item A string for conversion.}
%  {\item One string, converted according to the option toggle
% pushed
% after the string.}
% {\option{n} yields the string unchanged.
% \option{t} yields
% all lower case letters except for the first.
% \option{l} yields all lower case letters.
% \option{u} yields all
% upper case letters.}
%    \begin{macrocode}
FUNCTION {change.letter.case}
    { 't :=
      duplicate$ empty$
        'skip$
        { t chr.to.int$ "n" chr.to.int$ =
          'skip$
          { t change.case$ }
          if$
        }
      if$
    }
%    \end{macrocode}
% \DescribeEnv{n.dashify}
% The "n.dashify" function makes each single "-" in a string a
% double "--".
% if it's not already.\footnote{This comment by Oren Patashnik.}
% \spec{\item A single string for conversion.}
% {\item A single converted string.}
% {\option{none}}
%    \begin{macrocode}
 FUNCTION {n.dashify}
 { 't :=
   ""
     { t empty$ not }
     { t #1 #1 substring$ "-" =
      { t #1 #2 substring$ "--" = not
          { "--" *
            t #2 global.max$ substring$ 't :=
          }
          {   { t #1 #1 substring$ "-" = }
           { "-" *
             t #2 global.max$ substring$ 't :=
           }
            while$
          }
        if$
      }
      { t #1 #1 substring$ *
        t #2 global.max$ substring$ 't :=
      }
       if$
     }
   while$
%    \end{macrocode}
% \DescribeEnv{gather.chars}
% The following function was designed for date parsing, but
% may find other uses as well.  It finds the first letter or
% numeral in a given string, then proceeds to build a substring
% until it hits a different character type or the end of the
% string,
%  at which point it
% stops parsing and terminates.
% \spec{\item A string for parsing.}
% {\item A string toggle showing the character type of the
% first homogenous substring in the string for parsing.
% \item A string of characters from the front of the given string
% which are letters or numbers only.
% \item The remainder of the string.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION { gather.chars }  
  { 't :=
    { t empty$ not
      t #1 #1 substring$ get.character.type
      "other" = and }
    { t #2 global.max$ substring$ 't :=
    }while$
    t #1 #1 substring$ duplicate$ get.character.type
    "" swap$ duplicate$ 'u :=
    t #2 global.max$ substring$ 't :=
    { u = }
    { * t #1 #1 substring$ duplicate$ get.character.type
      duplicate$ u =
      { t #2 global.max$ substring$ 't :=  }  
      { swap$ pop$
      }if$  
    }while$
    t swap$ u
  }  
%    \end{macrocode}
% \DescribeEnv{tie.or.space.connect}
% We use this function from the standard styles.  It 
% adds a tie if the string it is applied to is three
% characters or less in length.
% \spec{\item A string
% \item A second string}
% {\item One string}
% {\item {none}}
%    \begin{macrocode}
 FUNCTION {tie.or.space.connect}
 { duplicate$ text.length$ #3 <
     { "~" }
     { " " }
   if$
   swap$ * *
%    \end{macrocode}
% \DescribeEnv{format.pages}
% This function accepts one toggle option currently,
% but it is open to expansion.  If the "short" toggle is
% fed to this function, it reads through the field until
% it hits a non-integer character.  Otherwise, it "n.dashify"s
% the
% whole field.  This needs robustification; the page number
% might be a roman numeral.  Might it be simplest to just 
% drop the scan when we hit a "-"?  Deserves some thought.
% \spec{\item A toggle string.
% \item page number or possibly
% a range of pages.}
% {\item One string, constituting a finished page number, or {\tt
% empty\$}.}
% {\option{short} will return only the
% first number given in the string.
% \option{full}
% will return the page number, n-dashified.}
% \changes{1.0e}{94/12/05}{Altered {\tt format.pages} so
% that the {\tt short} scan stops at a {\tt-} character,
% rather than a non-integer.  Simpler and more robust.}
%    \begin{macrocode}
FUNCTION {format.pages}
    { swap$ duplicate$ empty$
      { pop$ pop$ "" }
      { swap$ duplicate$ "short" =
        { pop$
          's :=
          ""
          s #1 #1 substring$
          { "-" = not }
            { s #1 #1 substring$ *
              s #2 global.max$ substring$ 's :=
              s #1 #1 substring$
              duplicate$ "" =
                { pop$ "-" }
                'skip$
                if$
            }
          while$
        }
      { "full" =
        { pages n.dashify }
        { "invalid switch fed to the format.pages function"
           warning$
        }if$
      }if$
    }if$
%    \end{macrocode}
% \paragraph{Names}
% The code for parsing and formatting names is extremely
% economical,
% thanks to the powerful built-in functions that \BibTeX\
% supplies
% for this purpose.
% \DescribeEnv{format.names}
% This is based on Oren Patashnik's original function of the same
% name.  His comment was as follows:
% \begin{quote}
% The format.names function formats the argument (which should
% be in
% \BibTeX name format) into ``{\tt First Von Last, Junior}'',
% separated by commas
% and with an "and" before the last (but ending with "et~al." if
% the last
% of multiple authors is "others").  This function's argument
% should always
% contain at least one name.
% The format.authors function returns the result of
% format.names(author)
% if the author is present, or else it returns the null string.
% \end{quote}
% This function is used to format any name field that is
% thrown at it.  It is based on the 1988 release, but with
% modifications
% to permit toggling, since the Blue Book requires different
% author
% formats for different types of material.  The toggling strategy
% should also make modification of this style a pretty simple
% matter.
% \spec{\item The contents of one name field
% \item  string toggle
% (either {\tt firstinitial}, {\tt lastonly} or {\tt full})}
% {\item A single string, containing a formatted name or names,
% or {\tt empty\$}}
% {\option{firstinitial} Yields the form ``F.~Bennett, Jr.''
% \option{lastonly}
% Yields the form ``Bennett''
% \option{full} yields ``Frank Bennett, Jr.''}
%    \begin{macrocode}
FUNCTION {format.names}
    { 'u :=
      duplicate$ empty$
        'skip$
        { 's :=
          #1 'nameptr :=
          s num.names$ 'numnames :=
          numnames 'namesleft :=
          { namesleft #0 > }
            { u "lastonly" =
              { s nameptr "{vv~}{ll}" format.name$ 't := }
              { u "firstinitial"  =
                { s nameptr "{f.~}{vv~}{ll}{, jj}" format.name$
't := }
                { u "full" =
                  { s nameptr "{ff~}{vv~}{ll}{, jj}" format.name$ 't :=}
                  { "style error; invalid or non-existent toggle" warning$ }
                  if$
                }
                if$
              }
              if$
              nameptr #1 >
                { namesleft #1 >
                  { ", " * t * }
                  { numnames #2 >
                    { "," * }
                    'skip$
                    if$
                    t "others" =
                      { " et~al." * }
                      { " and " * t * }
                    if$
                  }
                if$
                }
                't
              if$
              nameptr #1 + 'nameptr :=
              namesleft #1 - 'namesleft :=
            }
          while$
        }
      if$
    }
%    \end{macrocode}
% \paragraph{Dates}
% The style code for parsing and formatting dates is much more
% complex than that for names.  This is due to the need to
% build the necessary tools out of \BibTeX\ primitives, since
% no built-in tools for this task are supplied.  There are
% two date formatting routines below.  The simpler of the two
% is "format.month.year".  This is basically just the function
% supplied with the standard \BibTeX\ styles, under a new name.
% The \LexiBib\ version of "format.date" is underpinned by
% a whole set of new
% functions, and allows great flexibility in the syntax
% for entering dates.
% \subparagraph{Old format routine}
% \DescribeEnv{format.month.year}
% The "format.date" function is for the month and year, but we
% give a warning if
% there's an empty year but the month is there, and we return the
% empty string
% if they're both empty.\footnote{This comment by Oren
% Patashnik.}
% This is not changed over the "format.date" function in the 
% original standard \BibTeX{} styles.  It is retained, although
% the distributed \LexiBib\ styles won't be making use of it.
% \spec{\item none}
% {\item A single string.}
% {\option{none}}
%    \begin{macrocode}
 FUNCTION {format.month.year}
 { year empty$
     { month empty$
      { "" }
      { "there's a month but no year in " cite$ * warning$
        month
      }
       if$
     }
     { month empty$
      'year
      { month " " * year * }
       if$
     }
   if$
%    \end{macrocode}
% \subparagraph{New date parsing routines}
% The "format.date" function depends upon a number of
% supporting functions.  Some of these are of general utility,
% and are presented above.  Those presented here are specific
% to this particular function.
% \DescribeEnv{fillout.a.year}
% This adds a leading "19" to a year entered in two-digit
% form.
% \spec{\item A single string.}
% {\item A single string.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {fillout.a.year}
 { duplicate$ character.length #2 =
   { pop$ "19" swap$ * }
   'skip$
   if$
%    \end{macrocode}
% \DescribeEnv{parse.month}
% This function is a simple parser, used in converting
% database entries that have been identified as probable
% abbreviated month entries into numerical string form.
% \spec{\item A single string, which should consist of
% exactly three alphabetic characters.}
% {\item A single string of numbers.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {parse.month}
 { duplicate$ "jan" =
   { pop$ "1" }
   { duplicate$ "feb" =
     { pop$ "2" }
     { duplicate$ "mar" =
       { pop$ "3" }
       { duplicate$ "apr" =
         { pop$ "4" }
         { duplicate$ "may" =
           { pop$ "5" }
           { duplicate$ "jun" =
             { pop$ "6" }
             { duplicate$ "jul" =
               { pop$ "7" }
               { duplicate$ "aug" =
                 { pop$ "8" }
                 { duplicate$ "sep" =
                   { pop$ "9" }
                   { duplicate$ "oct" =
                     { pop$ "10" }
                     { duplicate$ "nov" =
                       { pop$ "11" }
                       { duplicate$ "dec" =
                         { pop$ "12" }
                         { "invalid month in " cite$ * warning$
                         }if$
                       }if$
                     }if$
                   }if$
                 }if$
               }if$
             }if$
           }if$
         }if$
       }if$
     }if$
   }if$
%    \end{macrocode}
% \DescribeEnv{extract.date}
% This function takes a labelled stack of numeric items,
% and places its contents into the appropriate variables
% representing elements of a date.  It will probably
% tolerate further enhancement, and should probably 
% deliver its result to the stack, rather than into
% variables.
% The behaviour of this function is a little too complex
% to describe its input and output behaviour using our
% usual description box, so we revert to prose here.
% The stack is always left empty, and there are no
% toggles that affect the overall behaviour of the
% function.  The expectation of what will appear on the
% stack depends on the contents of the integer
% variable "itemcount".  For each date item counted, the
% function expects to find two items on the stack: a
% string possibly flagging the variable into which the
% item should be placed; and a numerical string.
% If "itemcount" is "0", a warning of an empty date
% is issued.
% If "itemcount" is "1", the flag string is ignored,
% and the numerical item is assumed to be a year.
% If "itemcount" is "2", the flag strings are both
% ignored, and the numerical items are assumed to be
% a year and a month, in that order.
% If "itemcount" is "3", the first item is assumed to
% be a year.  The assignment of the second two items
% depends first upon the contents of the integer
% variable "date.specials".  If this is "0", then
% the flags are ignored, and the numerical items are
% assumed to be a day and a month, in that order.
% If "date.specials" is "1" (`true'), then the
% flag of the first item is checked to see whether it
% is ``"month"''.  If so, the first item is assigned to
% the "themonth" variable, and the next to the "theday"
% variable.  Otherwise the assignments are reversed.
%    \begin{macrocode}
FUNCTION {extract.date}
  { "" 'theyear := "" 'themonth := "" 'theday :=
    itemcount #0 =
    { "some date or other is COMPLETELY empty in " cite$ * warning$ }
    { itemcount #1 =
      { pop$ fillout.a.year 'theyear := "" 'theday := "" 'themonth := }
      { itemcount #2 =
        { pop$ fillout.a.year 'theyear :=
          pop$ 'themonth :=
          "" 'theday :=
        }
        { itemcount #3 =
          { date.specials
            { pop$ fillout.a.year 'theyear := 
              "month" =
              { 'themonth :=
                pop$ 'theday := }
              { 'theday :=
                pop$ 'themonth :=
              }if$
            }
            { pop$ fillout.a.year 'theyear :=
              pop$ 'theday :=
              pop$ 'themonth :=
            }if$
          }
          { "too many items for date in " cite$ * warning$
          }if$
        }if$
      }if$
    }if$
%    \end{macrocode}
% \DescribeEnv{format.month.name}
% \changes{\LexiBib{} 1.0b}{94/08/09}{Added {\tt swap\$} as
appropriate
% to expose string on stack for {\tt empty\$} check.}
% This takes a numerical string and converts it to either an
% abbrviated or a spelled-out month name.
% \spec{\item A toggle string.
% \item A numerical string.}
% {\item A string.}
% {\option{long} The month name placed on the stack will be
% spelled out.
% \option{short} The month name will be abbreviated.}
%    \begin{macrocode}
FUNCTION { format.month.name }
 { swap$ duplicate$ empty$
   { pop$ pop$ "" }
   { swap$ "long" = 
     { duplicate$ "1" =
       { "January" }
       { duplicate$  "2" =
         { "February" } 
         { duplicate$  "3" =
           { "March" } 
           { duplicate$ "4" =
             { "April" } 
             { duplicate$  "5" =
               { "May" } 
               { duplicate$ "6" =
                 { "June" } 
                 { duplicate$  "7" =
                   { "July" } 
                   { duplicate$  "8" =
                     { "August" } 
                     { duplicate$  "9" =
                       { "September" } 
                       { duplicate$  "10" =
                         { "October" } 
                         { duplicate$  "11" =
                           { "November" } 
                           { duplicate$  "12" =
                             { "December" }
                             { "invalid month in " cite$ * warning$ "" 
                             }if$
                           }if$
                         }if$
                       }if$
                     }if$
                   }if$
                 }if$
               }if$
             }if$
           }if$
         }if$
       }if$
     }
     { duplicate$ "1" =
       { "Jan." }
       { duplicate$  "2" =
         { "Feb." } 
         { duplicate$  "3" =
           { "Mar." } 
           { duplicate$ "4" =
             { "Apr." } 
             { duplicate$  "5" =
               { "May" } 
               { duplicate$ "6" =
                 { "Jun." } 
                 { duplicate$  "7" =
                   { "Jul." } 
                   { duplicate$  "8" =
                     { "Aug." } 
                     { duplicate$  "9" =
                       { "Sept." } 
                       { duplicate$  "10" =
                         { "Oct." } 
                         { duplicate$  "11" =
                           { "Nov." } 
                           { duplicate$  "12" =
                             { "Dec." }
                             { ""
                             }if$
                           }if$
                         }if$
                       }if$
                     }if$
                   }if$
                 }if$
               }if$
             }if$
           }if$
         }if$
       }if$
     }if$
swap$ pop$
   }if$
%    \end{macrocode}
% \DescribeEnv{topup.date}
% This is used to perform an addition operation on a
% string of numerals.
% \spec{\item An integer to be added.
% \item A string consisting of numerals only.}
% {\item A string consisting of the sum of the integer and
% the numeric string.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {topup.date}
 { 'a :=
   duplicate$ character.length #2 =
   { duplicate$ #1 #1 substring$ chr.to.int$ #48 -
     times.ten swap$ #2 #1 substring$ chr.to.int$ #48 -
     +
   { duplicate$ character.length #1 =
     { chr.to.int$ #48 - }
     { pop$ #0 "I can't cope with more than two Japanese year digits in "
       cite$ * warning$
     }if$
   }if$
 a + int.to.str$
%    \end{macrocode}
% \DescribeEnv{format.jdate}
% This rather specialized function converts a Japanese
% Imperial date written
% in a fixed syntax into the \LexiBib\ internal date stack,
% for onward handling by "extract.date".
% \spec{\item A string in the form: {\tt s57.9.27}, where
% the first letter indicates the Imperial reign of the year
% given, the first numeric item is the year, the second the
% month, and the third the day.  Any non-alphabetic, non-numeric
% character may be used as a separator.}
% {\item {\tt itemcount} is set to {\tt 2} (one more is added
% by the {\tt format.date} function in which is function is
% nested).
% \item Six stack items are output---see {\tt extract.date}
% for details on stack syntax.}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {format.jdate}
 { duplicate$ #2 global.max$ substring$ gather.chars
   pop$ 't := swap$
   #1 #1 substring$ duplicate$ "s" =
   { pop$ t #1925 topup.date }
   { duplicate$ "m" =
     { pop$ t #1867 topup.date }
     { duplicate$ "t" =
       { pop$ t #1911 topup.date }
       { duplicate$ "h" =
         { t #1988 topup.date }
         { "invalid Imperial calendar code in " cite$ * warning$
         }if$
       }if$
     }if$
   }if$
   swap$
   gather.chars
   pop$ swap$
   gather.chars
   pop$ swap$ pop$ 't := swap$ "default" swap$ t swap$ "default"
   swap$ "default"
   #2 'itemcount :=
%    \end{macrocode}
% \DescribeEnv{format.date}
% This function makes use of "gather.chars" and other supporting
% functions to build a stack of date items, which it then parses
% using "extract.date".  The syntax for entering dates is
% described in the user guide.
% \spec{\item A single string in appropriate date syntax.}
% {\item none}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {format.date}
  { empty.to.null 's :=
    #0 'charcount :=
    #0 'itemcount :=
    #0 'date.specials :=
    { s empty$ not }
      { s gather.chars
        duplicate$ "letter" =
        { pop$ duplicate$ character.length #1 = itemcount not and
          { pop$ pop$ s format.jdate "" 's := }
          { duplicate$ character.length #3 =
            { swap$ 's :=
              parse.month "month" #1 'date.specials := }
            { swap$ 's :=
              pop$ "1" "invalid date in " cite$ * warning$
              "default"
            }if$
          }if$
        }
        { "numeral" =
          { duplicate$ character.length #1 =
            { swap$ 's := "default" }
            { duplicate$ character.length #2 =
              { swap$ 's := "default" }
              { duplicate$ character.length #4 =
                { swap$ 's := "default" }
                { swap$ 's := pop$ "1"
                  "invalid numerical element in date for " cite$ * warning$
                  "default"
                }if$
              }if$
            }if$
          }
          { swap$ 's := pop$ "1"
            "failed to parse date in " cite$ * warning$
            "default"
          }if$
        }if$
        itemcount #1 + 'itemcount :=
      }while$
    extract.date
%    \end{macrocode}
% \subsubsection{Entry type specific functions}
% The functions below should not be used as general utilities;
% they are designed specifically for use with a particular entry
% type.  While they could be placed directly into the entry
% functions to which they apply, defining them separately
% helps improve the transparency of the code.
% \DescribeEnv{j.format.division}
% This function assembles the various fields relevant to the
% division of a Japanese court into a single string.
% \spec{\item none}
% {\item A single string}
% {\option{none}}
%    \begin{macrocode}
FUNCTION {j.format.division}
 { division empty$
   { "" }
   { dc.. court "end" first.in.second
     { " no.~" 's := "endlabel" 't := }
     { sc.. court "start" first.in.second
       { "No.\ " 's := "frontlabel" 't := }
       {  " " 's :=  "endlabel" 't :=
       }if$
     }if$
   }if$
   division s divno t field.tag.no.combine
%    \end{macrocode}
% \DescribeEnv{format.thesis.type}
% The function "format.thesis.type" returns either the
% (case-changed) type field,
% if it is defined, or else the default string already on the
% stack
% (like ``Master's thesis'' or ``PhD thesis'').\footnote{This
% comment by Oren Patashnik.}
% \spec{\item One string, representing the default label for a
% thesis}
% {\item One string, to stand as the label for a thesis}
% {\option{none}}
%    \begin{macrocode}
%<*oldbdoc>
FUNCTION {format.thesis.type}
{ type empty$
'skip$
{ pop$
type "t" change.case$
%</oldbdoc>
%    \end{macrocode}
% \DescribeEnv{format.tr.number}
% The function "format.tr.number" makes a string starting with
% ``Technical Report''
% (or type, if that field is defined), followed by the number if
% there is one;
% it returns the starting part (with a case change) even if there
% is no number.
% This is used at the beginning of a sentence.\footnote{This
% comment by Oren Patashnik.}
% \spec{\item none}
% {\item One string, the label for a technical report}
% {\option{none}}
%    \begin{macrocode}
%<*oldbdoc>
FUNCTION {format.tr.number}
{ type empty$
{ "Technical Report" }
'type
number empty$
{ "t" change.case$ }
{ number tie.or.space.connect }
%</oldbdoc>
%    \end{macrocode}
% \subsubsection{Entry type functions}
% Now we define the type functions for all entry types that may
% appear
% in the {\tt*.bib} file---e.g., functions like `"article"' and 
% `"book"'.  These
% are the routines that actually generate the "*.bbl" file output
% for
% the entry.  These must all precede the "READ" command.  In
% addition, the
% style designer should have a function `"default.type"' for
% unknown types.\footnote{This comment by Oren Patashnik.}
% \DescribeEnv{article}
% This function performs the necessary operations for
% exporting a valid \LexiTeX\ citation to an article.
% For this and for all citation types defined in the
% \LexiBib{} style, the goal is to provide reasonably
% complete commentary, so that anyone wanting to
% alter the behaviour of the style can set to work
% with a fair degree of confidence about what needs to
% be done to achieve a particular result.
%    \begin{macrocode}
FUNCTION { article }
%    \end{macrocode}
% We write the citation leader, to prepare for outputting the
% actual content of the citation text.
%    \begin{macrocode}
{ "\lexibib{article}{" cite$ "}{" must.must.must 
%    \end{macrocode}
% The author name is pushed to the stack, followed by a toggle
% to trigger last-name-only formatting.  Then the "format.names"
% reduces this to a single, appropriately-formatted string,
% possibly the null string.  All that is left to do is push a
% set a braces, a null string to make up three arguments to the
% export routine, and write the lot on the output
% file unconditionally, using "must.must.must".
%    \begin{macrocode}
author  "lastonly" format.names
must.must.must
%    \end{macrocode}
% Next comes the title of the article.  This is not specially
% formatted; we simply push the title field, then a warning
% string
% followed by a check for whether it is empty or not, then
% braces and a null string, and write again.
%    \begin{macrocode}
title "title" check
must.must.must
%    \end{macrocode}
% The next bit is actually rather thorny.  Our first
% formatting decision is based on whether or not there
% is a "volume" for this source.  If not, we adopt a
% Commonwealth-style citation scheme, using the bracketed
% year for the volume.  This will be coupled with suppression
% of the year in the final LexiTeX ``field''.
%    \begin{macrocode}
volume empty$
  { "[" year format.date theyear "] " must.must.must
    number  "\ " journal "journal" check might.ifone.must
%    \end{macrocode}
% If there {{\em is\/} a volume, we make different formatting
% decisions, depending on
% whether an issue number exists for this source.
%    \begin{macrocode}
  { number empty$
%    \end{macrocode}
% If there is no number, we do as follows.  We
% push the volume, providing
% a warning if it it is empty, then
% a separating space, then the name of the journal, writing
% all three on the output, or just the journal name if the
% volume number is empty.  This is a nice illustration of
% the utility of the "might.ifone.must" export routine.
% The braces are pushed and written separately.
%    \begin{macrocode}
    { volume "volume" check
      "\ "
      journal  "journal" check
      might.ifone.must }
%    \end{macrocode}
% If the "number" field was non-empty, then we do as follows.
% The journal name goes up first, and is checked.  Then the
% bridge for the volume number, and the volume number itself
% go up, both being checked.  The "must.ifthree.might" export
% routine will drop the bridge and number, if the volume number
% was empty, although this will probably leave us with an
incomplete
% (but nice-looking) citation anyway.  The we push a bridge to
% the number, the issue number, and a null string,
% and force all three on the
% output (no need for conditions in this last action, since we
% know already that "number" is non-empty.
%    \begin{macrocode}
    { journal "journal" check
      ", vol.~"
      volume "volume" check
      must.ifthree.might
      ", no.~" number "" must.must.must
    }if$
  }if$
%    \end{macrocode}
% Finally, we complete the second \LexiTeX{} argument
% and start the next.  The emply strings just take up space
% of course.
%    \begin{macrocode}
  "}{" "" "" must.must.must
%    \end{macrocode}
% The Blue Book does not like page ranges
% so we need to clean
% out anything following a dash in the "pages" field.  The
% "short" option
% toggles this behaviour on.  We also check to see that the page
% is not empty.  This is followed by braces, a null string, and
% output.
%    \begin{macrocode}
pages "short" format.pages "pages" check
must.must.must
%    \end{macrocode}
% We add the year next, but only if the "volume" field is
% non-empty (if "volume" is empty, we'll have put the year
% is as a bracketed volume number, Commonwealth-style.
% We could use just the year (this is normal for Blue Book
% style, but we'll add the month for good measure, if it's
% been provided in the "year" field.
%    \begin{macrocode}
volume empty$
  'skip$
  { year format.date themonth "short" format.month.name
  "\ " theyear might.ifone.must
  }if$
%    \end{macrocode}
% Finally, we have to tangle with cross-references.
% Yuck.  (Not a complaint about \BibTeX, just a general
% response to the design problems inherent in the task).
% Formatting depends on whether there is a "crossref" entry.
% Hear that, guys?  If you don't use the "crossref" field,
% we'll short-change you on formatting service.
%    \begin{macrocode}
crossref empty$
%    \end{macrocode}
% If there is no "crossref", we just push a brace and a couple
% of nulls, and write.  Done!  Hurray!
%    \begin{macrocode}
{ "}" "" "" must.must.must }
%    \end{macrocode}
% But if there {\em was\/} a "crossref", we've go work to do.
% Darn.  The first thing we do is have a look at "booktitle".
% This should be non-null in this situation; there's no sense
% setting up a cross-reference to an individual volume of
% a journal unless there's something special to be said
% about it.
%    \begin{macrocode}
{ booktitle empty$
%    \end{macrocode}
% So if no "booktitle" is found, we whinge and format
% as for a no-cross-reference entry.
%    \begin{macrocode}
  { "no booktitle (name of special issue) for "
     cite$ "/" crossref * * * warning$
    "}" "" "" must.must.must }
%    \end{macrocode}
% If there {\em is\/} a "booktitle", though, 
% and if the "volume" field is non-empty (which
% means that we just printed, or at least should
% have printed, the year), we close the
% parens following the year (which is opened by \LexiTeX),
% and open another (which will be closed by \LexiTeX).
% That's it for conditional punctuation; we follow
% with "booktitle", which should be 
% the title or subject description of the special
% issue of the journal.
%    \begin{macrocode}
  { volume empty$ {""} {") ("} if$ booktitle "}" must.must.must}
  if$
%    \end{macrocode}
% Add a fresh new line in the export file, and we're done!  Whew!
%    \begin{macrocode}
newline$
%    \end{macrocode}
% \DescribeEnv{book}
% This is the entry for books, which includes individual
% volumes in a series, and multi-volume works with a single
% title.  Correct me if I'm wrong, but I think this latter
% citation type is not supported by \BibTeX.  \LexiBib{}
% manages it by allowing the user to specify the volume
% number in the text using the optional ":<number>:" argument to
% the "\lexicite" tag.
%    \begin{macrocode}
FUNCTION {book}
%    \end{macrocode}
% After the opener, we push the opening macro tag for a book
% entry, the nickname of the citation, and a couple of
% braces.  This is all mandatory and can safely be given
% unconditional export.
%    \begin{macrocode}
{ "\lexibib{book}{" cite$ "}{" must.must.must
%    \end{macrocode}
% A non-empty "volume" field means we need a leading
% volume number in the Blue Book style.  If a volume
% number (or anything else) is found in the "volume"
% field of a book entry, we replace it with a 
% "volno" macro.  This will expand in the document to
% whatever the author has specified using the optional
% ":<number>" argument to "\lexicite".  For example,
% volume 8 of "holdsworth" would be: "\lexicite:8:{holdsworth}".
%    \begin{macrocode}
  volume empty$
  { "" }
  { "\volno " }
  if$
%    \end{macrocode}
% The author comes first.  We push the contents of the
% "author" field, then the toggle string ``"firstinitial"'',
% and run  the "format.names" function to produce the name
% formatted properly for a book entry.  Then we push
% fa couple of braces
% and force all three items ("\volno" or null, author, and
% braces) onto the output.
%    \begin{macrocode}
  author
  "firstinitial" format.names "}{" must.must.must
%    \end{macrocode}
% The title is very straightforward.  We push the title, then
% check to be sure it's non-empty, then a couple of braces
% go up, filled out with a null string, and force all three
% onto the output.
%    \begin{macrocode}
  title "title" check "}{" "" must.must.must 
%    \end{macrocode}
% We're now in the final ``field'' of the \LexiTeX{} entry.
% This is mainly for the year, but we also give the 
% name of the editor(s) or translator(s) 
% if present.  If there is no
% editor or translator, we'll put a series name here, to help
identify
% the source.  We don't put both, because this would
% confuse things (there can be book editors and
% series editors too, and the Blue Book style is 
% too streamlined to distinguish the two elegantly).
% So our first task is to see if there is an editor\ldots
%    \begin{macrocode}
    editor empty$ translator empty$ and
%    \end{macrocode}
% \ldots{} and if there is none, we put in a
% series name if it exists.  
% The "series" \BibTeX{} field should be used for the
% name of the series of which a volume forms a part,
% but some folks might accidentally use "booktitle".
% We'll be forgiving and accept it anyway.
%    \begin{macrocode}
    { series empty$ booktitle empty$ and
%    \end{macrocode}
%  If both
% "series" and "booktitle" are empty, we've nothing to do.
%    \begin{macrocode}
      'skip$
%    \end{macrocode}
% If at least one is non-empty, we push both,
% and cull them to just one using the "either.or" function.
% If only one is non-empty, this function leaves that one;
% if both are non-empty, the "either.or" function whinges
% and chooses one arbitrarily.
% Next, we push a bridge and a series item number,
% and a toggle for the "field.tag.no.combine" function.
% The "endlabel" toggle causes this function to put
% the bridge and the number after the series name,
% if a number exists, and push the lot back as a single
% item on the stack.  Otherwise it leaves just the series
% name.
% And last, we put up a comma to close, and do a mandatory
% export of the lot.
%    \begin{macrocode}
      { series booktitle either.or " No.~" number
      "endlabel" field.tag.no.combine ", " "" must.must.must }
      if$
    }
%    \end{macrocode}
% If either the editor or the translator fields were not empty,
% we format the editor or translator name
% instead, and put those details here.
%    \begin{macrocode}
    { editor translator either.or "firstinitial" format.names
%    \end{macrocode}
% We need to append the correct designator, either ``ed.''\ or
% ``trans.''.  The "either.or" function will use the second
% item pushed if both are non-empty, so we take advantage of
% this ``feature'' in making our choice of designators; the
% "ed." or "eds." strings are only used if the "translator"
% field is empty.  And finally, we push a null string to
% round out, and do a compulsory export.
%    \begin{macrocode}
      translator empty$
      { editor num.names$ #1 >
        { " eds.\ " }
        { " ed.\ " }
        if$
      }
      { " trans.\ " }
      if$
      "" must.must.must
    }if$
%    \end{macrocode}
% We also need to indicate the edition, if any.
%    \begin{macrocode}
  edition " ed.\ " "" might.ifone.must
%    \end{macrocode}
% The year itself is easy.  We push the year, do a check to issue
% a warning if necessary, then run format.date over it,
% which yields the year in "theyear", which can be pushed
% back onto the stack.  Then we fill out to six
% \LexiTeX{} fields in
% all, and do a compulsory export.
%    \begin{macrocode}
  year "year" check format.date theyear "}{}{}" "" must.must.must
%    \end{macrocode}
% A new line for a new macro, and we're done!  Rejoice!
% On to the next function definition!
%    \begin{macrocode}
  newline$
%    \end{macrocode}
% \DescribeEnv{incollection}
% This is for those nasty entries that are created when someone
% publishes an article in a collection of essays edited by
someone
% else.
%    \begin{macrocode}
FUNCTION{incollection}
{ "\lexibib{article}{" cite$ "}{" must.must.must
  author "lastonly" format.names "author" check
  "}{" "" must.must.must
  title "title" check
  "}{" "" must.must.must
  chapter empty$
    { "" }
    { "\\" chapter " of \\" * * }
    if$
  booktitle "booktitle" check
  "}{" must.must.must
  pages "short" format.pages "pages" check
  "}{" "" must.must.must
% We're now in the final ``field'' of the \LexiTeX{} entry.
% The coding here is the same as for a "book" entry; the
% reader is referred to that entry for the commentary
% on the following code.
%    \begin{macrocode}
    editor empty$ translator empty$ and
    { series empty$ booktitle empty$ and
      'skip$
      { series booktitle either.or " No.~" number
      "endlabel" field.tag.no.combine ", " "" must.must.must }
      if$
    }
    { editor booktranslator either.or "firstinitial" format.names
      booktranslator empty$
      { editor num.names$ #1 >
        { " eds.\ " }
        { " ed.\ " }
        if$
      }
      { " trans.\ " }
      if$
      "" must.must.must
    }if$
  edition " ed.\ " "" might.ifone.must
  year "year" check format.date theyear "}" "" must.must.must
  newline$
%    \end{macrocode}
%    \begin{macrocode}
FUNCTION{booklet}
{ "\lexibib{man}{" cite$ "}{" must.must.must
  author "}{" "" must.must.must
  "\\" title "\\}{" must.must.must
  howpublished ", " "" might.ifone.must
  year "year" check format.date theyear
  "}{}{}" "" must.must.must
  newline$
%    \end{macrocode}
% \paragraph{Cases} Law cases are all entered using the
% {\tt\string@CASE} entry type.  The formatting of citations
% varies from jurisdiction to jurisdiction, so the behaviour
% of citations of this type is controlled via a "jurisdiction"
% field.  Supported jurisdictions are listed in the user
% guide.  Below, the functions for each jurisdiction are
% defined first, followed by the "case" function itself.
% \DescribeEnv{j.case}
% This function applies to cases from Japan.
%    \begin{macrocode}
FUNCTION { j.case }
{ "\lexibib{jcase}{" cite$ "}{" must.must.must
 "Decision of the " court "court" check "" must.must.must
 " (" j.format.division ")" iftwo.might.iftwo  
 ", " "" "" must.must.must
  casedate format.date themonth "short" format.month.name "\ "
  theday must.must.must
  ", " theyear "}{" must.must.must
  volume empty$ not number empty$ not and
  { journal "journal" check ", no.~" number must.must.must
    ", vol.~" volume "" must.must.must }
  { volume number either.or "\ " journal "journal" check
  might.ifone.must
    "}{" pages "}{" must.must.must
  }if$
  year "year" check empty.to.null format.date
  themonth "short" format.month.name " " "" might.ifone.must
  theday ", " "" might.ifone.must
  theyear "}" "{}" must.must.must
  newline$
%    \end{macrocode}
% \DescribeEnv{e.case}
% This function applies to English cases.
%    \begin{macrocode}
FUNCTION { e.case }
{ "\lexibib{ecase}{" cite$ "}{" must.must.must
  title "title" check "}{" "" must.must.must
  year "year" check format.date "[" theyear "] " must.must.must
  journal "}{" "" must.must.must
  pages "short" format.pages "}" "{}{}{}" must.must.must
  newline$
%    \end{macrocode}
% \DescribeEnv{p.case}
% This function applies to Pakistani cases.
%    \begin{macrocode}
FUNCTION {p.case}
{e.case}
%    \end{macrocode}
% \DescribeEnv{us.case}
% This function applies to U.S. cases.
%    \begin{macrocode}
FUNCTION { us.case }
{"\lexibib{case}{" cite$ "}{" must.must.must
 title "title" check "}{" "" must.must.must
 volume "\ " journal must.must.must
 "}{" pages "}{" must.must.must
 court "court" check "\ " year format.date theyear
 might.ifone.must
 "}{}{}" "" "" must.must.must
 newline$
%    \end{macrocode}
% \DescribeEnv{case}
% This function simply reads the contents of the "jurisdiction"
% field, and invokes the appropriate case function.
%    \begin{macrocode}
FUNCTION { case }
 { jurisdiction empty.to.null duplicate$ "jurisdiction (sine qua non!)" check
   "japan" =
   { pop$ j.case }
   { duplicate$ "england" =
     { pop$ e.case }
     { duplicate$ "pakistan" =
       { pop$ p.case }
       { duplicate$ "singapore" =
         { pop$ e.case }
         { pop$ us.case
         }if$
       }if$
     }if$
   }if$
%    \end{macrocode}
% The following item adds annotations; this may be eliminated by
% stripping with "noannotes".
%    \begin{macrocode}
%<*!noannotes>
annote empty.to.null write$ newline$
%</!noannotes>
%    \end{macrocode}
% \DescribeEnv{j.statute}
% This function applies to Japanese statutory materials.
% \changes{\LexiBib 1.0c}{94/08/03}{Added support for Japanese
statutes.}
%    \begin{macrocode}
FUNCTION {j.statute}
 { "\lexibib{jstatute}{" cite$ "}{" must.must.must
   title "title" check "}{" "" must.must.must
   "Law no.~" number "number" check " of "
   iftwo.might.iftwo
   year format.date theyear "}{}{}{}" "" must.must.must
   newline$
%    \end{macrocode}
% \DescribeEnv{s.statute}
% This function formats a statute entry for Singapore.
%    \begin{macrocode}
FUNCTION { s.statute }
{ "\lexibib{statute}{" cite$ "}{" must.must.must
  title "title" check ", No.~" number
  "got it" warning$ "endlabel" field.tag.no.combine
  number empty$
    { "\ " * }
    { "\ of " * }
    if$
  year format.date theyear "}{" must.must.must
  "}{}{}{}{}" "" "" must.must.must
  newline$
%    \end{macrocode}
% \DescribeEnv{statute}
% This function selects the correct statute entry function.
%    \begin{macrocode}
FUNCTION { statute }
 { jurisdiction empty.to.null duplicate$
   "japan" =
   { pop$ j.statute }
   { duplicate$ "singapore" =
     { pop$ s.statute }
     { pop$ "IMPORTANT: unknown jurisdiction for " cite$ * warning$
     }if$
   }if$
%    \end{macrocode}
% \DescribeEnv{default.type}
% We use the "book" type as our default type.  When "manual" is
% completed, we should probably use that type instead.
%    \begin{macrocode}
FUNCTION {default.type} { book }
%</newbst>
%    \end{macrocode}
% \subsection{Macro definitions}
% We don't define any macros for abbreviating law journal
% names.  Instead, we will use Blue Book abbreviations
% ``native'', with a special syntax (probably the full
% form in syntax: 
% ``"\gobble{Accountant}{}"'' immediately after
% the abbreviation) for resolving
% ambiguous abbreviations.  Meanwhile, trust me: use the
% Blue Book abbreviations and take this upcoming facility
% on faith.  And besides, do you ever {\em need\/} to spell
% out journal and reporter names?
% \subsection{Execution}
% With all preliminaries out of the way, our first act is
% to read in the entries from {\tt *.bib}..
%    \begin{macrocode}
%<*newbst>
%    \end{macrocode}
% Then we say ``Hi'' to the user.  It would be nice to make this
% the first message, but the structure of \BibTeX{} style files
% dictates that it will follow any warnings about missing
% entries.
%    \begin{macrocode}
EXECUTE {hello}
%    \end{macrocode}
% \DescribeEnv{begin.bib}
% \DescribeEnv{end.bib}
% We need a couple of functions to write the starting and
% finishing strings on to output file.
%    \begin{macrocode}
FUNCTION {begin.bib}
 { "\begin{lexilist}" write$ newline$
FUNCTION {end.bib}
 { "\end{lexilist}" write$ newline$
%    \end{macrocode}
% Now we put it all together by running the entry-type
% function on each entry selected from the database in turn,
% bracketed by the starting and ending output functions
% just defined.
%    \begin{macrocode}
EXECUTE {begin.bib}
ITERATE {call.type$}
EXECUTE {end.bib}
%</newbst>
%    \end{macrocode}
% And, er, that's it!
% \section{Extraction utilities}
% \subsection{The Driver}
% Here is a simple driver for extracting the files in the
% package.
%    \begin{macrocode}
%<*driver>
\def\batchfile{lexibib.ins}
\input docstrip
\preamble
\endpreamble
\generateFile{lexibib.man}{t}{\from{lexibib.doc}{manual}}
\generateFile{lexitest.tex}{t}{\from{lexibib.doc}{testdoc}}
\postamble
\endpostamble
\generateFile{lexitest.bib}{t}{\from{lexibib.doc}{testbib}}
\generateFile{lexibib.bst}{t}{\from{lexibib.doc}{newbst}}
%</driver>
%    \end{macrocode}
% \subsection{The Manual}
% This is a simple driver that generates the documentation.
%    \begin{macrocode}
%<*manual>
\let\lawhline=\hline
\documentclass{ltxdoc}
\usepackage{lexitexm}
% \makebooktable
\EnableCrossrefs
\RecordChanges                  % Gather update information
\CodelineIndex                  % Index code by line number
%\OnlyDescription  % comment out for implementation details
\begin{document}
\DocInput{lexibib.doc}
\end{document}
%</manual>
%    \end{macrocode}
%    \begin{macrocode}
%<*testbib>
@STRING{tdc..="Tokyo District Court"}
@STRING{minjibu..="Civil Division"}
@STRING{ksh..="Kin'y{\=u} sh{\=o}ji hanrei"}
@PREAMBLE{ "\newcommand{\kk}{K.K. } "
          # "\newcommand{\v}{v.~} " }
@book{tanaka-tekisuto,
   author    = {Akira Tanaka and others},
   title     = {Tekisutobukku Tegata,  Kogitte H{\=o}},
   year      = {1987},
   publisher = {Y{\=u}hikaku Bukkusu},
   subject   = {commtrans}
@case{tdc.s57.9.27,
   journal   = ksh..,
   volume    = {663},
   annote    = {
       \holding{
       \item The Business Suspension Sanction of the Notes
Clearing Center falls
       within the terms of the explicit exemption of Notes and
Cheque Clearing
       Centers from the terms of Anti-Monopoly Act section 8, as
a practice
       necessary to the specialized function of such Centers.
       }},
   court     = tdc..,
   division  = minjibu..,
   divno     = {30},
   casedate  = {s57.9.27},
   page      = {35},
   jurisdiction = {j},
   subject   = {commtrans: aml attack on clearing system}
@book{horitsu-jiten,
   title     = {H\=oritsu jiten},
   year      = {s11.3.10},
   editor    = {? Suehiro and ? Tanaka},
   publisher = {Iwanami sh\=oten},
   volume    = {1},
   subject   = {general}
%</testbib>
%    \end{macrocode}
%    \begin{macrocode}
%<*testdoc>
\documentclass{article}
\usepackage{lexitex}
\begin{document}
\bibliography{lexitest}
\bibliographystyle{lexibib}
Waffle burble.\footnote{\lexicite{tanaka-tekisuto}.}
More waffle and burble.\footnote{\lexicite:2:{horitsu-jiten}.}
\lexicite{tdc.s57.9.27}:
\theholding
\end{document}
%</testdoc>
%    \end{macrocode}
%   \changes{no ver}{84/12/16}{   (HWT)       Original `plain'
% version, by Howard Trickey.}
%   \changes{no ver}{84/12/23}{     (LL)     Some comments made
% by Leslie Lamport.}
%    \changes{no ver}{85/2/16}{     (OP)     Changes based on
% LL's comments, Oren Patashnik.}
%    \changes{no ver}{85/2/17}{     (HWT)     Template file and
% other standard styles made.}
%    \changes{0.98b}{85/3/28}{     (OP)     First
% release,  for BibTeX 0.98f.}
%    \changes{0.98c}{85/5/9}{     (OP)      for BibTeX
% 0.98i:
%               fixed Theoretical Computer Science macro name;
%               fixed the format.vol.num.pages function.}
%    \changes{0.99a}{88/1/24}{     (OP)      for BibTeX
% 0.99a, main changes:
%               assignment operator (:=) arguments reversed;
%               the preamble$ function outputs the database
% PREAMBLE;
%               entry.max$ and global.max$ (built-in) variables
% replace
%               entry.string.max and global.string.max functions;
%               alphabetizing by year then title, not just title;
%               many unnecessary ties removed; \it ==> \em;
%               the `alpha' style uses a superscripted `+'
% instead of a
%               `*' for unnamed names in constructing the label;
%               the `abbrv' style now uses "Mar." and "Sept.";
%               the functions calc.label and presort now look at
% just
%               the fields they're supposed to;
%               BOOKLET, MASTERSTHESIS, TECHREPORT use nonbook
% titles;
%               INBOOK and INCOLLECTION take an optional type
% (e.g.
%               type = "Section"), overriding the default
% "chapter";
%               BOOK, INBOOK, INCOLLECTION, and PROCEEDINGS now
% allow
%               either volume or number, not just volume;
%               INCOLLECTION now allows an edition and series
% field;
%               PROCEEDINGS and INPROCEEDINGS now use the address
% field
%               to tell where a conference was held;
%               INPROCEEDINGS and PROCEEDINGS now allow either
% volume
%               or number, and also a series field;
%               MASTERSTHESIS and PHDTHESIS accept types other
% than
%               "Master's thesis" and "PhD thesis";
%               UNPUBLISHED now outputs, in one block, note then
% date;
%               MANUAL now prints out the organization in
%               the first block if the author field is empty;
%               MISC can't be empty---it requires some optional
% field.}
%    \changes{0.99b}{88/3/23}{     (OP)      For BibTeX
% 0.99c---changed the three
%               erroneous occurrences of `{\tt cite }'
% to `{\tt cite\$ }';
% this
%               change didn't affect the four standard styles,
% so the
%               0.99a versions of those styles are still
% current.}
%    \changes{\LexiBib{} 1.0}{94/6/12}{  (FB)   Copied into
% {\tt btxbst.doc} to {\tt lexibib.doc} and commenced editing.}
% \changes {\LexiBib{} 1.0a}{94/7/25}{(FB) First successful
% trial of \LexiBib.}
% \changes{\LexiBib{} 1.0b}{94/8/9}{(FB) Completed first
% operational version, with date parsing, for use with
% articles, books and Japanese case materials.}
% \Finale \PrintIndex \PrintChanges