%%% ====================================================================
%%% @LaTeX-doc-source-file{
%%%   filename  = "amsrefs.dtx",
%%%   version   = "1.23",
%%%   date      = "2002/02/28",
%%%   time      = "09:56:29 EST",
%%%   author    = "Michael J Downes",
%%%   address   = "American Mathematical Society,
%%%                Publications Technical Group,
%%%                PO Box 6248,
%%%                Providence, RI 02940,
%%%                USA",
%%%   email     = "tech-support@ams.org",
%%%   URL       = "http://www.ams.org/",
%%%   abstract  = "A LaTeX package that permits bibliography style
%%%                to be controlled completely from the LaTeX side
%%%                instead of being determined chiefly by the BibTeX
%%%                style file.",
%%%   checksum  = "38825 4863 19408 161569",
%%%   docstring = "The checksum field, produced by Robert Solovay's
%%%                checksum utility, gives CRC-16 checksum, lines,
%%%                words, and characters.",
%%%  }
%%% ====================================================================
% \iffalse
%<*driver>
\NeedsTeXFormat{LaTeX2e}
\documentclass{amsdtx}
\let\mdash=\relax \let\ndash=\relax \let\qq=\relax
%\OnlyDescription
\usepackage{amsrefs}
\usepackage{multicol}

%    Some or all of these are provided by amsdtx.cls version 2.
\providecommand{\qq}[1]{\textquotedblleft#1\textquotedblright}
\providecommand{\url}[1]{\texttt{#1}}
\providecommand{\secref}[1]{Section~\ref{#1}}
\providecommand{\bibtex}[1]{Bib\protect\TeX{}}
\providecommand{\lat}[1]{\protect\LaTeX{}}

\DeclareTextSymbol{\lbracechar}{OT1}{123}
\DeclareTextSymbolDefault{\lbracechar}{OT1}
\DeclareTextSymbol{\rbracechar}{OT1}{125}
\DeclareTextSymbolDefault{\rbracechar}{OT1}

\newcommand{\rdag}{{\normalfont\dag}}
\newcommand{\rast}{{\normalfont *}}
\providecommand{\embrace}[1]{%
  \begingroup \ntt\lbracechar#1\rbracechar\endgroup}

% One mandatory argument: cnm; two: cnmm; one optional and one
%    mandatory: cnom; throwing in a star: cnsom; and so on.
\providecommand{\cnbreak}{\penalty999\hskip0pt\relax}
\providecommand{\cnm}[2]{\cn{#1}\cnbreak\embrace{#2}}
\providecommand{\cnmm}[3]{\cn{#1}\cnbreak\embrace{#2}\cnbreak\embrace{#3}}
\providecommand{\cnom}[1]{\cn{#1}\cnbreak\ommitude}
\providecommand{\ommitude}[2][]{{\ntt[#1]}\cnbreak\embrace{#2}}
\providecommand{\cnmsm}[4]{%
  \cn{#1}\cnbreak\embrace{#2}\cnbreak *\embrace{#4}}

\providecommand{\strong}{\textbf}
\newcommand{\rpack}[1]{\pkg{amsrefs} package}
\newcommand{\bibtexnote}[1]{\par \noindent\kern-2em
  \textbf{\bibtex/ Note:} #1\par}

% non-stupid hline: support for adding some space below
\newcommand{\nshline}{\hline\omit\\}

\newcommand{\SeveralOptions}[1]{%
  \vbox{\def\\{\unskip,\egroup\hbox\bgroup\ignorespaces}%
    \hbox{\strut#1}}}

% URL tilde character:
\newcommand{\urltilde}{\symbol{126}}

% Make this a self-contained document without separate .bbl file.
\newcommand{\PrintBibliography}{%
  \begin{bibsection}\begin{biblist}\raggedright
  \bib{A97}{article}{
    author={Aphalo, Pedro J.},
    title={A proposal for citation commands in \lat/3},
    journal={TUGboat},
    volume={18},
    number={4},
    date={1997-12},
    pages={297\ndash 302},
  }
  \bib{B}{misc}{
    author={Beebe, Nelson},
    title={\fn{xbtxbst.doc}},
    url={http://www.math.utah.edu/pub/tex/bibtex/},
    note={software documentation for extended bst files},
  }
  \bib{CMS}{book}{
    title={Chicago Manual of Style},
    edition={13},
    date={1982},
    publisher={University of Chicago Press},
  }
  \bib{J}{misc}{
    author={Jacobsen, Dana},
    title={\textnormal{\ttfamily bp}, a Perl Bibliography Package},
    note={version 0.2.97 beta},
    date={1996-12-19},
    url={http://www.ecst.csuchico.edu/\til jacobsd/bib/bp/index.html},
  }
  \bib{PKK99}{article}{
    author={Piotrowski, Michael},
    author={Kl\"{o}cker, Jens},
    author={Knappen, J\"{o}rg},
    title={Is \LaTeX 2$_\epsilon$ markup sufficient for
           scientific articles?},
    conference={Euro\TeX{}'99 (Giessen, Augsburg, 1999)},
    isbn={1438-9959},
  }
  \end{biblist}\end{bibsection}
}
\begin{document}
\title{The \pkg{amsrefs} package}
\author{Michael Downes\\American Mathematical Society}
\date{Version \fileversion, \filedate}
\hDocInput{amsrefs.dtx}
\end{document}
%</driver>
% \fi
%
% \MakeShortVerb{\|}
%
% \maketitle
%
% \section{Introduction}
%
%    The \rpack/ is a \lat/ package for bibliographies that provides an
%    archival data format similar to the format of \bibtex/ database
%    files, but adapted to make direct processing by \lat/ easier. The
%    package can be used either in conjunction with \bibtex/ or as a
%    replacement for \bibtex/.
%
%    As compared to the standard methods for putting a bibliography into
%    a \lat/ document (running \bibtex/ to extract selected items from a
%    \fn{.bib} database into a formatted \fn{.bbl} file), the chief
%    advantages of the \rpack/ for an author or for someone writing a
%    document class are as follows:
%    \begin{description}
%    \item[Preservation of structure]
%
%    The internal structural information of the bibliography entries is not
%    lost when they are imported from the database file into the \lat/
%    document. This takes on its greatest significance when archiving
%    documents in \lat/ form or transmitting them to another user (such as a
%    publisher).
%
%    \item[Deferred formatting]
%
%    This means that the style of the bibliography can be readily changed
%    without reimporting everything from the original database(s).
%
%    \item[More natural handling of titles]
%
%    Proper nouns do not need to have braces added to prevent capitalization
%    problems.
%
%    \item[Less ambiguous format for author names]
%
%    When author names are given in inverted order (last name first) it is
%    possible for \lat/ to unambiguously identify the last name without any
%    further markup, even in cases like Saunders Mac Lane (Mac Lane, Saunders)
%    versus Stephen H. Lane (Lane, Stephen H.), or Cam Van Tran (Tran, Cam
%    Van) versus Bert Van Keulen (Van Keulen, Bert) that require the
%    addition of extra braces in \bibtex/.
%
%    \item[Setup requires only \lat/ knowledge]
%
%    All bibliography setup can be done in \lat/; learning another
%    programming language (such as, the one used in \bibtex/ bst files) is
%    unnecessary.
%
%    \item[Self-printable database files]
%
%    A \lat/ document that contains only a bibliography in amsrefs forms can
%    be used as a database for exporting entries to other documents. And
%    because it is a \lat/ document, the database can be printed directly at
%    any time simply by running it through \lat/ in the usual way.
%
%    \item[Self-contained]
%
%    In many cases it seems possible to do without \bibtex/ entirely. For
%    example, if the entries are extracted from a single database file that
%    is maintained in sorted order, the bibliography can be printed directly
%    by \lat/ on the first pass and the citations resolved on the second
%    pass.
%    \end{description}
%
% \section{Using the \pkg{amsrefs} package}
%
%    There are three main ways of using the \rpack/:
% \begin{enumerate}
% \item Using the standard \cn{bibliography} command to import a
%    \fn{.bbl} file that was created by \bibtex/ with the \bst{amsxport}
%    bibliography style.
%
% \item Using the \cn{bibselect} command. This is similar to using
%    \cn{bibliography}, but with an extension of \fn{.ltb} (``\lat/
%    bibliography'') for database files, rather than \fn{.bib}, and with
%    the extraction of the selected entries being done by \lat/ itself
%    on the first pass.
%
% \item Writing bibliography items in \pkg{amsrefs} form directly into
%    your \lat/ document.
%
% \end{enumerate}
%
% \subsection{Using the \rpack/ with \bibtex/}
%
%    When \bibtex/ is used to generate a bibliography for use with the
%    \rpack/, your main document will look almost exactly the
%    same as usual, with the \cn{bibliography} command used in the
%    normal way. The only overt difference would be one line at the
%    beginning to load the \rpack/. The bibliography style to use
%    would be a generic one called \fn{amsxport} (see the \fn{amsxport}
%    documentation for more details).
%\begin{verbatim}
%\documentclass{article}
%\usepackage{amsrefs}
% ...
%\begin{document}
% ...
%\bibliographystyle{amsxport}
%\bibliography{bib-file-name}
%\end{verbatim}
%
%    If you look at the \fn{.bbl} file generated by \bibtex/, you will
%    see that there are two main differences:
% \begin{enumerate}
% \item Instead of a \env{thebibliography} environment, the bibliography
%    is wrapped in two environments:
%\begin{verbatim}
%\begin{bibdiv}
%\begin{biblist}
% ...
%\end{biblist}
%\end{bibdiv}
%\end{verbatim}
%
% \item Each entry in the bibliography is done with \cn{bib} instead of
%    \cn{bibitem}, as discussed below in more detail.
% \end{enumerate}
%
% \subsection{Using \cn{bibselect} instead of \cn{bibliography}}
%
%    One feature of the \rpack/ is the possibility of maintaining your
%    bibliography data in the form of a \lat/ document that can be
%    directly printed, and using \lat/ itself instead of \bibtex/ for
%    importing cited items into other documents. If you write
%\begin{verbatim}
%\bibselect{combinatorics}
%\end{verbatim}
%    instead of \cnm{bibliography}{combinatorics}, then \lat/ will look
%    for a file \fn{combinatorics.ltb} (instead of
%    \fn{combinatorics.bib}), and search through it for a \env{biblist}
%    environment. The \cn{bib} entries in the list will be checked to
%    see if they match \cn{cite} commands in the calling document, and
%    if so they will be copied into a \fn{.bbl} file.
%
%    Because this is all being handled by \lat/, the contents of the
%    bibliography can be printed on the first pass; cite numbers are
%    resolved, consequently, on the second pass.
%
%    When \cn{bibselect} is used, the \cn{bibliographystyle} command is
%    irrelevant since its purpose is merely to pass information to
%    \bibtex/.
%
%    There is one significant difference with the \cn{bibselect}
%    command: it produces only the actual contents of the bibliography,
%    without the section title and without any bibliography environment
%    wrapper. This is to make it possible to put them in your main
%    document and freely customize them there, including (for example)
%    putting some introductory text at the beginning of the
%    bibliography. In other words, you can write something like
%\begin{verbatim}
%\begin{bibchapter}[Annotated Bibliography]
%(Some introductory text here.)
%\begin{biblist}
%\bibselect{combinatorics}
%\end{biblist}
%\end{bibchapter}
%\end{verbatim}
%
%    The \cn{bibselect} command always regenerates the \fn{.bbl} file
%    unless the \cn{nofiles} switch is used. When a time comes that you
%    no longer want it to be regenerated, comment out the
%    \cn{bibselect} command and copy the bbl file into your \lat/ file.
%    This is particularly relevant if you put any line break or page
%    break commands into the bibliography.
%
% \subsection{Direct entry of bibliography items}
%
%    When the bibliography is written by hand each entry is written with
%    \cn{bib} instead of \cn{bibitem}, and has the form
%\begin{verbatim}
%\bib{BW}{article}{
%    author={Bertram, A.},
%    author={Wentworth, R.},
%     title={Gromov invariants for holomorphic maps on Riemann
%            surfaces},
%      date={1996},
%   journal={J. Amer. Math. Soc.},
%    volume={9},
%    number={2},
%     pages={529\ndash 571},
%}
%\end{verbatim}
%    The first and second arguments give the citation key and type; the
%    remainder is a list of bibliographic field names and field values
%    in more or less standard \latex/ key-value syntax. The syntax for
%    \cn{bib} is intentionally restricted, however, to eliminate certain
%    variations---such as omitting the braces around the value---that
%    are not really a good idea when the values can contain arbitrary
%    \latex/ text.
%
%    If you try leaving out an equal sign or the comma at the end of a
%    line, you will see that the warning messages for syntax errors
%    usually identify the type and location of the error precisely
%    enough that making the corrections is straightforward. (For more
%    details see the documentation of the \pkg{rkeyval} package.)
%
% \section{Package options}
%
% The recognized options for \pkg{amsrefs} include:
% \begin{description}
%
% \item[?] Informational option. This causes \lat/ to produce some extra
%    informational messages in the \lat/ log, including a list of the
%    package options.
%
% \item[lite] Omit the loading of the auxiliary packages \pkg{textcmds}
%    and \pkg{mathscinet} (they can still be loaded separately if that
%    is what it takes to clear up timing problems). The \pkg{textcmds}
%    package defines a number of short forms for various text symbols,
%    which might lead to a name conflict with one of your own
%    definitions, or (possibly) with another package.
%
% \item[\SeveralOptions{numeric\\alphabetic\\author-year}]
%    These three options switch between the three main
%    bibliography styles supported ``out of the box'' by the
%    \pkg{amsrefs} package. If you want anything significantly
%    different, see \secref{other-styles}. The \opt{numeric} style is
%    the default.
%
% \item[short-journals] Print short form instead of full form for
%    journal names. If the data contains only the short form, that of
%    course is all you will get in either case.
%
% \item[short-months] Print short version of month names (e.g., Jan.\@
%    instead of January).
%
% \item[initials] Print author\slash editor names with initials instead
%    of full given name. This only affects the primary names for each
%    entry; anything more complicated is considered a matter of variant
%    style and must be specified in detail by whoever wants the
%    different style\mdash see \secref{name-style}.
%
% \item[\SeveralOptions{traditional-quotes\\logical-quotes}]
%    With the \textit{traditional quotes} option
%    (default), quotation marks produced by \cs{bibquotes}
%    (\secref{misc-cmds}) fall outside of other punctuation, ``like
%    this,'' whereas with the \textit{logical quotes} option the order
%    is reversed, ``like this''.
%
% \item[\SeveralOptions{sorted-cites\\non-sorted-cites\\
%       non-compressed-cites\\compressed-cites}]
%
%    These four options apply only when the \opt{numeric} option is in
%    effect. They cause multiple cites written with \cn{cites} or
%    \cn{citelist} to be sorted into numerical order and/or consecutive
%    numbers (three or more) to be written in range form:
%    [11,9,10,1,16,12] will be printed as [1,9--12,16] if both the
%    \opt{sorted-cites} and \opt{compressed-cites} options are in
%    effect. And by default they are.
%
% \item[backrefs] This option causes ``back-references'' to be printed
%    at the end of each bibliography entry to show where it was cited in
%    the body of the document.
%
% \item[\SeveralOptions{sorted\\citation-order}]
%
%    Should the bbl file be compiled in sorted order or citation order?
%    When using \bibtex/: don't forget that changing this option will
%    have no effect until after the next \bibtex/ run. When importing
%    from an \fn{.ltb} file: No real sorting is done (at the time of
%    this writing: 2002-02-27). You get whatever ordering was present in
%    the \fn{.ltb} file. The easiest way of working is therefore to
%    maintain the \fn{.ltb} file in sorted order.
%
% \item[\SeveralOptions{beta\\jpa}]
%
%    Obsolete; these applied only to the beta version of the
%    \pkg{amsrefs} package.
%
% \end{description}
%
% \section{More about the \cn{bib} command}
%
%    The \cn{bib} command is a drastically different replacement for the
%    \cn{bibitem} command. Compare the example shown above as it would
%    look if done with \cn{bibitem}:
%\begin{verbatim}
%\bibitem{BW} A. Bertram and R.~Wentworth,
%  \emph{Gromov invariants for holomorphic maps on {R}iemann surfaces},
%  J. Amer. Math. Soc. \textbf{9} (1996), no.~2, 529--571.
%\end{verbatim}
%
%    With the \cn{bib} command, the form of the data in the \lat/
%    document is fairly close to the form used in \bibtex/ database
%    files. There are, however, some significant differences.
% \begin{description}
%
% \item[braces] Always use braces to enclose the value of each field
%    (standard \lat/ syntax), never double quotes (\bibtex/ allows both
%    quotes and braces).
%
% \item[repeated fields] Fields can be repeated (and should be, where
%    applicable). In this example each author name is given separately.
%    The task of combining the names as needed for the current
%    publication is handled by \lat/.
%
% \item[inverted names] Author and editor names should be written
%    uniformly in \emph{Last, First} order. This is the form that
%    provides the maximum flexibility with the least ambiguity and least
%    extra markup. When printed, the names will be automatically
%    uninverted in whatever manner is specified by the bibliography
%    style in use. (For example, some styles have the first author's
%    name inverted and the remaining names not inverted.)
%
% \item[capitalization] In titles, do not capitalize words
%    that are not proper nouns. Capitalization will be applied
%    automatically as needed.
%
%    (In contrast to \bibtex/, proper nouns do not need to be written
%    with extra braces: you can simply write \texttt{Riemann} instead of
%    \texttt{\string{R\string}iemann} or
%    \texttt{\string{Riemann\string}}.)
%
% \item[ndash] Use of \cn{ndash} instead of \verb'--' for en-dashes is
%    recommended (the \verb'--' notation, which is tied to
%    idiosyncrasies of the Computer Modern \qq{\TeX{} text} font
%    encoding, hinders document translation and the use of alternate
%    fonts).
%
% \item[date] It is recommended to use a \fn{date} field to give the
%    year of publication; although \fn{year} is also accepted, \fn{date}
%    is more general. If the date includes a month, it is recommended to
%    use numbers for months in ISO form, e.g., 1987-12. This allows
%    month names to be printed in full or abbreviated at will. For
%    \qq{Winter}, \qq{Spring}, \qq{Summer}, \qq{Fall}, either use month
%    numbers of 13, 14, 15, 16 (respectively), or just put in the text
%    before the year:
%\begin{verbatim}
%  date={Summer 1987},
%\end{verbatim}
%
% \end{description}
%
% \subsection{Field names for the \cn{bib} command}
%
%    The following field names are currently supported by the \rpack/:
%\begin{multicols}{3}\ttfamily\noindent
%author\rdag\\
%editor\rdag\\
%isbn\rdag\\
%review\rdag\\
%title\\
%booktitle\rast\\
%series\\
%conference\\
%publisher\\
%organization\\
%institution\\
%place\\
%address\rast\\
%date\\
%year\rast\\
%journal\\
%volume\\
%number\\
%pages\\
%part\\
%issn\\
%doi\\
%language\\
%hyphenation\\
%\mbox{}\\
%status\\
%eprint\\
%preprint\\
%edition\\
%note\\
%url\\
%type\\
%\mbox{}\\
%xref\\
%label\\
%setup\\
%name
%\end{multicols}
%    The \rdag\ indicates that a field is repeatable. The
%    \rast\ indicates an alternative field name supported
%    for compatibility with \bibtex/; the field name just above a
%    starred name is the equivalent provided by \pkg{amsrefs}. For
%    example, the \fn{date} field can contain not only the year but also
%    month and day, if relevant, in ISO 8601 form: 1987-06-21. If month
%    and/or day are present, they can be printed in different forms
%    depending on the current bibliography style.
%
%    Although the meaning of many fields should be self-evident, for
%    some of them explanatory remarks are in order.
%\begin{description}
%\item[author] This field can be repeated. Multiple author names will be
% concatenated into a single field value. Names should be given in
% \emph{Last, First} order (\emph{von Last, First, Jr} for more
%    complicated cases).
%\item[date] This is a generalization of the \fn{year} and \fn{month}
%  fields. Its value should be written in ISO 8601 format, e.g.,
%  1987-06-05; but the day and month are omissible, so this can be
%  used instead of the \fn{year} field.
%\item[doi] Digital Object Identifier
%
%\item[edition] For books. If the value of this field is a simple
%  number, \cs{bib} will convert it to cardinal form and add \qq{ed.}
%  (or alternative text if specified by the bibliography style).
%  Otherwise it will be printed as~is.
%
%\item[editor] Like \fn{author}; can be repeated.
%\item[eprint] Electronic preprint information such as for
%  \url{www.arXiv.org}. See
%  \url{http://xxx.lanl.gov/help/faq/references} for recommended form.
%\item[isbn] International Standard Book Number. Can be repeated.
%\item[issn] International Standard Serial Number.
%\item[language] Language of the work.
%  The language name should be the printed form, not a Babel-style language
%  name, since in principle this field could contain more complicated
%  remarks such as \qq{Russian, with French abstract}. Cf.\@
%    \fn{hyphenation}.
%
%\item[hyphenation] This corresponds to the Babel package notion of
%    `language'. The hyphenation language used for a given
%    \cn{bib} entry is determined from various clues, which are checked
%    in the following order:
% \begin{enumerate}
% \item The Babel language specified by the \fn{hyphenation} field.
% \item The Babel language specified by the \emph{first word} of the
%    \fn{language} field (after lowercasing).
% \item The current Babel language that was in effect before the
%    \cn{bib} command started.
% \item The current hyphenation patterns of the document, if
%    there are no Babel language modules loaded.
% \end{enumerate}
%
%\item[organization] The \bibtex/ documentation says that
%  \fn{institution} should be used for technical reports and
%  \fn{organization} for other entry types, whereas \fn{school} should
%  be used for theses. Having three different field names for these
%  strikes me as overkill, so I omitted \fn{school}.
%
%\item[part] This is for a long journal article that is published in
%  separate issues of the same journal.
%
%\item[place] A synonym for address. Or to put it another way,
%  \fn{address} is supported as a \bibtex/-compatible synonym for
%  \fn{place}.
%
%\item[review] A review number or similar pointer, e.g., for
%  Mathematical Reviews or Zentralblatt. Can be repeated.
%
%\item[setup] This is a special field that can be used to give
%  arbitrary commands to be executed at the beginning of the current
%  \cs{bib} entry, after all the fields have been read. The idea is
%  that one can alter the formatting of an individual entry through
%  this field, to handle special cases.
%
%\item[status] Typically used for notes such as \qq{to appear} or \qq{in
%  preparation} with journal articles.
%
%\item[subtitle] Typically used with a multipart journal article to
%  give a subtitle for each part.
%
%\item[translator] Like \fn{author}; can be repeated.
%
%\item[url] Universal Resource Locator.
%
%\item[xref] This is used when cross-referring to another \cn{bib} entry.
%    All of the fields of the xref'd entry are copied into the data of the
%    current entry. If an article has a cross-reference to an entry of
%    type \fn{collection}, the collection title is imported into the
%    article's \fn{booktitle} field, not the \fn{title} field.
%
%\end{description}
%
% \subsection{Bibliography entry types}
%    It behooves a bibliography package that has ambitions of serving
%    for general use to support a rather wide range of entry types,
%    e.g.,
%\begin{multicols}{3}\ttfamily\noindent
%book\\
%collection\\
%series\\
%set\\
%article\\
%periodical\\
%report\\
%thesis\\
%manual\\
%reference\\
%supplement\\
%fragment\\
%review\\
%play\\
%screenplay\\
%personal\\
%lecture\\
%speech\\
%reading\\
%story\\
%image\\
%artwork\\
%map\\
%audio\\
%music\\
%video\\
%film\\
%software\\
%miscellaneous
%\end{multicols}
%
%    The current release of \pkg{amsrefs} does not, however, provide
%    support out of the box for all of the above types, but only the
%    ones that are commonly used in AMS publications. On the other hand,
%    adding support for a new type is scarcely any more work than
%    modifying an existing type: in either case if you write down the
%    desired output style in a \cn{BibSpec} command, voil\`a, the type
%    is now supported. If it needs any unusual fields, you will need to
%    define the extra field names as well, but adding a new field
%    normally requires only one line, a \cn{DefineSimpleKey} statement
%    (see the examples in \fn{amsrefs.sty}).
%
%    Discussion of individual types:
%    \begin{description}
%
%    \item[book] A written work by one or more authors where the authors
%    share credit for the work as a whole.
%
%    \item[collection] A written work by multiple authors where distinct
%    parts are credited to particular authors. This includes conference
%    proceedings volumes. Collected works of a single author would also
%    be classed as a collection, even though everything is written by
%    the same person.
%
%    \item[series] A sequence of \emph{books} or \emph{collections} with
%    a common series name given by the publisher. New releases are
%    published as they become ready, rather than following a definite
%    regular schedule (compare to \emph{periodical}).
%
%    \item[set] Like \emph{book} or \emph{collection} but split into
%    multiple volumes, where the volumes might be published either
%    simultaneously or sequentially as they are finished; as compared to
%    \emph{series}, a \emph{set} has a more definite (usually
%    pre-established) number of volumes, and is usually the work of a
%    single author, or collaborating authors, rather than multiple
%    independent authors. A set may consist of books, articles
%    collections, reports, etc.---e.g., a multi-part article published
%    over several issues of a journal can be classified as a set to
%    facilitate citing it as a whole---but it would be unusual for a set
%    to contain periodicals or series.
%
%    \item[article] A relatively short---usually less than 50
%    pages---written item that appeared as part of a periodical or
%    collection; or, possibly, as a World Wide Web document, in which
%    case the URL should be included in the citation info as well as the
%    date on which it was read.
%
%    \item[periodical] Like a \emph{series} of \emph{collections} but
%    following a definite schedule of publication, usually two times or more
%    per year. This includes newspapers, magazines, and journals.
%
%    \item[report] Technical description, white paper, or the like.
%    Similar to \emph{article} but produced by the author(s) for a
%    particular organization, usually their employer, whose primary
%    business is usually not publishing.
%
%    \item[thesis] Like \emph{report} but produced by the author for an
%    educational institution to satisfy the requirements for a degree.
%    Not (normally) produced by a commercial publisher for sale.
%
%    \item[manual] A special case of \emph{book}: Instructions for dealing
%    with a particular kind of software or manufactured product.
%
%    \item[reference] This could be considered as a general type subsuming
%    indexes into various kinds of literature, dictionaries, and other
%    kinds of books where items are looked up by name or title.
%
%    \item[story] Short story, longer story, fiction
%
%    \item[review] Like \emph{article}, usually, but specifically to
%    discuss some other citable item.
%
%    \item[personal] A personal communication in the form of mail, e-mail,
%    phone call, face-to-face conversation, telegram, etc.; unlike
%    most other cited items, personal communications are usually not
%    available in libraries.
%
%    \item[lecture] This type provides for a kind of ``publication'' that
%    consists primarily in the verbal presentation, or possibly in printed
%    lecture notes given out at the time of the lecture. It is somewhat rare
%    for these to have a formal publisher. A lecture might be published in
%    audio or video form.
%
%    \item[speech] Political or academic address intended for persuasive
%    rather than pedagogic purposes. A commencement address would not
%    normally be considered a lecture.
%
%    \item[reading] A public reading of poetry or fiction. This could
%    be subsumed under lecture in respect of the etymology.
%
%    \item[image] Visual image in some form or another. See also \emph{map}.
%
%    \item[artwork] Similar to image, but could also include sculpture and
%    other types of visual art.
%
%    \item[map] Like \emph{image} but specifically for spatial and
%    navigational information about physical locations.
%
%    \item[audio] Speech captured in the form of sounds (rather than
%    transcription) on audio-cassette or other media. See also \emph{music},
%    \emph{lecture}, \emph{speech}.
%
%    \item[music] Like \emph{audio} but restricted to items that are
%    generally held to be \emph{music}.
%
%    \item[video] Moving pictures, usually also accompanied by sound.
%
%    \item[film] A slight variation of \fn{video}: Moving pictures that
%    were released ``on the big screen''.
%
%    \item[software] Computer programs and electronic files whose main
%    purpose is to contribute to the operation of a program, rather than
%    to be read by a human being. \emph{\TeX: the Program} (Knuth, 1986)
%    is an example of a computer program published in book form; when
%    such a form exists it could be cited either as a book or as
%    software. Consider what will best serve the reader.
%
%    \item[play] Drama.
%
%    \item[screenplay] A variant of \emph{play}.
%
%    \item[miscellaneous] None of the above, including pamphlets,
%    restaurant menus, patents, grant proposals, court records, business
%    plans, or what have you.
%
%    \item[supplement] Anything added to a parent published item
%    (including supplements, corrigenda, errata, etc.), that appeared
%    separately from the original item.
%
%    \item[fragment] Some part of one of the above types, such as a
%    chapter from a book, an appendix from an article, a page from a
%    World Wide Web document, etc.
%    \end{description}
%
%    Some readers will surely be wondering at this point: ``Why the
%    absence of a \fn{url} type? An \fn{incollection} type?'' The answer
%    is that it seems better normally to use a type that indicates the
%    intrinsic nature of the writing, rather than the context in which
%    it occurs. An article that appears in a larger collection or on the
%    World Wide Web is still an article and should be cited as such,
%    using the place of publication or xref fields to indicate the
%    manner of publication. Consider how silly it would be to cite all
%    kinds of material that might appear on a web page as type
%    \fn{url}: screenplay, map, video??
%
% \section{Citing a source}
%
%    The standard \cn{cite} command of \lat/ mixes single-cite and
%    multiple-cite functionality in a single command; although most of
%    the time this is convenient, it is something of a departure from
%    the \lat/ philosophy of logical markup, and gives rise to awkward
%    complications in certain situations. You can write, for example,
%    \cnm{cite}{xx,yy} to get [13, 15] and
%    \cnom{cite}[Theorem 4.9]{xx} to get [13, Theorem 4.9] but
%    there are difficulties in getting [13, Theorem 4.9; 15] or [13; 15,
%    Theorem 4.9].
%
%    The \pkg{amsrefs} package therefore provides two primary citing
%    commands: \cn{cite} for single cites and \cn{citelist} for multiple
%    cites. Some additional variants of those two commands are provided
%    for convenience.
% \begin{description}
% \item[] \cnm{cite}{xyz} works the same as ever, but for the sake of
%    consistency it is recommended that \cn{cite} be used only for
%    single cites and that multiple cites be handled with
%    \cn{citelist} (or \cn{cites}\mdash see below).
%    The square bracket notation \cnom{cite}[...]{...} is also
%    deprecated, because a superior alternative is provided:
%\begin{verbatim}
%\cite{xyz}*{Theorem 4.9}
%\end{verbatim}
%    Not only is this ordering of the arguments more natural, the
%    star\slash braces syntax prevents certain scanning problems that
%    afflict the bracketed form\mdash for example, if one wants
%    to put a cite command into the optional argument of a theorem
%    environment:
%\begin{verbatim}
%\begin{thm}[\cite[Theorem 4.9]{xyz}]
%\end{verbatim}
%    Although the square-bracket syntax continues to be supported for
%    compatibility's sake, users who stick to the
%    \cnmsm{cite}{...}*{...} form will never be confronted unexpectedly
%    with the puzzling \TeX{} error messages that result from scanning
%    problems as in this example.
%
% \item[] \cn{citelist} takes a list of \cn{cite} commands as its
%    argument; this makes it possible for any member of the list to be a
%    qualified cite, not just the last one. For example:
%\begin{verbatim}
%\citelist{\cite{xx} \cite{yy}*{Theorem 4.9} \cite{zz}}
%\end{verbatim}
%    Note that the members of the list \emph{are not separated by
%    commas}. The punctuation between individual cites is supplied
%    automatically (resist the temptation to add it by hand, since that
%    will seldom produce the intended results).
%
%    When the cites are numerical, multiple cites are sorted and ranges
%    of length three or more are compressed, as is done by the renowned
%    \pkg{cite} package, unless the \pkg{amsrefs} option
%    \opt{non-sorted-cites} is used. However, range compression is
%    turned off when the hyperref package is used, following its usual
%    practice.
%
% \item[] \cn{cites} is a variant of \cn{citelist}, provided for
%    convenience:
%\begin{verbatim}
%\cites{aa,bb,cc}
%\end{verbatim}
%    is equivalent to \verb'\citelist{\cite{aa}\cite{bb}\cite{cc}}'.
%    This can only be used when all the citations are simple ones with
%    no options.
%
% \item[] \cn{ycite} (year only), \cn{ocite} (object cite), and some
%    related commands are variants of \cn{cite} that provide the
%    distinctions needed in an author-year citing scheme. See the
%    discussion in Section~\ref{author-year} below.
% \end{description}
%
% \section{Author-year citation schemes}\label{author-year}
%
%    In author-year citation schemes three main citing forms are
%    required to cover the cases that in other schemes require only one
%    form. The first form is used when the citation serves as a
%    parenthetical annotation\mdash i.e., it could be omitted without
%    harming the grammatical structure of the sentence containing it.
%    For example:
% \begin{quote}
% The question first arose in systems theory (Rupp and Young, 1977).
% \end{quote}
%
%    The second form is like the first but is used when the author name
%    is already present as a natural part of the sentence, and the cite
%    therefore ought to supply only the year:
% \begin{quote}
% Rupp and Young (1977) have investigated \dots
% \end{quote}
%
%    The third form is used when the citation serves as a direct object
%    or other inomissible noun-like object within its sentence. For
%    instance, dropping the citation from the following sentence leaves
%    a grammatically incomplete remainder:
% \begin{quote}
% \dots\ for further details, see Rupp and Young (1977).
% \end{quote}
%
%    The logic of requiring three forms is perhaps most clearly seen if
%    we envision replacing the author-year citations by numerical
%    citations. The type of text replaced by [14] is different in each
%    case:
% \begin{quote}
% \dots\ arose in systems theory [14].\par
% Rupp and Young [14] have investigated \dots\par
% \dots\ for further details, see [14].
% \end{quote}
%
%    We delegate \cn{cite} to produce the primary parenthetical form
%    (Author, Year) and provide \cn{ycite} (\qq{year cite}) and
%    \cn{ocite} (\qq{object cite}) as the other forms. Plural forms
%    \cn{ycites} and \cn{ocites} are also provided in parallel with
%    \cn{cites}. And \cn{citeauthor} can be used to produce the list of
%    author names without the year.
%
%    These correspond with command names from other commonly used
%    author-year packages as follows:
% \begin{center}
% \begin{tabular}{l@{\qquad}l@{\qquad}l}
% \pkg{amsrefs}& \pkg{harvard}& \pkg{natbib}\\
% \nshline[3pt]
% \cn{cite}, \cn{cites}&    \cn{cite}&        \cn{citep}\\
% \cn{ycite}, \cn{ycites}&  \cn{citeyear}&    \cn{cite}\\
% \cn{ocite}, \cn{ocites}&  \cn{citeasnoun}&  \cn{citet}\\
% \cn{citeauthor}&          (none)&           \cn{citeauthor}\\
% \cn{citeauthory}&         (none)&           \cn{citet}
% \end{tabular}
% \end{center}
%
%    Some people like to use \cn{citet} or \cn{citeasnoun} when the
%    author name serves as the subject of a sentence; to my mind this
%    is a questionable blurring of the boundary between the essential
%    text of the sentence and the bibliography pointer. Those who don't
%    feel a need to bother about such distinctions can write
%    \cnm{citeauthory}{xyz} if they choose as an abbreviation for
%\begin{verbatim}
%\citeauthor{xyz} \ycite{xyz}
%\end{verbatim}
%
%    I.e., the command name is \cn{citeauthor}+\cn{ycite} except for
%    dropping the redundant second \texttt{cite}.\footnote{Why not
%    simply call it \cn{aycite}, you may ask? Well, I could hardly pass
%    up the opportunity to get all six vowels in a single command name,
%    could I? (If you really want a shorter command name you can always
%    provide it yourself with \cn{newcommand}.)}
%
%    When an author-year scheme is in use, parens are normally added by
%    \cn{cite}, \cn{citelist} and their variants unless the character
%    immediately following the command's argument is a closing paren.
%    This suffices for most cases; for special cases one can use the
%    \cn{parenthesize} command.
%
%    The ``full'' variants \cn{fullcite} and \cn{fullocite} print the
%    full list of author names instead of an abbreviated list, when an
%    (Author, Year) style of citation is in use. For other citation
%    schemes they produce the same output as the normal forms. Some
%    citation styles require the first cite to use the full list and
%    subsequent ones to use the abbreviated version. With proper setup
%    this can be done automatically, so that the ``full'' variants
%    should seldom be necessary in practice.
%
% \section{Section titles for bibliographies}
%
%    The \rpack/ provides three environments for producing the section or
%    chapter title of a bibliography:
%\begin{verbatim}
%bibchapter
%bibsection
%bibdiv
%\end{verbatim}
%    If you write
%\begin{verbatim}
%\begin{bibchapter}
% ...
%\end{bibchapter}
%\end{verbatim}
%    it is equivalent to writing
%\begin{verbatim}
%\chapter*{\bibname}
% ...
%\end{verbatim}
%    Similarly \env{bibsection} corresponds to
%    \verb'\section*{\refname}', whereas \env{bibdiv} produces one or
%    the other depending on context.
%
% \section{The \protect\env{biblist} environment}
%
%    The \pkg{amsrefs} package provides a \env{biblist} environment
%    which is used for the body of the bibliography. This makes it
%    easier to add introductory text after the section or chapter
%    heading but before the first bib-item. Instead of using
%    \env{thebibliography}, you would write, for example
%\begin{verbatim}
%\begin{bibchapter}[Annotated Bibliography]
%(Optional introductory text)
%\begin{biblist}
%\bib{...}{...}{
% ...
%}
%\bib{...}{...}{
% ...
%}
%...
%\end{biblist}
%\end{bibchapter}
%\end{verbatim}
%
%    The \env{biblist} environment is based on the standard \lat/
%    \env{list} environment and has an optional argument to allow changing
%    the list parameters. To change the amount of space allotted for
%    entry numbers to (e.g.) three digits, use the \cn{resetbiblist}
%    command:
%\begin{verbatim}
%\begin{biblist}[\resetbiblist{000}]
%\end{verbatim}
%
%    The name of the list counter is \fn{bib}. So you could (for
%    example) make the bibliography start at number 0 instead of 1 if
%    you wrote:
%\begin{verbatim}
%\begin{biblist}
%\setcounter{bib}{-1}
%\end{verbatim}
%
% \section{Line breaks in the bibliography}
%
%    When you need to force a line break between two fields, you can put
%    the \cn{linebreak} command at the end of the first field. Although
%    this would seem to invite, in most cases, a quite erroneous line
%    break preceding the punctuation that is automatically added, there
%    is special handling behind the scenes to move the line break
%    further along to the right place.
%\begin{verbatim}
%   title={Bayesian analysis ... econometrics\linebreak},
%\end{verbatim}
%    It should be fairly obvious that any change in the bibliography
%    style might invalidate existing forced breaks. ``Suggesting'' line
%    breaks with, e.g., \verb'\linebreak[3]' is usually a better idea
%    than forcing them.
%
% \section{Author names}
%
%    In general \BibTeX{} does quite a good job of dealing with names.
%    But there are a few problem cases. Consider the following examples
%    given in normal and inverted order.
% \begin{ctab}{r@{\ $\leftrightarrow$\ }l}
% John Q. Smith& Smith, John Q.\\
% John Q. Smith, Jr.& Smith, John Q., Jr.\\
% John Q. Smith III& Smith, John Q., III\\
% John Quincy Smith& Smith, John Quincy\\
% Ch. B. Chase& Chase, Ch. B.\\
% Rip van Winkle& van Winkle, Rip\\
% Thomas W. de la Ware& de la Ware, Thomas W.\\
% J. Marshall Ash& Ash, J. Marshall\\
% Jakob Ben Isaak& Ben Isaak, Jakob\\
% Anna Palamara Orsi& Palamara Orsi, Anna\\
% Yin Wuxiang& Yin Wuxiang\\
% Wuxiang Yin& Yin, Wuxiang
% \end{ctab}
%
% \section{Customizing bibliography style}\label{other-styles}
%
%    If you use the \rpack/ as is, the bibliography style you get is the
%    kind of style customarily seen in AMS publications. The recommended
%    way to get a different bibliography style is to write a \lat/
%    package which loads the \rpack/ with \cn{RequirePackage} and
%    then makes the desired changes by using suitable \cn{BibSpec}
%    commands as explained below. Thus, the general form of the custom
%    package will be
%\begin{verbatim}
%\ProvidesPackage{xyzbib}[2002/02/28 v1.23]
%\RequirePackage{amsrefs}\relax
%\BibSpec{article}{
%  ...
%}
%\BibSpec{book}{
%  ...
%}
%\end{verbatim}
%
%    The interior formatting within entries is specified by \cs{BibSpec}
%    commands, one for each entry type. To illustrate, let's look at an
%    example style spec for entries of type \fn{article}:
%    \begin{verbatim}
%    \BibSpec{article}{%
%      +{}{\PrintAuthors}  {author}
%      +{,}{ \textit}      {title}
%      +{,}{ }             {journal}
%      +{}{ \textbf}       {volume}
%      +{}{ \parenthesize} {date}
%      +{,}{ }             {pages}
%      +{,}{ }             {note}
%      +{.}{}              {transition}
%      +{}{ }              {review}
%    }
%    \end{verbatim}
%
%    It should be pretty obvious that each line specifies the formatting
%    for a particular field. After reading the data for a particular
%    \cn{bib} command, \lat/ steps through the style spec and for each
%    field listed, prints the field with the given formatting \emph{if
%    and only if the field has a nonempty value}. The \verb'+' character
%    at the beginning of each field spec must be followed by three
%    arguments: the punctuation to be added if the field is nonempty;
%    space and/or other material to be added after the punctuation; and
%    the field name. It is permissible for the second part to end with a
%    command that takes an argument, such as \cn{textbf}, in which case
%    it will receive the field's value as its argument. By defining a
%    suitable command and using it here you can place material after the
%    field contents as well as before; \cn{parenthesize} is an example
%    of this.
%
%    The reason that the punctuation and the following space are specified
%    separately is that between them there is a crucial boundary for line
%    breaks. If you put a \cn{linebreak} command at the end of a field value,
%    the break point will actually be carried onward to a suitable point after
%    the next bit of punctuation (whose actual value may vary depending on
%    which of the following fields is the first to turn up with a nonempty
%    value).
%
%    The meaning of the \cn{parenthesize} command, supplied by
%    \pkg{amsrefs}, should be obvious. The meaning of the \cs{PrintAuthors}
%    command is a different story. But I don't think it is all that hard to
%    understand. If we have two or three author names which were given separately,
%    and we need to combine them into a conventional name list using commas
%    and the word \qq{and}, then it would be nice if we had a command which
%    could take a list of names and Do The Right Thing. And that is just
%    what \cs{PrintAuthors} is.
%
%    The \pkg{rkeyval} package allows keys to be defined as additive: if the
%    key occurs more than once, each successive value will be concatenated to
%    the previous value, along with a prefix. The setup done by
%    \pkg{amsrefs} for the \fn{author} field is
%    \begin{verbatim}
%    \DefineAdditiveKey{bib}{author}{\name}
%    \end{verbatim}
%    This means that if two names are given, as in
%    \begin{verbatim}
%    author={Bertram, A.},
%    author={Wentworth, R.},
%    \end{verbatim}
%    then the final value of the \fn{author} field seen when \lat/
%    processes the style spec will be
%    \begin{verbatim}
%    \name{Bertram, A.}\name{Wentworth, R.}
%    \end{verbatim}
%
%    The \fn{transition} field in our BibSpec example is a dummy field to be
%    used when punctuation or other material must be added at a certain point
%    in the bibliography without regard to the emptiness or non-emptiness of
%    the fields after it. The \fn{transition} field always tests as non-empty
%    but has no printed content. So when you use it you always get the
%    indicated punctuation and space at the indicated point in the list of
%    fields. If it were the last thing in this BibSpec example, it could
%    serve just to put in the final period that is always wanted. But in AMS
%    bibliographies, if a Mathematical Reviews reference is given, it is
%    conventionally printed \emph{after} the final period. Using the
%    \fn{transition} field as shown here ensures that the final period will
%    be always printed, even when the \fn{review} field is empty.
%
%    \subsection{Miscellaneous commands provided by the \pkg{amsrefs}
%    package}\label{misc-cmds}
%
%    Most of the following commands are helper commands for use in
%    \cs{BibSpec} statements. The others are intended for use in bibliography
%    data.
%    \begin{description}
%
%    \item[\cn{parenthesize}] This command adds parentheses around its
%      argument. It is useful in \cs{BibSpec} statements because there is no
%      special provision for adding material after the field value.
%
%    \item[\cs{bibquotes}] This command is much like \cs{parenthesize} but
%      it adds quotes around its argument and it has one other important
%      difference: there are special arrangements to print the closing quote
%      \emph{after} a following comma or similar punctuation (unless the
%      \pkg{amsrefs} package is invoked with the \opt{logical-quotes} option,
%      in which case \cs{bibquotes} puts the closing quote immediately after
%      the quoted material).
%
%    \item[\cs{voltext}] The normal definition of this command is
%      \verb'vol.~' to supply the text that precedes a volume number.
%
%    \item[\cs{editiontext}] This command produces \verb'ed.' following an
%      edition number. See \cs{PrintEdition} for more information.
%
%    \item[\cs{DashPages}] This command is similar in spirit to
%    \cs{voltext} but more complicated in its implementation. It takes
%    one argument which is expected to contain one or more page numbers
%    or a range of page numbers. The argument is printed with a prefix
%    of \qq{p.} if it seems to be a single page number, otherwise with a
%    prefix of \qq{pp.}.
%
%    \item[\cn{tsup}, \cn{tsub}, \cn{tprime}] These are for text subscripts
%      and superscripts, with \cn{tprime} producing a superscript prime
%      symbol. Unlike the standard \cn{textsuperscript} and
%      \cn{textsubscript} functions provided by \lat/, these do not use
%      math mode at all.%
%    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%      \footnote{There is one drawback: If you don't want to get the prime
%        symbol for \cn{tprime} from the cmsy font, you will need to redefine
%        \cn{tprime} in some suitable way.}
%
%    \item[\cs{nopunct}] This command causes following punctuation to be
%      omitted if it is added with the internal function \cs{@addpunct}
%      (which is used throughout the \cs{BibSpec} handling when appending
%      a non-empty field).
%
%    \item[\cs{PrintAuthors}] This is a relatively complicated function that
%      tests a list of author names in order to decide whether
%      \cs{sameauthors} or \cs{AuthorList} should be called.
%
%    \item[\cs{AuthorList}] This takes a list of author names in the form
%    \begin{verbatim}
%    \name{Jones, Sam}\name{Smith, John}...
%    \end{verbatim}
%      and prints them in standard series form with the names uninverted,
%      e.g., \qq{Sam Jones and John Smith}, or \qq{Sam Jones, John Smith, and
%        James Taylor}.
%
%    \item[\cs{sameauthors}] This is a function of one argument. If you use
%      the default set of \cs{BibSpec}s from the \pkg{amsrefs},
%      \cs{sameauthors} is applied to the author name for a given \cn{bib}
%      command if it matches exactly the author name of the preceding
%      \cn{bib} command. Change the definition of \cs{sameauthors} if you
%      don't want to get a bysame dash.
%
%    \item[\cs{bysame}] This is a horizontal rule of length 3 em. The default
%      definition of \cs{sameauthors} prints \cs{bysame} instead of the
%      author names.
%
%    \item[\cs{PrintEditorsA}] This is similar to \cs{AuthorList} but adds
%      \verb'(ed.)' following the editor name (or \verb'(eds.)' if
%      applicable).
%
%    \item[\cs{PrintEditorsB}] This is like \cs{PrintEditorsB} but puts
%      parentheses around the entire list of editor names.
%
%    \item[\cs{Plural}, \cs{SingularPlural}] These are helper functions
%      that allow you to conditionally
%      print singular or plural forms such as \verb'(ed.)' or \verb'(eds.)'
%      depending on the number of names in the current name list.
%      The definition of \cs{PrintEditorsA} reads, in part,
%    \begin{verbatim}
%    ... (ed\Plural{s}.) ...
%    \end{verbatim}
%
%    \item[\cs{ReviewList}] This is similar to \cs{AuthorList} but is used for
%      printing (possibly multiple) MR numbers given in the \fn{review}
%      field.
%
%    \item[\cs{inicap}] This command applies initial capitalization to
%    its argument. The rules used are based on the ones given in the
%    Chicago Manual of Style (see \secref{capitalization}).
%
%    \item[\cs{EnglishInitialCaps}] This command will call \cs{inicap}
%    if and only if the language of the current reference is English.
%    For this purpose it is ``language'' in the Babel sense that is
%    significant. If English is the default language, you need to
%    specify
%    \begin{verbatim}
%    hyphenation={german},
%    \end{verbatim}
%      (for example) for non-English references to ensure that nothing will
%      be initial-capped.
%
%    \item[\cs{BibField}] This is for more complicated programming tasks such
%      as may be necessary for some BibSpecs.  It takes one argument, a field
%      name, and yields the contents of that field for the current \cn{bib}
%      entry.
%
%    \item[\cs{IfEmptyBibField}] If one writes
%    \begin{verbatim}
%    \IfEmptyBibField{isbn}{A}{B}
%    \end{verbatim}
%    then the commands in A will be executed if the \fn{isbn} field is empty,
%    otherwise the commands in B.
%
%    \item[\cs{PrintEdition}]
%    If a bibliography entry has
%    \begin{verbatim}
%      edition={2}
%    \end{verbatim}
%    and the bib-spec used \cs{PrintEdition} to handle this field, then the
%    edition information will be printed as \qq{2nd ed.}\mdash that is, the
%    number is converted to cardinal form and \qq{ed.} is added (taken from
%    \cs{editiontext}).
%
%    \item[\cs{CardinalNumeric}] This provides the conversion to cardinal
%      number form used by \cs{PrintEdition}.
%
%    \item[\cs{PrintDate}, \cs{PrintYear}] These functions convert a date in
%      canonical form (ISO 8601) to the form required by the current
%      bibliography style.  You can get your preferred date form by
%      redefining these functions or by changing your \cs{BibSpec} statements
%      to use another function of your own devising. The original definition
%      of \cs{PrintDate} adds parens (as for the year of a journal article
%      in normal AMS style), whereas the \cs{PrintYear} function simply
%      prints the year without any additional material (as for a book's year
%      of publication in normal AMS style).
%
%    \item[\cn{mdash}, \cn{ndash}] These are short forms for \cn{textemdash}
%      and \cn{textendash}, recommended instead of the more usual \verb'---'
%      and \verb'--' notation. From the \pkg{textcmds} package.
%
%    \item[\cn{cite}, \cn{cites}, \cn{citelist}, \cn{ycite}, \cn{ocite},]
%      etc. See the section on \textbf{Citations}.
%
%    \item[et cetera \dots] [mjd,2002-01-03] See the dtx files for
%    further possibilities that I have not managed to get properly
%    documented yet!
%    \end{description}
%
% \section{Capitalization of titles}\label{capitalization}
%
%    Here is a summary of the rules given in the \emph{Chicago Manual
%    of Style} \cite{CMS} for capitalizing titles of written
%    works in English:
% \begin{quote}
%    Capitalize each word, including pronouns and subordinate
%    conjunctions, except for articles, coordinate conjunctions, and
%    prepositions, or the word \emph{to} in infinitives. Always
%    capitalize the first and last word of the title and the first and
%    last word of any subtitles that it may contain. Don't capitalize
%    the second or later word in a hyphenated compound unless it is a
%    noun or proper adjective, or it has equal force with the first
%    word.
% \end{quote}
%    These are the rules followed by the \cn{EnglishInitialCaps}
%    command; see the documentation of the \pkg{inicap} package for more
%    details.
%
% \StopEventually{\PrintBibliography}
% \clearpage
% \typeout{@@@ That was the last page of the OnlyDescription}
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% \section{Implementation}
%
% \subsection{Overview}
%
%    It will be a while yet before we get to any actual code. First we
%    need to understand what the code needs to accomplish in order to
%    provide the user interface described above in a way that is as
%    compatible as possible with existing \lat/ mechanisms.
%
% \subsubsection{Normal \lat/ processing of cites}
%
% \paragraph{First \lat/ pass}
%
%    Various commands are written to the aux file that are mostly used by
%    \bibtex/.
% \begin{enumerate}
% \item A \cnm{cite}{moo} command writes one line to the aux file:
%    \cnm{citation}{moo}. This indicates to \bibtex/ that it should
%    include `moo' in the list of cited items to be searched for.
%    The \cn{cite} command then checks to see if \cs{b@moo} is defined:
%    If it is, its contents are printed as the cite `number'; if it
%    isn't, \lat/ gives an `Undefined citation' warning and prints
%    question marks instead.
%
% \item A \cnm{bibliographystyle}{har} command writes one line to the
%    aux file: \cnm{bibstyle}{har}. This indicates to \bibtex/ that it
%    should use \fn{har.bst} to determine the style for sorting and
%    formatting the bib items.
%
% \item A \cnm{bibliography}{hij,klm,...} command writes one line to the
%    aux file: \cnm{bibdata}{hij,klm,...}. This indicates to \bibtex/
%    that it should look in \fn{hij.bib}, \fn{klm.bib}, \dots\ for
%    bibliographic data. The \cn{bibliography} also tries to input
%    the \fn{.bbl} file, but on the first pass it won't exist
%    yet.
%
% \end{enumerate}
%
%    On the first pass all \cn{cite}'s normally are reported as
%    undefined because the \fn{.bbl} file has not yet been created.
%
% \paragraph{\bibtex/ pass}
%
%    For a document named \fn{xyz.tex}, the command \verb'bibtex xyz' is
%    used to invoke \bibtex/. It looks in \fn{xyz.aux} to find the
%    citation info written there by \lat/. For each \cn{citation} line,
%    \bibtex/ searches for a corresponding entry in the specified
%    \fn{.bib} files and formats it. The entire list is then sorted in
%    whatever way dictated by the bibliography style, and written out to
%    the file \fn{xyz.bbl}. This normally produces entries that look
%    something like:
%\begin{verbatim}
%\bibitem{BGL} P. Busch, M. Grabowski and P. J. Lahti:
%{\it Operational Quantum Physics.}
%Springer Verlag, New York (1995).
%\end{verbatim}
%
% \paragraph{Second \lat/ pass}
%
%    Now the \fn{.bbl} file exists and contains some \cn{bibitem}
%    commands. At \cnm{begin}{document}, \lat/ reads the aux file,
%    hoping to find some \cn{bibcite} commands, but it will not find
%    them until the next time around. \cn{citation}, \cn{bibstyle}, and
%    \cn{bibdata} commands in the aux file are simply ignored by \lat/.
%    Then \lat/ proceeds to typeset the body of the document.
% \begin{enumerate}
% \item Instances of \cn{cite} still print question marks.
%
% \item The \cn{bibliography} command causes \lat/ to input \fn{xyz.bbl}
%    and typeset its contents.
%
% \item A \cnm{bibitem}{moo} command writes one line to the
%    aux file: \cnmm{bibcite}{moo}{9}, where 9 is the current item
%    number.
%
% \item A \cnom{bibitem}[Moody]{moo} command writes one line to the
%    aux file: \cnmm{bibcite}{moo}{Moody}, using the supplied label
%    instead of a number.
%
% \end{enumerate}
%
% \paragraph{Third \lat/ pass}
%
%    Now the aux file contains some \cn{bibcite} commands. Once again,
%    \lat/ reads the aux file when it reaches \cnm{begin}{document}.
% \begin{enumerate}
%
% \item A \cnmm{bibcite}{moo}{Moody} causes \lat/ to define \cs{b@moo}
%    with `Moody' as the replacement text.
%
% \item If two \cn{bibcite} commands have the same citation key, \lat/
%    gives a warning message. This happens at \cnm{begin}{document},
%    during the reading of the aux file.
%
% \item Instances of \cn{cite} in the body of the document will print
%    the appropriate numbers obtained from the aux file.
%
% \item If there are any \cn{cite} commands for which the aux file did
%    not have a \cn{bibcite} command, \lat/ will give an `Undefined
%    citation' warning. This often happens if the aux file is incomplete
%    due to a \TeX{} error on the preceding pass.
% \end{enumerate}
%
% \subsection{How cites are processed by \pkg{amsrefs}}
%
%    In order to support its additional features (e.g., author-year
%    citations and the backref option), the \rpack/ stores additional
%    information for each cite in the macro \cs{b@whatever}. Instead of
%    simply using the defined or undefined status of this macro to
%    trigger the standard warnings, we add some boolean flags to allow
%    us to discriminate more finely what the current situation is.
% \begin{itemize}
% \item Each time an item is cited in the body of the document, a
%    backref entry is added to the info of that item. The backref info
%    is the current page and section location. Section location is a bit
%    hard to get right without better support from the documentclass.
%    So we provide a hook to allow it to work better when the support is
%    there.
%
% \item When a cite occurs, if the info is undefined then a warning is
%    issued and the info structure is created. A \cs{citation} command
%    and a \cs{citedest} command (providing backref info) are written to
%    the aux file. Because the backref info includes page number, it has
%    to be a non-immediate write.
%
%    An undefined info structure would normally happen only on a first
%    pass when no \fn{.aux} file exists, or when a new cite is added.
%    I.e., when the corresponding \cs{citation} command is not yet
%    present in the aux file.
%
% \item When a citation command occurs in the aux file, it initializes
%    the info structure if necessary, setting the ``bib-info-present''
%    flag to 0.
%
% \item When a \cs{citedest} command occurs in the aux file, it
%    initializes the info structure if necessary---but this shouldn't
%    happen: if the corresponding \cs{citation} command did not already
%    get processed, then something is wrong. So normally, the
%    \cs{citedest} command merely needs to add its backref info to the
%    existing info structure.
%
% \item When a \cs{bibcite} command occurs in the aux file, it will
%    normally find that \cs{b@whatever} is already defined, if the
%    bibliography occurs after all the \cs{cite} commands. What it must
%    do is fill in the appropriate blank slots in the info structure
%    set up by a previous \cs{citation} command.
%
% \item The aux file is actually processed two times, once at the
%    beginning of the document and once at the end. In the latter case,
%    \cs{bibcite} should give a warning if the backref-list is empty,
%    since that means there were no \cs{cite} commands for the given
%    key.
%
% \item When processing the bibliography: The \cs{bib} command needs to
%    check if it is using a key that is already used by another \cs{bib}
%    command.
% \end{itemize}
%
%    We therefore have
%\begin{verbatim}
%\b@xyz -> \citesel 00{number}{year}{backref-list}
%\end{verbatim}
%    where the first number is 1 if there has already been another
%    citation for the same key earlier in the document (some citation
%    styles use abbreviated forms for all instances after the first),
%    and the second number is 1 if the same key was already used by an
%    earlier \cs{bib} command.
%
%    Because the backref-list often includes page number information, it
%    cannot be built on the fly as we go along; instead we have to write
%    the information to the aux file and read it in at the beginning of
%    the next run.
%
%    If there was no \cs{bibcite} in the aux file for a given key, then
%    the info is
%\begin{verbatim}
%\b@xyz -> \citesel 00{}{}{backref-list}
%\end{verbatim}
%    If there was neither \cs{citation} nor \cs{bibcite} in the aux file
%    for a given key, then the \cs{cite} command should find that
%    \cs{b@xyz} is undefined.
%
%    If the author-year option is in effect, the \qq{number} contains
%    the author names instead of a number.
%\begin{verbatim}
%\b@xyz -> \citesel 00{
%            \name{Smith, J. K.}\name{Jones, T. J.}
%          }{...}{...}
%\end{verbatim}
%    Full name information is included in the data because some citation
%    styles give full names at the first citation and abbreviated forms
%    for subsequent instances.
%
% \subsection{Data structures}
%
%    The result of scanning the key/value pairs of a \cs{bib} command is
%    an assignment statement for \cs{rsk@toks}. (Cf the \pkg{rkeyval}
%    package.) For example, consider the entry
%\begin{verbatim}
%\bib{miller83}{article}{
%  author={Miller, G.},
%  title={Eine Bemerkung zur Darstellung von Polynomen \"{u}ber
%    Verb\"{a}nden},
%  journal={J. Math. Sent.},
%  volume={10},
%  year={1983},
%  pages={26\ndash 30},
%}
%\end{verbatim}
%    The scanned result is to assign
%\begin{verbatim}
%\global\rsk@toks{%
%  \@append\name\bib'author{Miller, G.}%
%  \DSK@def\bib'title{Eine Bemerkung zur Darstellung von Polynomen
%    \"{u}ber Verb\"{a}nden}%
%  \DSK@def\bib'journal{J. Math. Sent.}%
%  \DSK@def\bib'volume{10}%
%  \DSK@def\bib'year{1983}%
%  \DSK@def\bib'pages{26\ndash 30}%
%}
%\end{verbatim}
%    The code in the last arg of \cs{RestrictedSetKeys} then invokes
%    \cs{bib@exec} to do something with the value of \cs{rsk@toks}.
%\begin{verbatim}
%\bib@exec{miller83}{\the\rsk@toks}{\setbib@article}{}
%\end{verbatim}
%
%    Now the second-best thing to do with the value, if there weren't
%    any memory limits to worry about, might be to store it directly
%    into \cs{b@miller83}, reading the bibliography data at the
%    beginning of the document. Then cite commands would have access to
%    every field in the reference, and the bibliography could be printed
%    wherever desired by a suitable dumping command. Since this could
%    potentially use up a rather large chunk of main mem, depending on
%    the size of the bibliography, and since versions of \tex/ without
%    dynamic memory allocation are still in common use, and the fatal
%    error given by \tex/ when it runs out of memory is unlikely in this
%    case to provide much of a clue to the true source of the problem,
%    it seems dangerous to make this behavior the default.
%
%    The first, best alternative, however, might be to have cites and
%    bibliography filled in dynamically by the editing software during
%    the writing process, and leave to \latex/ only some final
%    consistency checks. The Emacs Ref\tex/ package is a good example of
%    this sort of thing.
%
% \subsubsection{Associating properties with field values}
%
%    The data example given above doesn't tell the whole story, because
%    it is in fact sometimes necessary to attach properties to a
%    particular value of a field. The premier example of this is
%    language. The title of a bibliography item, above all, may need to
%    have complicated bits of information supplied about its language:
%    not only whether it is an English or German or Spanish
%    title---which affects hyphenation and capitalization---but for a
%    Russian title (say) which input encoding does it use (since there
%    are several in common use), or whether it is an ASCII
%    transliteration (and if so, what kind), or whether it is a
%    translation into, say, English, of the original title. Arabic and
%    Chinese titles can be equally problematic.
%
%    It would be unnatural to plunk this information directly into the
%    contents of the field, but it needs to be tied to the field in some
%    way by our data structures. The simplistic approach would be to use
%    an auxiliary field ``extra-foo'' for each field foo. But this tends
%    to burn up a lot of \tex/ control sequence names, since the set of
%    field names is quasi-unbounded. So instead I have decided that if
%    a bibliography item has extra data for any of its fields, all such
%    extra data will be stored in another field named ``aux'', with
%    embedded tags to indicate which field each piece of the aux field
%    is associated with. The extra bits can be extracted on demand
%    using standard techniques, and the primary value of each field is
%    not burdened with any attachments, so that comparisons or scanning
%    of the field contents can remain as simple as possible.
%
%    Thus in practice there is at least one bit of auxiliary information
%    in every bib item, and our previous example would have the title
%    language indicated:
%\begin{verbatim}
%  \DSK@def\bib'title{Eine Bemerkung zur Darstellung von Polynomen
%    \"{u}ber Verb\"{a}nden}%
%  \@append\bib'title\bib'aux{\selectlanguage{german}}%
%\end{verbatim}
%
%    That provides a half-way decent \tex/ solution to the problem of
%    storing field properties. The problem of how the user should
%    specify field properties in the bib file is another story.
%
% \subsection{Preliminaries}
%
%    \begin{macrocode}
%<*pkg>
%    \end{macrocode}
%
%    Standard declaration of package name and date.
%    \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}[1995/12/01]
%    \end{macrocode}
%    Backward handling for beta version. There should be a better way of
%    doing this \dots
%    \begin{macrocode}
\@ifpackagewith{amsrefs}{beta}{%
  \expandafter\edef\csname opt@amsrbeta.sty\endcsname
    {\@ptionlist{amsrefs.sty}}%
  \def\@currname{amsrbeta}%
  \expandafter\let\csname amsrbeta.sty-h@@k\endcsname\@empty
  \def\@tempa{\input{amsrbeta.sty}\endinput}%
}{%
  \let\@tempa\@empty
}
\@tempa
%    \end{macrocode}
%
%    \begin{macrocode}
\@ifundefined{NormalCatcodes}{%
  \RequirePackage{pcatcode}\relax\PushCatcodes\NormalCatcodes
}{}
\ProvidesPackage{amsrefs}[2002/02/28 v1.23]
%    \end{macrocode}
%
%    \begin{macrocode}
%% WARNING WARNING WARNING: Catcode of apostrophe ' is letter
%% throughout this file.
\catcode`\'=11 % letter
%    \end{macrocode}
%
% \subsection{Some general tools}
%
%    Many of these useful functions are found also in AMS
%    documentclasses.
%
%    Some short forms to reduce code clutter.
%    \begin{macrocode}
\let\@xp=\expandafter
\let\@nx=\noexpand
%    \end{macrocode}
%
%    Some frequently used tests for empty arguments.
%    \begin{macrocode}
\long\def\@ifempty#1{\@xifempty#1@@..\@nil}
\long\def\@xifempty#1#2@#3#4#5\@nil{%
  \ifx#3#4\@xp\@firstoftwo\else\@xp\@secondoftwo\fi}
\long\def\@ifnotempty#1{\@ifempty{#1}{}}
%    \end{macrocode}
%
%    A really old version of \lat/ might still choke on \cs{newtoks} if
%    it is written directly in the false branch of a conditional.
%    \begin{macrocode}
\@ifundefined{@emptytoks}{\csname newtoks\endcsname\@emptytoks}{}
%    \end{macrocode}
%
%    Not in the \lat/ kernel yet.
%    \begin{macrocode}
\long\def\@gobblethree#1#2#3{}
\long\def\@nilgobble#1\@nil{}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\macrotext{\expandafter\strip@prefix\meaning}
\def\vdef#1#2{%
  \def#1{#2}\edef#1{\macrotext#1}%
}
%    \end{macrocode}
%
%    Test for a trailing option marked by a star. Usage:
%\begin{verbatim}
%\newcommand{\blub}[1]{\star@{\blubaux{#1}}{default}}
%\end{verbatim}
%    Arg 1 of \cs{star@} is the code to
%    be run, arg 2 is the default value of the option (could be empty).
%    If arg 1 is \cs{moo}, this test discards a star and expands to
%    \cs{moo} if a star is found, or expands to \verb'\moo{#2}' if
%    not. As the example shows, arg 1 need not be a single token.
%    \begin{macrocode}
\let\star@char=*
\def\star@#1#2{%
  \def\star@a##1{#1}\def\star@b{#1{#2}}%
  \futurelet\@let@token\star@test
}
\def\star@test{\ifx\@let@token\star@char \let\star@b\star@a\fi \star@b}
%    \end{macrocode}
%
%    Please note: If there is a space before the star, then the star is
%    not treated as an option char.
%
%    Why use a star? I dunno, it's one of the characters least likely to
%    suffer from catcode changes I guess, since it's already part of
%    standard \lat/ command syntax.
%
%    Why not just put the star at the beginning in the usual way? I
%    dunno, I guess it seemed to me that the lack of a trailing option
%    feature was a deficiency in current \lat/ and could be given an
%    experimental implementation in a package like this without any
%    adverse effect on existing documents.
%
%    \begin{macrocode}
\def\amsrefs@warning{\PackageWarning{amsrefs}}
%    \end{macrocode}
%
% \subsection{Package Options}
%
%    We call the \pkg{ifoption} package to facilitate some option tests.
%    \begin{macrocode}
\RequirePackage{ifoption}[2000/02/15]
%    \end{macrocode}
%
%    Options beginning with \qc{\*} require an additional \BibTeX{} run
%    in order to take proper effect. We use \qc{\*} as the marker
%    because it is one of the few characters that are pretty much
%    guaranteed to be free of catcode changes (because \cs{@ifstar}
%    relies on it).
%
%    The standard \fn{unsrt} biblio style prints the bibliography in
%    citation order, not sorted by author names.
%
%    \emph{Need to do an .ltb implementation for citation-order option,
%    and redo the \bibtex/ version too.} [mjd,2002-02-26]
%    \begin{macrocode}
\DeclareExclusiveOptions{sorted,citation-order}
%    \end{macrocode}
%    The \opt{alphabetic} option corresponds to the standard \fn{alpha}
%    biblio style with labels like Knu66 (three letters from name plus
%    two digits of year). Maybe should provide an alias LllYY for this
%    option. Numeric is the default since it is commoner in AMS
%    publications.
%    \begin{macrocode}
\DeclareExclusiveOptions{alphabetic,author-year,numeric}
%    \end{macrocode}
%
%    The standard \fn{abbrv} bibliography style uses abbreviations for
%    month names and journal names, and first names of people are
%    abbreviated to their initials. Since the second test bibliography
%    that I tested with had unabbreviated month names but abbreviated
%    journal names, perhaps it is a good idea to let these choices be
%    specified separately.
%    \begin{macrocode}
\DeclareBooleanOption{short-journals}
%    \end{macrocode}
%    \begin{macrocode}
\DeclareBooleanOption{short-months}
%    \end{macrocode}
%
%    \begin{macrocode}
\DeclareBooleanOption{initials}
%    \end{macrocode}
%
%    In the bibliography, if a title or something is enclosed in quotes,
%    should the closing quotes go inside the punctuation (logical
%    position) rather than outside (traditional)? These options give you
%    a choice.
%    \begin{macrocode}
\DeclareExclusiveOptions{traditional-quotes,logical-quotes}
%    \end{macrocode}
%
%    A sequence of cites will be sorted and ranges of length three or
%    greater will be compressed if these options so indicate.
%    \begin{macrocode}
\DeclareExclusiveOptions{sorted-cites,non-sorted-cites}
\DeclareExclusiveOptions{non-compressed-cites,compressed-cites}
%    \end{macrocode}
%
%    In the bibliography, print page numbers showing where each
%    entry was cited.
%    \begin{macrocode}
\DeclareBooleanOption{backrefs}
%    \end{macrocode}
%
%    Option for giving information about the available options:
%    \begin{macrocode}
\DeclareBooleanOption{?}
%    \end{macrocode}
%
%    This option means to forgo loading of the \pkg{textcmds} and
%    \pkg{mathscinet} packages.
%    \begin{macrocode}
\DeclareBooleanOption{lite}
%    \end{macrocode}
%
%    This option can be used by later releases as a sign that fall-back
%    adaptations need to be done.
%    \begin{macrocode}
\DeclareBooleanOption{beta}
%    \end{macrocode}
%
%    This one is obsolete now.
%    \begin{macrocode}
\DeclareBooleanOption{jpa}
%    \end{macrocode}
%
%    \begin{macrocode}
\ExecuteOptions{numeric,sorted,traditional-quotes,%
                sorted-cites,compressed-cites}
\ProcessOptions\relax
\ProcessExclusiveOptions
%    \end{macrocode}
%
%    Note that in the following auxiliary package list, ``getwidth'' is not
%    (yet) included.
%    \begin{macrocode}
\IfOption{?}{%
  \begingroup\catcode`\%=9 \catcode`\#=14 \catcode`\ =12
  \let\\\@backslashchar \catcode\endlinechar=12 \endlinechar=`\^^J
}{%#
  \typeout{Try \string\usepackage[?]{amsrefs} to get more information %#
           about amsrefs.}%#
}%#
%%\typeout{^^J#
%%The amsrefs package makes use of the following subordinate packages:
%%
%%  pcatcode      textcmds
%%  ifoption      mathscinet
%%  rkeyval
%%  inicap
%%
%%Package options for amsrefs are:
%%
%%  ?             Show this information
%%
%%  numeric       Numbered citations, standard AMS journal style
%%  alphabetic    Alphabetic citations, standard AMS journal style
%%  author-year   Author-year citation style
%%
%%  short-journals  "Duke Math. J." instead of "Duke Mathematical Journal"
%%  short-months    "Jan." instead of "January"
%%  initials        "Smith, J. Q." instead of "Smith, John Quincy"
%%
%%  traditional-quotes   \\bibquotes quotes text ``like this,''
%%  logical-quotes       \\bibquotes quotes text ``like this'',
%%
%%  sorted-cites      [9,1,16,4] --> [1,4,9,16] (numeric cites only)
%%  non-sorted-cites  No sorting of numeric cites
%%
%%  compressed-cites      [1,2,3,4] --> [1-4] (numeric cites only)
%%  non-compressed-cites  No range compression of numeric cites
%%
%%  backrefs      Backward references from bibliography entries
%%                to citations
%%
%%Defaults are: numeric, traditional-quotes, sorted-cites, and
%%compressed-cites.
%%}\aftergroup\@gobble\endgroup
%    \end{macrocode}
%
% \subsection{Auxiliary packages}
%    Now, if these other packages make use of the \pkg{pcatcode} package
%    like they should, then we don't need to make any fuss here about
%    the special catcode of \texttt{'}. Just load the packages.
%    \begin{macrocode}
\RequirePackage{rkeyval}[2001/12/22]
\RequirePackage{inicap}[2002/01/01]
\IfOption{lite}{% True? Then don't load the next two packages.
}{%              False? OK, let's load them:
  \RequirePackage{textcmds}[2001/12/14]
  \RequirePackage{mathscinet}[2002/01/01]
}
%    \end{macrocode}
%
% \subsection{The \env{biblist} environment}
%
%    The \env{biblist} environment can be used with a section or chapter
%    heading.
%
%  \begin{environment}{biblist}
%    Arg 1 was intended for key-value options, but support for that is
%    not done yet, instead it is just handled as additional list setup
%    code.
%    [mjd,2000/03/07]
%    \begin{macrocode}
\newenvironment{biblist}[1][]{%
  \normalfont \footnotesize \labelsep .5em\relax
  \list{\BibLabel}{%
    \resetbiblist{00}%
    \usecounter{bib}%
    #1\relax
  }%
  \sloppy
%    \end{macrocode}
%    Discourage page breaks within bibliography entries and disable them
%    completely for entries that are less than four lines long.
%    \begin{macrocode}
  \interlinepenalty\@m  \clubpenalty\@M \widowpenalty\clubpenalty
  \frenchspacing
  \ResetCapSFCodes
  \let\@bibdef\normal@bibdef
}{%
%    \end{macrocode}
%    Change error for empty list (no items) to warning, to allow authors
%    to leave their bibliography temporarily empty during writing:
%    \begin{macrocode}
  \def\@noitemerr{\@latex@warning{Empty bibliography list}}%
  \endlist
}
%    \end{macrocode}
%  \end{environment}
%
%    In case this is undefined sometimes.
%    \begin{macrocode}
\def\CurrentBib{??}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\BibLabel}{%
  [\hyper@anchorstart{cite.\CurrentBib}\relax\thebib\hyper@anchorend]%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\resetbiblist}[1]{%
  \settowidth\labelwidth{\def\thebib{#1}\BibLabel}%
  \leftmargin\labelwidth
  \ifdim\labelwidth=\z@
    \leftmargin=1em \itemindent=-\leftmargin
  \else
    \advance\leftmargin\labelsep
  \fi
}
%    \end{macrocode}
%
% \subsection{Printing a bibliography entry}
%
%    These two things (\cs{bib@start} and \cs{bib@end}) used to be
%    useful, but I haven't entirely made up my mind to get rid of them
%    yet. [mjd,2001-11-28]
%    \begin{macrocode}
\def\bib@start{\begingroup}
%    \end{macrocode}
%    Instead of being handled by \cs{bib@end}, ending punctuation is
%    normally handled via the `transition' field qv.
%    \begin{macrocode}
\def\bib@end{\relax
  \@xp\PrintBackRefs\@xp{\CurrentBib}%
  \par\endgroup
}
%    \end{macrocode}
%
%    For list macros such as \cn{name}, we want them to be unexpandable
%    except when special processing is done.
%    \begin{macrocode}
\let\name=?
\let\isbn=?
\let\review=?
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\partialbib}[1]{%
  \begingroup
  \RestrictedSetKeys{\let\rsk@nocomma\amsrefs@nocomma}{bib}%
    {\bib@partial{}{\the\rsk@toks}{}\endgroup}{#1}%
}
%    \end{macrocode}
%
%    And here are the predefined key names. You could always add some
%    more if you needed them. Only worry is about compatibility if you
%    want to share your data with other people.
%    \begin{macrocode}
\let\bib@reset\@empty % this gets modified by the following statements.
%    \end{macrocode}
%
%    First the fields that could be repeated more than once in a single
%    entry. Maybe publisher should be allowed to repeat also, for
%    co-published works. But then need to worry about the address
%    handling.
%    \begin{macrocode}
\DefineAdditiveKey{bib}{author}{\name}
\DefineAdditiveKey{bib}{editor}{\name}
\DefineAdditiveKey{bib}{isbn}{\isbn}
\DefineAdditiveKey{bib}{review}{\review}
\DefineAdditiveKey{bib}{part}{\partialbib}
%    \end{macrocode}
%
%    \begin{macrocode}
\DefineSimpleKey{bib}{title}
\DefineSimpleKey{bib}{subtitle}
\DefineSimpleKey{bib}{booktitle}
\DefineSimpleKey{bib}{series}
\DefineSimpleKey{bib}{conference}
\DefineSimpleKey{bib}{publisher}
\DefineSimpleKey{bib}{organization}
\DefineSimpleKey{bib}{institution}
\DefineSimpleKey{bib}{address}
\DefineSimpleKey{bib}{place}
\DefineSimpleKey{bib}{year}
\DefineSimpleKey{bib}{date}
\DefineSimpleKey{bib}{journal}
\DefineSimpleKey{bib}{fulljournal}
\DefineSimpleKey{bib}{issn}
\DefineSimpleKey{bib}{volume}
\DefineSimpleKey{bib}{number}
\DefineSimpleKey{bib}{pages}
\DefineSimpleKey{bib}{doi}
\DefineSimpleKey{bib}{status}
\DefineSimpleKey{bib}{eprint}
\DefineSimpleKey{bib}{preprint}
\DefineSimpleKey{bib}{language}
\DefineSimpleKey{bib}{hyphenation}
\DefineSimpleKey{bib}{note}
\DefineSimpleKey{bib}{url}
\DefineSimpleKey{bib}{xref}
\DefineSimpleKey{bib}{label}
\DefineSimpleKey{bib}{type}
\DefineSimpleKey{bib}{edition}
\DefineSimpleKey{bib}{setup}
\DefineSimpleKey{bib}{name}
%    \end{macrocode}
%
%    The \fn{transition} key is used when we want to insert punctuation
%    or other material at a given point in the sequence, even if all
%    following fields happen to be empty. The key appears
%    to have a non-empty value but its value is effectively empty.
%    \begin{macrocode}
\DefineDummyKey{bib}{transition}
%    \end{macrocode}
%
%    This provides access for BibSpec functions to individual fields.
%    \begin{macrocode}
\newcommand{\BibField}[1]{\csname bib'#1\endcsname}
\newcommand{\IfEmptyBibField}{\rkvIfEmpty{bib}}
%    \end{macrocode}
%
%    Here is the sequence of macro calls when \cn{bib} runs in a normal
%    way to print a bib item.
%\begin{verbatim}
%\bib{abk89}{proceedings.article}{...}
%  \bib ->\@ifstar and call \BibItem
%
%\\BibItem{abk89}{proceedings.article}{...}
%  \\BibItem #1#2-> read key, type and call \@bibdef
%#1<-abk89
%#2<-proceedings.article
%
%\@bibdef \setbib@proceedings.article {proceedings.article}{abk89}
%  \@bibdef #1#2#3-> warn if #1 is undefined; call RestrictedSetKeys
%#1<-\setbib@proceedings.article
%#2<-proceedings.article
%#3<-abk89
%
%\\RestrictedSetKeys {...}{bib}{\bib@exec ...}{...}
%  Read three args and then futurelet to start key scanning.
%  Third arg is code to call after scanning is finished.
%  \\RestrictedSetKeys [#1]#2#3-> ... (futurelet to start key scanning)
%#1<-\amsrefs@nocomma
%#2<-bib
%#3<-\bib@exec {abk89}{\the \rsk@toks }{\setbib@proceedings.article }\endgroup
%\end{verbatim}

%\begin{verbatim}
%\rsk@finish ->\bib@exec {abk89}{\the \rsk@toks }{...
%
%\bib@exec #1#2#3->
%#1<-abk89
%#2<-\the \rsk@toks
%#3<-\setbib@proceedings.article
%
% Here bib@exec = bib@print: Start up a group, process the assignments
%    found in rsk@toks, run \bib@cite{KEY}, then run \setbib@moo
%\end{verbatim}
%
%    Here is where the rubber hits the road.
%    \begin{macrocode}
\newcommand{\bib}{%
  \begingroup
  \@ifstar{\let\bib@exec\bib@store
           \let\@bibdef\normal@bibdef
           \BibItem}%
          {\BibItem}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
%% #1 key #2 type
\newcommand{\BibItem}[2]{%
  \def\@tempa{#1}%
  \edef\@tempb{%
    \@nx\@bibdef\@xp\@nx\csname setbib@#2\endcsname{#2}%
      {\macrotext\@tempa}%
  }%
  \@tempb
}
\let\biblio@list\@empty
%    \end{macrocode}
%    Use a standard \lat/ counter for numbering biblio items.
%    \begin{macrocode}
\newcounter{bib}
%    \end{macrocode}
%
%    The standard \cs{NoCommaWarning} is not entirely apt here, so we
%    make a slightly different version.
%    \begin{macrocode}
\def\amsrefs@nocomma{\amsrefs@warning{%
  Missing comma: some text might print incorrectly\MessageBreakNS}}
%    \end{macrocode}
%
%    \begin{macrocode}
%% #1 \setbib@article #2 article #3 key
\def\normal@bibdef#1#2#3{%
  \ifx\relax#1%
    \PackageError{amsrefs}{Undefined entry type: #2}\@ehc
    \let#1\setbib@misc
  \fi
  \RestrictedSetKeys{\let\rsk@nocomma\amsrefs@nocomma}{bib}%
    {\bib@exec{#3}{\the\rsk@toks}{#1}\endgroup}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\let\@bibdef\normal@bibdef
%    \end{macrocode}
%
%    This is a variation that ignores anything not having a known
%    citation key.
%    \begin{macrocode}
%% #1 \setbib@article #2 article #3 key
\def\selective@bibdef#1#2#3{%
  \@xp\selbibdef@a\csname b@#3\endcsname{#1}{#2}{#3}%
}
\def\selbibdef@a#1{%
  \def\@tempa{\endgroup\@gobblefour}%
  \ifx\relax#1\else \@xp\selbibdef@b#1\@nil \fi
  \@tempa
}
\def\selbibdef@b#1#2#3\@nil{%
  \ifx 1#2\let\@tempa\copy@bibdef\fi
}
\def\copy@bibdef#1#2#3{%
  \ifx\relax#1%
    \PackageError{amsrefs}{Undefined entry type: #2}\@ehc
    \let#1\setbib@misc
  \fi
  \@open@bbl@file
  \bbl@write{\string\bib{#3}{#2}\string{\iffalse}\fi}%
  % empty arg means use current value of \rsk@set instead of overriding it.
  \RestrictedSetKeys{\let\rsk@nocomma\amsrefs@nocomma
    \global\let\rsk@set\bbl@copy}{}%
    {\bbl@write{\iffalse{\fi\string}}%
     \endgroup}%
}
\def\bbl@copy#1\endcsname#2{%
  \toks@{#2}%
  \bbl@write{ \space#1={\the\toks@},}%
  \rsk@resume
}
%    \end{macrocode}
%
%    The function \cs{EnglishInitialCaps} needs to verify that the
%    language is English before applying capitalization.
%    \begin{macrocode}
\providecommand{\languagename}{english}
\def\biblanguageEnglish{english}
\let\biblanguagedefault\biblanguageEnglish
\let\bib@language\@empty
%    \end{macrocode}
%
%    \begin{macrocode}
\def\@ifsame@patterns#1#2{%
  \@xp\@ifsamepat\csname l@#1\@xp\endcsname\csname l@#2\endcsname
}
\def\@ifsamepat#1#2{%
  \ifnum
    \ifx\relax#1\m@ne\else#1\fi
    =\ifx\relax#2\m@ne\else#2\fi
    \@xp\@firstoftwo
  \else
    \@xp\@secondoftwo
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@language@fixup{%
  \ifx\bib'hyphenation\@empty
    \ifx\bib'language\@empty \let\bib@language\biblanguagedefault
    \else \let\bib@language\bib'language
    \fi
  \else
    \let\bib@language\bib'hyphenation
  \fi
  \def\@tempa##1 ##2\@nil{\lowercase{\def\bib@language{##1}}}%
  \@firstofone{\@xp\@tempa\bib@language} \@nil
}
%    \end{macrocode}
%
%    For \cn{bib} purposes we are interested mainly in testing whether
%    the hyphenation patterns are the same. So we use an
%    if-same-patterns test (by which \pkg{babel}'s `english' and
%    `american' compare as equal) rather than an if-same-language test.
%    Also, the way that the \cs{selectlanguage} command checks to see
%    whether a language has been properly defined for babel use is to
%    see if \cs{dateLANGUAGE} is defined. And if we tried to select an
%    undefined language, the result would be a \lat/ error.
%    \begin{macrocode}
\def\bib@selectlanguage{%
  \@ifsame@patterns{\languagename}{\bib@language}{}{%
    \@ifundefined{date\bib@language}{}{%
      \@xp\selectlanguage\@xp{\bib@language}%
    }%
  }%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@field@patches{%
  \let\bib@author\bib'author
  \let\bib@editor\bib'editor
  \let\bib@date\bib'date
  \ifx\bib@date\@empty \let\bib@date\bib'year \fi
  \bib@parsedate
  \ifx\bib'date\@empty \ifx\bib'year\@empty
    \let\bib@date\bib'status
    \let\bib@year\bib@date
  \fi\fi
%    \end{macrocode}
%    \cs{bib'language} is used for producing the printed rendition of
%    the language. \cs{bib@language} needs to be in the form required by
%    \cs{selectlanguage}.
%    \begin{macrocode}
  \bib@language@fixup
  \ifx\bib'place\@empty \let\bib'place\bib'address\fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@bibcite#1#2#3{%
  \begingroup
    \let\bib@elt\relax
    \xdef\biblio@list{\biblio@list\bib@elt{#1}}%
  \endgroup
  \if\numeric@refs
    \stepcounter{bib}%
    \protected@edef\@currentlabel{\@nx\@nx\@nx\cite@x{\thebib}{}}%
  \else
    \bib@reset
    \edef\bib@@data{#2}%
    \bib@@data % execute definitions locally
%    \bib@resolve@xrefs % ??
    \bib@field@patches
    \def\@currentlabel{\@nx\cite@x{\bib@year}{\bib@author}}%
    \let\name\relax
  \fi
  \protected@edef\@tempa{\@nx\bibcite{#1}{\@currentlabel}}%
  \@tempa
}
%    \end{macrocode}
%
%    \begin{macrocode}
\let\bib@@data\@empty
%    \end{macrocode}
%
%    \begin{macrocode}
\let\all@bibs\@empty
%    \end{macrocode}
%
%    \begin{macrocode}
% #1 key #2 \the\rsk@toks #3 \setbib@article
\def\bib@store#1#2#3{%
  \begingroup
  \bib@reset
  \edef\bib@@data{#2}%
  \bib@@data % execute definitions locally
%    \bib@resolve@xrefs % ??
  \bib@field@patches
  \def\@currentlabel{\@nx\cite@x{\bib@year}{\bib@author}}%
  \global\@xp\let\csname bi@#1\endcsname\bib@@data
  \@xp\g@addto@macro\@xp\all@bibs\csname bi@#1\endcsname
  \endgroup
}
%    \end{macrocode}
%
%    If one of the names in a name list is discovered to be an
%    abbreviation, resolve it. I guess this requires rebuilding the
%    list.
%    \begin{macrocode}
% \xref@check@a\bib'author
%  --> \name{Smith, J. K.}\name{me.w.e}
\def\xref@check@a#1{%
  \ifx\@empty#1\relax
  \else
    \begingroup
    \toks@\@emptytoks
    \let\name\xref@check@aa
    #1\relax
    \edef\@tempa{\endgroup \def\@nx#1{\the\toks@}}%
    \@tempa
  \fi
}
\def\BibAbbrevWarning#1{\amsrefs@warning{Abbreviation '#1' undefined}}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\xref@check@aa#1{\def\@tempa{#1}\lowercase{\def\@tempb{#1}}%
  \ifx\@tempa\@tempb
    \ifx\@tempa\@empty
      \toks@\@xp{\the\toks@ \name{}}%
    \else
      \@ifundefined{bi@#1}{%
        \BibAbbrevWarning{#1}%
%    \end{macrocode}
%    I think the extra space and unskip helps to avoid a certain kind of
%    error. [mjd,2001-12-27]
%    \begin{macrocode}
        \toks@\@xp{\the\toks@ \name{#1 \unskip}}%
      }{%
        \xref@check@ab{#1}%
      }%
    \fi
  \else
    \toks@\@xp{\the\toks@ \name{#1}}%
  \fi
}
\def\xref@check@ab#1{%
  \let\bib'name\@empty
  \csname bi@#1\endcsname
  \ifx\@empty\bib'name
    \@temptokena{#1}%
  \else
    \@temptokena\@xp{\bib'name}%
  \fi
  \edef\@tempa{\toks@{\the\toks@\@nx\name{\the\@temptokena}}}%
  \@tempa
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\xref@check@b#1{%
  \ifx\@empty#1%
  \else
    \toks@\@xp{#1}%
    \edef\@tempb{\lowercase{\def\@nx\@tempa{\the\toks@}}}%
    \@tempb
    \ifx\@tempa#1\relax % all lowercase
      \let#1\@empty
      \csname bi@\the\toks@\endcsname
    \fi
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\XRefWarning#1{\amsrefs@warning{Xref '#1' undefined}}
%    \end{macrocode}
%
%    If any title occurs in an xref-d item, assume that it is a book
%    title. This might not always be the best assumption? Let's see how
%    it goes though. [mjd,2001-12-11]
%    \begin{macrocode}
\def\xref@add@toks#1#2{%
  \ifx#1\@empty
    \toks@\@xp{\the\toks@\DSK@def#1{#2}}%%
  \else
    \def\@tempa##1\bib'title##2##3\@nil{##2}%
    \if T\@tempa #1T\bib'title F\@nil
      \toks@\@xp{\the\toks@\DSK@def\bib'booktitle{#2}}%
    \fi
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\xref@check@c#1{%
  \begingroup
  \let\DSK@def\xref@add@toks
  \toks@\@emptytoks
  \@for\xref@ID:=#1\do{%%
    \@ifundefined{bi@\xref@ID}{%
      \XRefWarning{\xref@ID}%
    }{%
      \csname bi@\xref@ID\endcsname
    }%
  }%
  \edef\@tempa{\endgroup\the\toks@}%
  \@tempa
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@resolve@xrefs{%%
  \ifx\bib'xref\@empty \else \xref@check@c\bib'xref \fi
  \xref@check@a\bib'author \xref@check@a\bib'editor
  \xref@check@b\bib'journal \xref@check@b\bib'publisher
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@print#1#2#3{%
%% #1 citation key, #2 \the\rsk@toks, #3 \setbib@moo
  \bib@start
  \bib@reset
  \let\setbib@@#3%
  \edef\bib@@data{#2}%
  \bib@@data % execute definitions locally
  \bib@resolve@xrefs % modifies \bib@@data
  \bib@field@patches
  \bib@selectlanguage
  \bib'setup
  \bib@cite{#1}\kern\@ne sp\relax
  \ifx\setbib@@\setbib@article
    \ifx\bib'booktitle\@empty
    \else \let\setbib@@\setbib@incollection
    \fi
  \fi
  \setbib@@
  \bib@end
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@partial#1#2#3{%
%% #1 citation key, #2 \the\rsk@toks, #3 \setbib@moo
  \begingroup
  \bib@reset \let\bib'title\@empty \let\bib'author\@empty
  \edef\bib@@data{#2}%
  \bib@@data % execute definitions locally
  \bib@field@patches
  \bib'setup
  \setbib@@
  \endgroup
}
%    \end{macrocode}
%
%    This is just an alias for convenience.
%    \begin{macrocode}
\let\bib@exec\bib@bibcite
\AtBeginDocument{\let\bib@exec\bib@print}
%    \end{macrocode}
%
%    When \cs{bib@cite} is called, author name and year are available in
%    \cs{bib@author} and \cs{bib@year}.
%    \begin{macrocode}
\IfOption{author-year}{%
  \def\cite@label{%
    \ifx\bib@author\@empty \bib@editor \else \bib@author \fi
  }%
}{%
  \def\cite@label{\@currentlabel}%
}

\IfOption{alphabetic}{%
  \def\alpha@label{%
    \ifx\@empty\bib'label
      \def\thebib{\CurrentBib}%
    \else
      \let\thebib\bib'label
    \fi
  }%
}{%
  \let\alpha@label\relax
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@cite#1{%
  \alpha@label % modify \thebib if necessary
  \def\CurrentBib{#1}%
  \item\leavevmode
  \@xp\bib@cite@a\csname b@#1\endcsname
  \bibcite@write{#1}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@cite@a#1{%
  \ifx\relax#1%
    \protected@edef\@tempa{%
      \gdef\@nx#1{\@nx\citesel 01{\cite@label}{\bib@year}{}}}%
    \@tempa
  \else
    \@xp\bib@cite@check\@xp#1#1%%
    \@empty\@empty\@empty\@empty\@empty
  \fi
}
%    \end{macrocode}
%
%    For the citation key we want to check if it is already defined. But
%    there is a slight problem. There is already one control
%    sequence in use for each bibliography entry, to store the number or
%    the author/year information needed by \cs{cite}. If we introduce
%    another control sequence to check whether a particular cite is
%    multiply defined, then we double the number of control sequences
%    used. For a large bibliography in a book this is fairly serious.
%    This is addressed by using a \cs{citesel} function.
%
%    \begin{macrocode}
%  #1        #2        #3#4 #5     #6    #7
% \b@foo -> \citesel ??{number}{year}{backrefs}
\def\bib@cite@check#1#2#3#4#5#6#7{%
  \ifx 1#4\relax
    \DuplicateBibKeyWarning
  \else
    \protected@edef\@tempa{%
      \gdef\@nx#1{\@nx\citesel #31{\cite@label}%
        {\bib@year\ifx\citesel\citesel@authoryear \bib'label\fi}%
        {#7}}}%
    \@tempa
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\DuplicateBibKeyWarning{%
  \amsrefs@warning{%
    Duplicate \protect\bib\space key%
    `\CurrentBib ' detected\MessageBreakNS}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bibcite@write#1{%
  \if@filesw
    \let\citesel\citesel@write
    \csname b@#1\endcsname
  \fi
}
\def\citesel@write#1#2#3#4#5{%
  \begingroup
  \toks@{{#3}{#4}}%
  \immediate\write\@auxout{\string\amsRefs}%
  \immediate\write\@auxout{\string\bibcite{\CurrentBib}{\the\toks@}}%
  \endgroup
}
%    \end{macrocode}
%
%    Because duplicate bibs are caught immediately, we don't need
%    \cs{bibcite} to run \cs{@testdef}.
%    \begin{macrocode}
\AtEndDocument{\let\amsRefs@bibcite\@gobbletwo}
%    \end{macrocode}
%
% \subsection{Some abbreviations}
%
%    The commands \cs{DefineName}, \cs{DefinePublisher}, and \cs{DefineJournal}
%    are provided to make abbreviations a little easier.
%
%    \begin{macrocode}
\newcommand{\DefineName}[2]{%
  \begingroup
  \let\bib@exec\bib@store \BibItem{#1}{person}{name={#2},}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\DefineJournal}[2]{%
  \begingroup
  \let\bib@exec\bib@store \BibItem{#1}{periodical}{journal={#2},}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\DefinePublisher}[3]{%
  \begingroup
  \let\bib@exec\bib@store
  \BibItem{#1}{publisher}{%
    publisher={#2},%
    place={#3},
  }%
}
%    \end{macrocode}
%
%    This is what the standard \fn{book.cls} has for the bibliography
%    title:
%\begin{verbatim}
%\newenvironment{thebibliography}[1]
%     {\chapter*{\bibname
%        \@mkboth{\MakeUppercase\bibname}{\MakeUppercase\bibname}}%
%      \list{\@biblabel{\@arabic\c@enumiv}}%
%\end{verbatim}
%
%    \begin{macrocode}
\newcommand{\printbibliography}[1][\bibname]{%
  \begin{bibdiv}[#1]%
  \begin{biblist}%
  \let\bib@elt\bib@print
  \biblio@list
  \end{biblist}%
  \end{bibdiv}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\renewenvironment{thebibliography}[1]{%
  \bibdiv
  \biblist[\resetbiblist{#1}]%
}{%
  \endbiblist
  \endbibdiv
}
%    \end{macrocode}
%
%    \begin{macrocode}
\providecommand{\bibname}{Bibliography}
\providecommand{\refname}{References}
%    \end{macrocode}
%    We need to take a little extra trouble here to pre-expand the
%    \cs{bibname}.
%    \begin{macrocode}
\newenvironment{bibchapter}[1][\bibname]{%
  \begingroup
  \LoadBackRefs
  \protected@edef\@{\endgroup\protect\chapter*{#1}}%
  \@
}{\par}
\newenvironment{bibsection}[1][\refname]{
  \begingroup
  \LoadBackRefs
  \protected@edef\@{\endgroup\protect\section*{#1}}%
  \@
}{\par}%
\@ifundefined{chapter}{%
  \newenvironment{bibdiv}{\bibsection}{\endbibsection}
}{%
  \newenvironment{bibdiv}{\bibchapter}{\endbibchapter}
}
%    \end{macrocode}
%
%    An extremely crude way of carrying forward the current font series
%    (basically, bold or non-bold) for the next punctuation object. This
%    needs work I guess. [mjd,2001-12-11] As things currently stand,
%    when you specify a font in the punctuation argument of a bibspec
%    line, it interferes with the stuff in \cs{SwapBreak} that prevents
%    redundant punctuation.
%    \begin{macrocode}
\let\bib@lastseries\@empty
\def\CarryFontSeries{\global\let\bib@lastseries\f@series}
\def\SameFontSeries#1{%
  \begingroup
  \ifx\bib@lastseries\@empty
  \else \fontseries\bib@lastseries\selectfont
    \global\let\bib@lastseries\@empty
  \fi
  #1\endgroup
}
%    \end{macrocode}
%
%    The arg has to start with a simple punctuation character in order
%    for this to work as intended.
%    \begin{macrocode}
\def\SwapBreak#1{%
  \@tempcnta\@MM
  \ifhmode
    \unskip
    \ifnum\lastpenalty=\z@
    \else \@tempcnta\lastpenalty \unpenalty
    \fi
  \fi
%    \end{macrocode}
%    Omit arg 1 (punctuation) if there is a special kern or spacefactor
%    to so indicate.
%    \begin{macrocode}
  \toks@{\SameFontSeries{#1}}%
  \edef\@tempa{%
    \@nx\deferredquoteslogical
    \ifnum\lastkern=\@ne
    \else
%    \end{macrocode}
%    If the preceding text ended with an abbreviation, we don't want to
%    redundantly add another period. Check the spacefactor.
%    \begin{macrocode}
      \ifnum\spacefactor=\sfcode\@xp\@xp\@xp`\@xp\@car\string#1)\@nil
      \else \the\toks@
      \fi
    \fi
    \@nx\deferredquotes
    \ifnum\@tempcnta=\@MM \else \penalty\number\@tempcnta\space\fi
    \ifnum\lastkern=\@ne \ignorespaces \fi
  }%
  \@tempa
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcount\series@index
%    \end{macrocode}
%
%    The \cn{PrintSeries} command prints a list of objects in series
%    form. The essential idea is to produce something like \qq{A, B, and
%    C} when we are given three elements A B C, with suitable variations
%    in the punctuation and other intervening material if the number of
%    elements is less or more than three.
%
%    To generalize this process, we envision \cn{PrintSeries} being
%    called as
%\begin{verbatim}
%\PrintSeries{i0}{i1}{i2}{i3}{i4}{\do{A}\do{B}...}
%\end{verbatim}
%    where i0, \dots, i4 are material to be interpolated and the last
%    arg is a list of indeterminate length where each element consists
%    of a macro and its argument. The output, depending on the number of
%    elements, will be
%\begin{verbatim}
% i0 A i4                 % 1 element
% i0 A i1 B i4            % 2 elements
% i0 A i2 B i3 C i4       % 3 elements
% i0 A i2 B i2 C i2 ... X i2 Y i3 Z i4   % 26 elements
%\end{verbatim}
%    That is the simple explanation but in practice there are some
%    additional complications. What if user-supplied line breaks have to
%    be supported at the boundaries between elements? What if in
%    addition to adding material between elements we also want to apply
%    some handy function to each element (e.g., \cn{textsc})? Even
%    worse, what if we want the function to be different depending on
%    the position of the element in the list?? Indeed if this did not
%    happen to be the case with the current application I would not have
%    gone to the extra trouble of supportint it. But if it must be so,
%    then the output that we need from a list \verb'\do{A}\do{B}...' is
%\begin{verbatim}
% f0{A}
% f0{A} p1 i1 f1{B}
% f0{A} p2 i2 f2{B} p3 i3 f3{B}
%\end{verbatim}
%    and so on, where
% \begin{itemize}
% \item $f_n$ is a macro taking one argument,
% \item $p_n$ is punctuation\mdash material that must precede a line
%    break if one occurs at this boundary,
% \item $i_n$ other interpolated material, as before.
% \end{itemize}
%    To reduce the number of distinct required objects we decree that
%    each element will get braces wrapped around it as a matter of
%    course; then it is possible for f1, f2, f3 to be assimilated into
%    the tail end of i1, i2, i3.
%
%    \begin{macrocode}
\def\PrintSeries#1#2#3#4#5#6#7#8{%
  \begingroup
  \def\series@add@a{#2}%
  \def\series@add@b{\SwapBreak{#3}#4}%
  \def\series@add@c{\SwapBreak{#5}#6}%
  \def\series@add@d{\SwapBreak{#7}#8}%
  \PrintSeries@a{#1}%
}
%    \end{macrocode}
%    \cn{Plural} takes one argument and prints it if there were two or
%    more elements in the current list. So, to get \qq{editors} instead
%    of \qq{editor} after printing a list of editor names, write
%    \verb'editor\Plural{s}'.
%
%    \cn{SingularPlural} takes two arguments and prints the first if
%    there was only one element, otherwise prints the second arg.
%    \begin{macrocode}
\newcommand{\SingularPlural}[2]{#1}
\newcommand{\Plural}{\SingularPlural{}}
%    \end{macrocode}
%    For \cn{PrintSeries@a} the first arg is the iterator function
%    present in the list which is arg 3. Args 2 and 4 are extra material
%    to be added before and after the list that may require the use of
%    \cn{Plural} or \cn{SingularPlural}.
%    \begin{macrocode}
\def\PrintSeries@a#1#2#3#4{%
  \series@index\z@ \def#1{\advance\series@index\@ne \@gobble}%
  #3\relax
  \ifnum\series@index=\@ne \let\SingularPlural\@firstoftwo
  \else \let\SingularPlural\@secondoftwo
  \fi
  \chardef\series@total=\series@index \series@index=\z@
  \let#1\series@add
  #2#3#4\relax
  \endgroup
}
%    \end{macrocode}
%
%    This is the inner function called by \cn{PrintSeries} that
%    carefully distributes all the material stored previously in
%    \cn{series@add@...} macros.
%    \begin{macrocode}
\def\series@add#1{%
  \advance\series@index\@ne
  \let\@tempa\relax
  \ifcase\series@index
  \or % material before name 1
    \let\@tempa\series@add@a
  \or % material before name 2
    \ifnum\series@total<\thr@@ \let\@tempa\series@add@b
    \else \let\@tempa\series@add@c
    \fi
  \else % material before name 3, 4, 5, ...
    \ifnum\series@index<\series@total \let\@tempa\series@add@c
    \else \let\@tempa\series@add@d
    \fi
  \fi
  \@tempa{#1}%
}
%    \end{macrocode}
%
%    The function \cn{bib@append} prints the value of a field, together
%    with associated punctuation and font changes, unless the value is
%    empty. Arg 1 is punctuation (that may need to be swapped with a
%    preceding line break), arg 2 gives the space to be added after the
%    punctuation, and possibly a function to be applied to the contents
%    of arg 3, which is
%    a macro containing the field
%    value. So if we have \cn{moo} and \cn{bib'pages}, from
%    \verb'pages={21\ndash 44}', then we want to arrange to call
%\begin{verbatim}
%\moo{21\ndash 44}
%\end{verbatim}
%    We don't want to simply call \cs{moo}\cs{bib'bar} because that makes
%    it rather difficult for \cs{moo} to look at the contents of
%    \cn{bib@bar}.
%    \begin{macrocode}
\def\bib@append#1#2#3{%
  \ifx\@empty#3\relax
  \else
    \ifx\relax#3\errmessage{#3=\relax}\fi
    \@ifempty{#1}{%
      \@temptokena{\ifnum\lastkern=\@ne\ignorespaces\fi #2}%
    }{%
      \@temptokena{\SwapBreak{#1}#2}%
    }%
    \toks@\@xp{#3}%
    \edef\@tempa{\the\@temptokena{\the\toks@}}%
    \@xp\@tempa
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\let\endbracket\@empty
\let\bracket@stack\@empty
\def\push@bracket#1{%
  \xdef\bracket@stack{#1\bracket@stack}%
}
\def\pop@bracket{\iffalse{\fi
  \@xp\pop@bracket@a\bracket@stack \@empty}}
\def\pop@bracket@a#1{\leavevmode\/\upn{#1}\xdef\bracket@stack{\iffalse}\fi}
%    \end{macrocode}
%
%    The \cn{parenthesize} function adds parens around its argument,
%    calling \cn{upn} to prevent italic parens from being used. But the
%    default definition of \cn{upn} is a no-op, meaning that this
%    refinement lies dormant unless the \pkg{upref} package or other
%    activation is done. (Probably better done via special fonts,
%    anyway.)
%    \begin{macrocode}
\providecommand{\upn}[1]{#1}
\newcommand{\parenthesize}[1]{%
  \leavevmode\push@bracket)\upn{(}#1\pop@bracket
}
%    \end{macrocode}
%
%    \begin{macrocode}
\let\deferredquotes\@empty
\IfOption{logical-quotes}{%
  \def\deferredquoteslogical{\deferredquotes}%
}{%
  \let\deferredquoteslogical\relax
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\bibquotes}[1]{%
  \textquotedblleft#1%
  \gdef\deferredquotes{%
    \global\let\deferredquotes\@empty \textquotedblright
  }%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\providecommand{\mdash}{\textemdash}
\providecommand{\ndash}{\textendash}
%    \end{macrocode}
%
%    For older versions of some AMS documentclasses, this patch is
%    needed.
%    \begin{macrocode}
\providecommand{\MRhref}[1]{}
%    \end{macrocode}
%
%    The \cn{@addpunct} function is defined by AMS documentclasses and
%    the \pkg{amsgen} package. But if we find it undefined we had better
%    define it.
%    \begin{macrocode}
\@ifundefined{@addpunct}{%
  \def\@addpunct#1{%
    \relax\ifhmode\unskip\ifnum\spacefactor>\@m \else#1\fi\fi
  }
  \def\frenchspacing{\sfcode`\.1007\sfcode`\?1006\sfcode`\!1005%
    \sfcode`\:1003\sfcode`\;1002\sfcode`\,1001\relax}
}{}
\providecommand{\nopunct}{\spacefactor 1004\relax}
\providecommand{\ResetCapSFCodes}{%
  \count@=`\A
  \def\@tempa{\sfcode\count@=\@m
    \advance\count@\@ne
    \ifnum\count@>`\Z\relax\expandafter\@gobble \fi
    \@tempa
  }%
  \@tempa
}
%    \end{macrocode}
%
%    The \cn{bibspec@scan} function scans one field spec
%    from the second arg of \cn{BibSpec}. Each field spec has the form
%\begin{verbatim}
%+{punctuation}{prelim material}{field name}
%\end{verbatim}
%    We provide a certain amount of error recovery by reading an extra
%    argument and checking if it is a plus sign as expected. If the plus
%    sign is missing for the next field, we skip arg 4 (printing the
%    tail of it if it is multiple tokens), give an error message, and
%    strive to continue anyway. This might produce hilarious
%    nonsense\mdash but that could be, after all, not such a bad thing
%    once in a while.
%    \begin{macrocode}
\def\bibspec@scan#1#2#3#4{%
  \toks@\@xp{\the\toks@\bib@append{#1}{#2}}%
  \edef\@tempa{%
    \toks@{\the\toks@ \@xp\@nx\csname bib'#3\endcsname}%
  }%
  \@tempa
  \ifx\@empty#4\@xp\@gobble % end the recursion
  \else \ifx +#4\else\bibspec@scan@error\fi
  \fi
  \bibspec@scan
}
%    \end{macrocode}
%
%    \begin{macrocode}
%    \end{macrocode}
%
%    Accumulate spec material in \cs{toks@}, then define \cs{setbib@moo}
%    from it.
%    \begin{macrocode}
\newcommand{\BibSpec}[2]{%
  \toks@\@emptytoks
  \@ifnotempty{#2}{%
    \@ifnextchar{+}{\@xp\bibspec@scan\@gobble}{\bibspec@scan}#2%
      \@empty\@empty\@empty\@empty
  }%
  \@xp\edef\csname setbib@#1\endcsname{\the\toks@}%
}
%    \end{macrocode}
%    I made this a \cs{def} rather than a \cs{let} because using
%    \cs{let} would make \cs{BibSpecAlias} statements order-sensitive in
%    a way that seems frequently to be a stumbling block to unwary
%    package writers. But then we should probably do at least the
%    simplest kind of infinite loop check.
%    \begin{macrocode}
\newcommand{\BibSpecAlias}[2]{%
  \@xp\def\@xp\@tempa\@xp{\csname setbib@#1\@xp\endcsname}%
  \@xp\ifx\csname setbib@#2\endcsname\@tempa
    \PackageError{amsrefs}{%
      Mirror alias #1->#2 not allowed (infinite loop)}\@ehc
  \else
    \@xp\def\csname setbib@#1\@xp\endcsname
      \@xp{\csname setbib@#2\endcsname}%
  \fi
}
%    \end{macrocode}
%
% \subsection{Name handling}\label{name-style}
%
%    When dealing with author or editor names in a \cs{BibSpec}, it is
%    not easy to design some way of specifying the desired style that is
%    flexible enough to handle the multitude of different styles needed
%    in practice. As a first level of indirection, we expect that the
%    \cs{BibSpec} will typically start by applying a function such as
%    \cs{PrintAuthors} to the list of names. At the point of
%    application, the name list will have the form
%\begin{verbatim}
%\name{firstauthor}\name{secondauthor}...
%\end{verbatim}
%    Because the formatting may well depend on the number of names that
%    are present, we first run a \cs{NameSetup} macro that looks at the
%    number of names and prepares a list of functions to be applied to
%    each name in the list. This method allows us to specify the
%    internal formatting of an author name with the same kind of
%    specifications as are used in a general \cs{BibSpec}.
%    \begin{macrocode}
\newcommand{\PrintNames}[4]{%
  \begingroup \NameSetup{#1}{#4}%
  #2#4#3\relax
  \edef\@tempa{\seriesdefault}\edef\@tempb{\f@series}%
  \ifx\@tempa\@tempb \else \CarryFontSeries \fi
  \endgroup
}
%    \end{macrocode}
%
%    \begin{macrocode}
\providecommand{\bysame}{\leavevmode\hbox to3em{\hrulefill}\thinspace}
\newcommand{\sameauthors}[1]{\bysame}
\def\PreviousAuthors{\relax}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\AuthorList}[1]{\PrintNames{author}{}{}{#1}}%
\newcommand{\PrintAuthors}[1]{%
  \ifx\PreviousAuthors\bib@author
    \begingroup \aftergroup\sameauthors \let\AuthorList\endgroup
  \else
    \global\let\PreviousAuthors\bib@author
  \fi
  \AuthorList{#1}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bib@name@init{%
  \let\bib'surname\@empty \let\bib'given\@empty
  \let\bib'jr\@empty \let\bib'initials\@empty
}
%    \end{macrocode}
%
%    When specifying the format of author names, we distinguish four
%    cases: first author, co-author, middle author, and last author.
%    By breaking down the specifications to this level, we make it
%    easier to handle special requirements such as inverting the first
%    author's name but printing the remaining names in normal order.
%    \begin{macrocode}
\IfOption{initials}{% TRUE:
  \BibSpec{firstauthor}{
   +{}{}{initials}
   +{}{ }{surname}
   +{}{ }{jr}
  }
  \BibSpec{coauthor}{
   +{}{ and}{transition}
   +{}{ }{initials}
   +{}{ }{surname}
   +{}{ }{jr}
  }
  \BibSpec{middleauthor}{
   +{,}{}{transition}
   +{}{ }{initials}
   +{}{ }{surname}
   +{}{ }{jr}
  }
  \BibSpec{lastauthor}{
   +{,}{ and}{transition}
   +{}{ }{initials}
   +{}{ }{surname}
   +{}{ }{jr}
  }
}{% initials? FALSE:
  \BibSpec{firstauthor}{
   +{}{}{given}
   +{}{ }{surname}
   +{}{ }{jr}
  }
  \BibSpec{coauthor}{
   +{}{ and}{transition}
   +{}{ }{given}
   +{}{ }{surname}
   +{}{ }{jr}
  }
  \BibSpec{middleauthor}{
   +{,}{}{transition}
   +{}{ }{given}
   +{}{ }{surname}
   +{}{ }{jr}
  }
  \BibSpec{lastauthor}{
   +{,}{ and}{transition}
   +{}{ }{given}
   +{}{ }{surname}
   +{}{ }{jr}
  }
} % end conditional code for initials option
%    \end{macrocode}
%
%    The Chicago Manual of Style suggests that it is slightly better not
%    to italicize `et al' and some other extremely common abbreviations
%    inherited from Latin. (Compare `etc'.)
%    \begin{macrocode}
\newcommand{\etaltext}{et al.}
\newcommand{\SubEtal}[1]{\etaltext}
%    \end{macrocode}
%
%    Some ``et alii'' requirements can also be handled this way, e.g.,
%    if two authors should be printed as ``A and B'' but three or more
%    should be printed as ``A et al.'', it could be done readily enough
%    by changing only the specs for middleauthor and lastauthor:
%\begin{verbatim}
%\BibSpec{middleauthor}{}
%\BibSpec{lastauthor}{
% +{}{ \SubEtal}{surname}
%}
%\end{verbatim}
%    For more complicated et alii rules, well, just throw your brain
%    into gear and think of something clever. I think you will find
%    enough points of attack to make it possible one way or another.
%
%    When we consider editor names we have to think about some further
%    complications. First, for the case of a book where editor names are
%    listed in place of author names, just copy the same style with a
%    bit of added text at the end.
%    \begin{macrocode}
\newcommand{\PrintEditorsA}[1]{%
  \PrintNames{editor}{\bfseries}{ (ed\Plural{s}.)}{#1}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\BibSpecAlias{firsteditor}{firstauthor}
\BibSpecAlias{coeditor}{coauthor}
\BibSpecAlias{middleeditor}{coeditor}
\BibSpecAlias{lasteditor}{coeditor}
%    \end{macrocode}
%
%
%    \begin{macrocode}
\newcommand{\PrintEditorsB}[1]{%
  \PrintNames{ed}{(}{\SwapBreak{,}~ed\Plural{s}.)}{#1}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\BibSpec{firsted}{
 +{}{}{initials}
 +{}{ }{surname}
 +{}{ }{jr}
}
\BibSpec{coed}{
 +{}{ and }{initials}
 +{}{ }{surname}
 +{}{ }{jr}
}
\BibSpec{middleed}{
 +{,}{ }{initials}
 +{}{ }{surname}
 +{}{ }{jr}
}
\BibSpec{lasted}{
 +{,}{ and }{initials}
 +{}{ }{surname}
 +{}{ }{jr}
}
%    \end{macrocode}
%
%    So what does \cs{NameSetup} do anyway (you may ask)?
%    \begin{macrocode}
\def\NameSetup#1{%
  \@xp\NameSetup@a
    \csname setbib@first#1\@xp\endcsname
    \csname setbib@co#1\@xp\endcsname
    \csname setbib@middle#1\@xp\endcsname
    \csname setbib@last#1\endcsname
}%
%    \end{macrocode}
%
%    \cs{NameSetup@a} first builds a queue consisting of, e.g.,
%\begin{verbatim}
%\setbib@firstauthor
%\setbib@middleauthor \setbib@middleauthor ...
%\setbib@lastauthor
%\end{verbatim}
%    and then checks to see if
%    \cs{name@queue} = \cs{setbib@firstauthor}
%    \cs{setbib@lastauthor}, because that means there were only two
%    authors; and then we substitute \cs{setbib@coauthor} for the second
%    token.
%    \begin{macrocode}
\def\NameSetup@a#1#2#3#4#5{%
  \def\name##1{\@nx#3}%
  \edef\name@queue{\@nx#1\@gobblefour#5\@nx#4\@empty\@empty}%
  \def\@tempa{#1}%
  \ifx\name@queue\@tempa \let\SingularPlural\@firstoftwo
  \else \let\SingularPlural\@secondoftwo
  \fi
  \def\@tempa{#1#4}%
  \ifx\name@queue\@tempa\def\name@queue{#1#2}\fi
  \let\name\doName
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\doName#1{%
  \iffalse{\fi \@xp\get@namer\name@queue}%
  \name@split#1,{},{},\@nil \setbib@@name
}
%    \end{macrocode}
%
%    Short for get-name-handler.
%    \begin{macrocode}
\def\get@namer#1{\let\setbib@@name#1%
  \@xp\def\@xp\name@queue\@xp{\iffalse}\fi
}
%    \end{macrocode}
%
%    Some tests of various author name forms.
%\begin{verbatim}
%                    ()->[,] [@] []
%           (Abiteboul)->[A] [biteboul] [@]
%        (S. Abiteboul)->[S] [. Abiteboul] [@]
%      ( S. Abiteboul )->[S] [. Abiteboul ] [@]
%     (S. T. Abiteboul)->[S] [. T. Abiteboul] [@]
%      (S.T. Abiteboul)->[S] [.T. Abiteboul] [@]
%     (Steve Abiteboul)->[S] [teve Abiteboul] [@]
%       (Abiteboul, S.)->[A] [biteboul] [S]
%    (Abiteboul, S. T.)->[A] [biteboul] [S]
%     (Abiteboul, S.T.)->[A] [biteboul] [S]
%       (P. Kanellakis)->[P] [. Kanellakis] [@]
%   (K. I. Be\u {i}dar)->[K] [. I. Be\u {i}dar] [@]
%  (\protect \SubEtal )->[\protect ] [\SubEtal ] [@]
%\end{verbatim}

%    If arg3 = \qc{\@} then the name is probably given in uninverted order.
%    This may be because it is a Chinese name, and that is the correct
%    order; but check to see if the name starts with an initial, which
%    probably should get a warning.
%
%    \begin{macrocode}
\def\EmptyNameWarning{\amsrefs@warning{Empty author or editor name}}
%    \end{macrocode}
%
%    Need to check for extra braces here.
%    \begin{macrocode}
\def\NameCheck#1#2#3\@nil{%
  \def\@tempa{#1#2}%
  \edef\@tempa{\@nx\NameCheck@a\macrotext\@tempa}%
  \@tempa ??\@nil
}
%    \end{macrocode}
%
%    This gets rid of the leading space in \cs{on@line}.
%    \begin{macrocode}
\def\MessageBreakNS{\MessageBreak\romannumeral`\^^@}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\NameCheck@a#1#2#3\@nil{%
  \if.#2\ifnum\catcode`#1=11
    \amsrefs@warning{%
Write `Lastname, First' to ensure proper formatting\MessageBreakNS}%
  \fi\fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\name@split#1,#2#3,#4#5,#6\@nil{%
  \bib@name@init
  \def\bib'surname{#1}%
  \ifx\bib'surname\@empty \EmptyNameWarning\fi
  \def\bib'given{#2#3}%
  \def\bib'jr{#4#5}%
  \ifx\@empty\bib'given \NameCheck #1 ??\@nil
%    \end{macrocode}
%    Throwing in an extra period at the end makes this work even when
%    the given names consist of a single letter with no period or
%    anything. I think. [mjd,2002-02-18]
%    \begin{macrocode}
  \else \extract@initials{#2#3\ignorespaces.}%
  \fi
}
%    \end{macrocode}
%
%    Extracting initials from the author name is an arduous chore,
%    because of the various cases that need to be handled:
%\begin{verbatim}
%\doName{Harish, \'{E}tienne-P\^{\i}erre}
%\doName{Jones, H. C.}
%\doName{Jones, H.-C.}
%\doName{Jones, H.C., Jr.}
%\doName{Jones, Harold, IV}
%\doName{\'Swiderski, Jacek},
%\doName{\'{S}widerski, Jacek},
%\doName{{\'{S}}widerski, Jacek},
%\doName{Cartan, \'Elie}
%\doName{Cartan, \'{E}lie}
%\doName{Cartan, \'E.}
%\doName{Shua Tae Yua}
%\doName{Shua, Tae Yua}
%\doName{Harish-Chandra}
%\end{verbatim}
%
%    So we do it in several steps. The result that we want to end up
%    with is defining \cs{bib'initials} to contain, e.g.,
%\begin{verbatim}
% Jones, Harold C.  --> H\InitialPunct C\InitialPunct
%\end{verbatim}

% \begin{enumerate}
% \item Expand accent commands in a way that replaces an accented letter
%    with only the base letter, if it is lowercase (keep the accent
%    command if the base letter is uppercase).
%
% \item Convert periods to spaces.
%
% \item Add a space before hyphens, to ensure that \cs{InitialPunct}
%    will get added in the middle of a hyphenated compound name.
%
% \end{enumerate}
%    \begin{macrocode}
\def\xi@accent#1#2{%
  \ifnum\uccode`#2=`#2 \@nx#1{#2}\else #2\fi
}
%    \end{macrocode}
%
%    This reads up to the next period and replaces it by a space, then
%    recurses. The \cs{@gobbletwo} terminates the recursion done by
%    \cs{xi@hyph}.
%    \begin{macrocode}
\def\xi@dot#1.{\xi@hyph#1@\space\@gobbletwo - \xi@dot}
%    \end{macrocode}
%
%    This reads up to the next hyphen and adds a space before it, then
%    recurses.
%    \begin{macrocode}
\def\xi@hyph#1-{#1 -\xi@hyph}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\extract@initials#1{%
  \begingroup
  \def\i{i}\def\j{j}%
  \def\ae{a}\let\oe\ae \let\o\ae \let\dh\ae \let\dj\ae \let\l\ae
  \let\ng\ae \let\ss\ae \let\th\ae
  \def\"{\xi@accent\"}\def\'{\xi@accent\'}\def\.{\xi@accent\.}%
  \def\={\xi@accent\=}\def\^{\xi@accent\^}\def\`{\xi@accent\`}%
  \def\~{\xi@accent\~}\def\b{\xi@accent\b}\def\c{\xi@accent\c}%
  \def\d{\xi@accent\d}\def\H{\xi@accent\H}\def\k{\xi@accent\k}%
  \def\r{\xi@accent\r}\def\t{\xi@accent\t}\def\u{\xi@accent\u}%
  \def\v{\xi@accent\v}%
%    \end{macrocode}
%    The use of \cs{@gobbletwo} here is to terminate the recursion done
%    by \cs{@xidot} and \cs{xi@hyph}.
%    \begin{macrocode}
  \protected@edef\i{\xi@dot#1\@gobbletwo -\@gobbletwo .\@empty\space}%
  \def\j{\uccode\fam=64 \advance\fam\@ne
    \ifnum\fam>122 \@firstoftwo \fi
    \j
  }%
  \fam=97 \j
  \uppercase\@xp{\@xp\def\@xp\i\@xp{\i}}%
  \let\InitialSpace=F%
  \@xp\extract@initials@b\i @@ \relax
  \edef\j{\def\@nx\bib'initials{\the\toks@}}%
  \expandafter\endgroup\j
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\extract@initials@b#1#2@{%
  \ifx @#1\@xp\@gobbletwo
  \else
    \ifx F\InitialSpace
      \toks@{#1#2\ini@punct}%
      \let\InitialSpace=T%
    \else
      \toks@\@xp{\the\toks@ \ini@space #1#2\ini@punct}%
    \fi
  \fi
  \extract@initials@c
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\extract@initials@c#1 {\extract@initials@b}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\ini@punct{\InitialPunct}
\newcommand{\InitialPunct}{.}
\def\ini@space{\futurelet\@let@token\ini@space@a}
\def\ini@space@a{\ifx -\@let@token \else\InitialSpace\fi}
\newcommand{\InitialSpace}{\nobreakspace}
%    \end{macrocode}
%
%    This should probably get some extra work to handle the possibility
%    that a name was entered in the bibliography in normal (uninverted)
%    order. [mjd,2000/04/07]
%    \begin{macrocode}
\def\lastName#1{\lnscan@a#1,\@nil}
\def\lnscan@a#1,#2\@nil{#1}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\discardName#1{}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\CiteNames}[1]{%
  \PrintSeries{\name}%
    {\lastName}%
    {}{ and \lastName}%
%    \end{macrocode}
%    For three or more names: print `et al' instead of the last name.
%    Have to putz around with the spacefactor a bit or the comma between
%    name and year will not be applied.
%    \begin{macrocode}
    {}{\discardName}%
    {}{ \etaltext\@\discardName}%
    {}%
    {#1}%
    {}%
}
%    \end{macrocode}
%
%    This needs to be called at the beginning of a list of cites to
%    reset a few things.
%    \begin{macrocode}
\def\cites@init{%
  \gdef\prev@names{???}\let\cites@init\@empty
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\PrintCiteNames#1{%
  \leavevmode
  \def\@tempa{#1}%
  \ifx\@tempa\prev@names
  \else
    \gdef\prev@names{#1}%
    \@xp\ifx\@car#1.\@nil\CitePrintUndefined
      #1\relax
    \else
      \CiteNames{#1}%
    \fi
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\PrintCNY#1#2{%
  \PrintCiteNames{#1}%
  \@ifnotempty{#2}{\@addpunct{,} #2}%
}
%    \end{macrocode}
%
% \subsection{Special formatting for other fields}
%
%    Reviews are handled as a list to support the theoretical
%    possibility of multiple reviews.
%    \begin{macrocode}
\newcommand{\ReviewList}[1]{%
  \PrintSeries{\review}{}{,}{ }{,}{ }{,}{ }{}{#1}{}%
}
%    \end{macrocode}
%
%    And similarly for ISBN numbers. There seem to be a few different
%    situations where one book might have two different ISBN numbers.
%    Here are the ones I know of so far [mjd,2002-02-18]: separate ISBN
%    numbers for hardback and paperback; separate ISBN numbers for U.S.
%    edition and European edition.
%    \begin{macrocode}
\newcommand{\ISBNList}[1]{%
  \PrintSeries{\isbn}{}{,}{ }{,}{ }{,}{ }{ISBN }{#1}{}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\voltext}{vol.~}
%    \end{macrocode}
%    Scan the contents of a page value to see if it is a single page.
%    Presence of \cs{ndash} or hyphen is taken to mean no. Probably
%    should test also for spaces and commas. [mjd,2000/01/24]
%    \begin{macrocode}
\newcommand{\DashPages}[1]{%
  p\pp@scan@a#1@\ndash p@\ndash{\pp@scan#1@-p@-{}\@nil}\@nil.~#1%
}
\def\pp@scan#1-#2@-#3#4\@nil{#3}
\def\pp@scan@a#1\ndash#2@\ndash#3#4\@nil{#3}
%    \end{macrocode}
%
%    If we have eprint info and pages info and no journal name, the
%    pages information is presumably the number of pages in the eprint.
%    \begin{macrocode}
\newcommand{\eprintpages}[1]{%
  #1\IfEmptyBibField{eprint}{}{\IfEmptyBibField{journal}{ pp.}{}}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\EnglishInitialCaps}[1]{%
  \ifx\@empty\bib@language \let\bib@language\biblanguagedefault\fi
  \ifx\bib@language\biblanguageEnglish \@xp\inicap\fi
  {#1}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\PrintThesisType#1{%
  \thesis@type#1?\@nil{#1}%
}
\def\thesis@type#1#2\@nil#3{%
  \ifx p#1Ph.D. Thesis%
  \else\ifx m#1Master's Thesis%
  \else #3\fi\fi
}
%    \end{macrocode}
%
%    Perhaps need to add allowbreak penalties at the parens in a DOI.
%    Also what about prohibiting a break after the leading S?
%    \begin{macrocode}
\newcommand{\PrintDOI}[1]{%
  DOI #1%
  \IfEmptyBibField{volume}{, (to appear in print)}{}%
}
%    \end{macrocode}
%
%    When this function is called, \cs{bib@year} has been copied into
%    \cs{bib@date} if necessary.
%
%    [This code has gotten sort of circular, it seems. Needs
%    reassessment [mjd,2000/04/07]]
%    \begin{macrocode}
\def\bib@parsedate{%
  \@xp\bib@parsedate@a\bib@date ---\@nil
}
\let\bib@year\@empty
\let\bib@month\@empty
\let\bib@day\@empty
\def\bib@parsedate@a#1-#2-#3-#4\@nil{%
  \def\bib@year{#1}\def\bib@month{#2}\def\bib@day{#3}%
  \ifx\@empty\bib@month \let\bib'date\bib@year
  \else \def\bib'date{#1-#2-#3}%
  \fi
}
%    \end{macrocode}
%
%    Print date in different forms depending on DOI and volume
%    information.
%    \begin{macrocode}
\newcommand{\PrintDatePV}[1]{%
  \IfEmptyBibField{doi}{\let\@tempa\PrintDate}{%
    \IfEmptyBibField{volume}{\let\@tempa\PrintDatePosted}{%
      \let\@tempa\PrintDate}}%
  \@tempa{#1}%
}
%    \end{macrocode}
%
%    Intent is to handle variations such as \texttt{1987},
%    \texttt{August 1987}, \texttt{1987-08}, and \texttt{1987-08-14}.
%    If the month is present, print August or Aug. or 08 or nothing,
%    at the behest of the bib style.
%
%    We've taken some special care to parse out the date info ahead of
%    time, so this function just discards arg 1 and uses the
%    already-parsed value.
%    \begin{macrocode}
\newcommand{\PrintDate}[1]{(\print@date)}
\newcommand{\PrintDateB}[1]{\print@date}
\def\print@date{%
  \ifx\bib@month\@empty \else\print@month@day \fi
  \bib@year
}
%    \end{macrocode}
%
%    With the Babel package, month names for a given language are
%    typically available in a macro \cs{month@}\emph{language}:
%\begin{verbatim}
%\def\month@german{\ifcase\month\or
%  Januar\or Februar\or M\"arz\or April\or Mai\or Juni\or
%  Juli\or August\or September\or Oktober\or November\or Dezember\fi}
%\end{verbatim}
%    However this is not true for English.
%    \begin{macrocode}
\def\print@month@day{%
  \ifcase 0\bib@month\ignorespaces
  \or January\or February\or March\or April\or May\or June\or
      July\or August\or September\or October\or November\or December\or
      Winter\or Spring\or Summer\or Fall\else Unknown Month%
  \fi
  \ifx\@empty\bib@day \else \space\number 0\bib@day,\fi
  \space
}
%    \end{macrocode}
%
%    You can use \cs{PrintYear} if you want to suppress month/day even
%    when supplied in the data.
%    \begin{macrocode}
\newcommand{\PrintYear}[1]{\bib@year}
%    \end{macrocode}
%
%    This one is special for AMS use.
%    \begin{macrocode}
\newcommand{\PrintDatePosted}[1]{\unskip,  posted on \print@date}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\CardinalNumeric}[1]{%
  \number#1\relax
  \if\ifnum#1<14 \ifnum#1>\thr@@ T\else F\fi\else F\fi T%
    th%
  \else
    \@xp\keep@last@digit\@xp#1\number#1\relax
    \ifcase#1th\or st\or nd\or rd\else th\fi
  \fi
}
\def\keep@last@digit#1#2{%
  \ifx\relax#2\@xp\@gobbletwo
  \else #1=#2\relax
  \fi
  \keep@last@digit#1%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\PrintEdition}[1]{%
  \afterassignment\print@edition
  \count@ 0#1\relax\@nil
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\editiontext}{ed.}
%    \end{macrocode}
%    If the number assignment swept up all the contents, produce a
%    cardinal number from \cs{count@}.
%    \begin{macrocode}
\def\print@edition#1#2\@nil{%
  \ifx\relax#1\relax
    \ifnum\count@>\z@ \CardinalNumeric\count@
    \else ??th%
    \fi
    \ \editiontext
  \else \ifnum\count@>\z@\number\count@\fi #1#2\relax
  \fi
}
%    \end{macrocode}
%
%    The braces here are just a simple way to terminate the number
%    scanning.
%    \begin{macrocode}
\newcommand{\SentenceSpace}{\relax\ifhmode{\spacefactor3000} \fi}
%    \end{macrocode}
%
%    For now, this does nothing. Could do a url/hyperlink or something.
%    \begin{macrocode}
\newcommand{\eprint}[1]{#1}
%    \end{macrocode}
%    The \fn{www.arXiv.org} recommendations for citing their eprints are
%    found at \url{http://xxx.lanl.gov/help/faq/references}, including
%    these examples:
%\begin{verbatim}
%arXiv:hep-th/9910001
%arXiv:math.AT/9910001
%arXiv:physics.acc-ph/9911027
%\end{verbatim}
%
%    \begin{macrocode}
\BibSpec{article}{%
  +{}{\PrintAuthors}       {author}
  +{,}{ \textit}      {title}
  +{.}{ \textit}      {subtitle}
  +{,}{ }             {journal}
%  +{,}{ ISSN }        {ISSN}
  +{} { \textbf}      {volume}
%    \end{macrocode}
%    The date form is tricky depending on presence or absence of DOI.
%    \begin{macrocode}
  +{} { \PrintDatePV} {date}
  +{,}{ \PrintDOI}    {doi}
  +{,}{ \eprint}      {eprint}
  +{} { \parenthesize}{status}
  +{,}{ \eprintpages} {pages}
  +{,}{ }             {note}
  +{.}{}              {transition}
  +{} {\SentenceSpace \ReviewList}   {review}
  +{}{ \parenthesize}{language}
  +{}{}{part}
}
\BibSpec{book}{%
  +{}{\PrintAuthors}{author}
  +{}{\PrintEditorsA}{editor}
  +{,}{ \textit}{title}
  +{:}{ \textit}{subtitle}
  +{,}{ }{type}
  +{,}{ \EnglishInitialCaps}{booktitle}
  +{,}{ \PrintEdition}{edition}
  +{,}{ }{series}
  +{,}{ vol.~}{volume}
  +{,}{ part~}{part}
  +{,}{ }{publisher}
  +{,}{ }{organization}
  +{,}{ }{place}
  +{,}{ \PrintDateB}{date}
  +{,}{ \ISBNList}{isbn}
  +{}{ \parenthesize}{status}
  +{,}{ }{note}
  +{.}{}{transition}
  +{}{\SentenceSpace \ReviewList}{review}
  +{}{ \parenthesize}{language}
}
\BibSpec{collection.article}{%
  +{}{\PrintAuthors}{author}
  +{,}{ \textit}{title}
  +{,}{ }{type}
  +{,}{ \EnglishInitialCaps}{booktitle}
  +{,}{ \PrintEdition}{edition}
  +{}{ \PrintEditorsB}{editor}
  +{,}{ Proc.\ }{conference}
  +{,}{ }{series}
  +{,}{ vol.~}{volume}
  +{,}{ part~}{part}
  +{,}{ }{publisher}
  +{,}{ }{organization}
  +{,}{ }{place}
  +{,}{ \PrintDateB}{date}
  +{,}{ \DashPages}{pages}
  +{,}{ \ISBNList}{isbn}
  +{}{ \parenthesize}{status}
  +{,}{ }{note}
  +{.}{}{transition}
  +{}{\SentenceSpace \ReviewList}{review}
  +{}{ \parenthesize}{language}
}
\BibSpec{report}{%
  +{}{\PrintAuthors}{author}
  +{}{\PrintEditorsA}{editor}
  +{,}{ \textit}{title}
  +{,}{ }{type}
  +{,}{ \EnglishInitialCaps}{booktitle}
  +{,}{ Technical Report }{number}
  +{,}{ }{series}
  +{,}{ vol.~}{volume}
  +{,}{ part~}{part}
  +{,}{ }{publisher}
  +{,}{ }{organization}
  +{,}{ }{institution}
  +{,}{ }{place}
  +{,}{ \PrintDateB}{date}
  +{,}{ \ISBNList}{isbn}
  +{,}{ \eprint}      {eprint}
  +{}{ \parenthesize}{status}
  +{,}{ }{note}
  +{.}{}{transition}
  +{}{\SentenceSpace \ReviewList}{review}
  +{}{ \parenthesize}{language}
}
\BibSpec{thesis}{%
  +{}{\PrintAuthors}{author}
  +{}{\PrintEditorsA}{editor}
  +{,}{ \textit}{title}
  +{,}{ \PrintThesisType}{type}
  +{,}{ part~}{part}
  +{,}{ }{organization}
  +{,}{ }{institution}
  +{,}{ }{place}
  +{,}{ \PrintDateB}{date}
  +{}{ \parenthesize}{status}
  +{,}{ }{note}
  +{.}{}{transition}
  +{}{\SentenceSpace \ReviewList}{review}
  +{}{ \parenthesize}{language}
}
%    \end{macrocode}
%
%    \begin{macrocode}
\BibSpecAlias{periodical}{book}
\BibSpecAlias{collection}{book}
\BibSpecAlias{proceedings}{book}
\BibSpecAlias{manual}{book}
\BibSpecAlias{misc}{book}
\BibSpecAlias{unpublished}{book}
\BibSpecAlias{proceedings.article}{collection.article}
\BibSpecAlias{techreport}{report}
%    \end{macrocode}
%
%    \begin{macrocode}
\edef\setbib@incollection{%
  \@xp\@nx\csname setbib@collection.article\endcsname
}
\edef\setbib@inproceedings{%
  \@xp\@nx\csname setbib@collection.article\endcsname
}
%    \end{macrocode}
%
%    Some more entry types for implementing abbreviations.
%    \begin{macrocode}
\BibSpec{journal}{%
  +{}{}{journal}
  +{}{ \parenthesize}{shortjournal}
  +{,}{ ISSN }{issn}
}
\BibSpec{person}{%
  +{}{\PrintAuthors}{name}
}
\BibSpec{publisher}{%
  +{,}{ }{publisher}
  +{,}{ }{organization}
  +{,}{ }{institution}
  +{,}{ }{place}
}
%    \end{macrocode}
%
%    Some special handling for ``et alii'' or ``and others''.
%    \begin{macrocode}
\DefineName{alii}{\etaltext}%
\DefineName{others}{\etaltext}%
%    \end{macrocode}
%
% \subsection{Extracting entries from an external database file}
%
%    If you have a \lat/ document that contains \pkg{amsrefs}-style
%    \cs{bib} entries, you can use it as a database and extract items
%    from it for use in another document. In typical relatively simple
%    scenarios, the extraction can be done by \lat/ itself on the first
%    pass, so that citations in the text will be successfully resolved
%    on the second pass (possibly even the first, depending on what kind
%    of bibliography sorting is used).
%    \begin{macrocode}
\newread\bib@dbfile
\newcommand{\ReadBibData}[1]{%
  \IfFileExists{#1.ltb}{%
    \openin\bib@dbfile=\@filef@und \relax
  }{%
    \IfFileExists{#1.ltx}{%
      \openin\bib@dbfile=\@filef@und \relax
    }{%
      \IfFileExists{#1.tex}{%
        \openin\bib@dbfile=\@filef@und \relax
      }{%
        \begingroup \NoBibDBFile{#1}%
        \let\ReadBibPrelim\endgroup
      }%
    }%
  }%
  \ReadBibPrelim
}
%    \end{macrocode}
%
%    Maybe this should have a star form to mean, read everything into
%    memory? Then it could be used at begin-document to make all the info
%    available from the get-go. [mjd,2001-12-13]
%    \begin{macrocode}
\def\bblname{\jobname}
\newcommand{\bibselect}{%
  \@ifstar{\let\@bibdef\copy@bibdef \BibSelect}{\BibSelect}%
}
\newcommand{\BibSelect}[2][\bblname]{%
  \if@filesw
    \typeout{Trying to create bbl file `#1.bbl' ...}%
    \def\bibselect@msg{\typeout{ ... rats. Unable to create bbl file.}}%
    \let\@open@bbl@file\OpenBBLFile
    \ifx\@bibdef\copy@bibdef \else \let\@bibdef\selective@bibdef \fi
    \ReadBibData{#2}%
  \fi
  \@close@bbl@file
  \let\@bibdef\normal@bibdef
  \input{#1.bbl}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\NoBibDBFile#1{%
  \amsrefs@warning{No data file #1.ltb (.ltx, .tex) found}%
}
%    \end{macrocode}
%
%    Start reading the db file one line at a time,
%    looking for \verb'\begin{biblist}'.
%    \begin{macrocode}
\def\ReadBibLoop{%
  \ifeof\bib@dbfile
    \closein\bib@dbfile
    \EmptyBibDBFile
    \@xp\@gobble
  \else
    \read\bib@dbfile to\CurLine
    \lowercase{\@xp\ReadBibLoop@a\CurLine} \@nil
  \fi
  \ReadBibLoop
}
\def\ReadBibPrelim{%
  \edef\bib@providefile{\@nx\ProvidesFile{\@filef@und}\relax}%
  \begingroup
  \let\do\@makeother \dospecials
%    \end{macrocode}
%    Spaces and comments can simply be ignored, and that makes our scanning
%    task easier.
%    \begin{macrocode}
  \catcode32=9 \catcode`\%=14 \relax
  \ReadBibLoop
  \endgroup
  \bib@providefile
}
\edef\BibListMarker{\@backslashchar begin\string{biblist\string}}
\long\def\ReadBibLoop@a#1 #2\@nil{%
  \def\@tempa{#1}%
  \ifx\@tempa\BibListMarker
    \begingroup \def\ReadBibLoop{\endgroup\ReadBibEntries}%
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\ReadBibEntries{%
  \endgroup % turn off the mostly-verbatim catcodes
  \ReadBibEntry
  \begingroup % pair for upcoming endgroup
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\ReadBibEntry{%
  \ifeof\bib@dbfile
    \closein\bib@dbfile
    \@xp\@gobble
  \else
    \read\bib@dbfile to\CurLine
    \ifx\space\CurLine
    \else \ifx\@empty \CurLine
    \else \@xp\ReadBibLoop@e\CurLine\@nil
    \fi\fi
  \fi
  \ReadBibEntry
}
\long\def\ReadBibLoop@e#1#2\@nil{%
  \ifx\bib#1%
    \CurLine % just exec it
  \else \ifx\end#1\begingroup\let\ReadBibEntry\endgroup\fi
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\let\bbl@out=\relax
\let\bbl@write\@gobble
\let\@open@bbl@file\relax
\let\@close@bbl@file\relax
\def\OpenBBLFile{%
  \if@filesw
    % Just use the next unused output stream
    \count@\count17 \advance\count@\@ne
    \ifnum\count@<\sixt@@n
      \global\chardef\bbl@out=\count@
      \immediate\openout\bbl@out=\bblname.bbl\relax
      \global\let\@close@bbl@file\CloseBBLFile
      \gdef\bbl@write{\immediate\write\bbl@out}%
    \else
      \ch@ck\count@\sixt@@n\write
    \fi
  \fi
  \global\let\@open@bbl@file\relax
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\CloseBBLFile{%
  \immediate\closeout\bbl@out\relax
  \global\let\@close@bbl@file\relax
  \global\let\bbl@write\@gobble
  \global\let\bbl@out\relax
}
%    \end{macrocode}
%
% \subsection{Cites}
%
%    These variables are named to follow the precedent set by Arseneau's
%    \pkg{cite} package. \cs{citepunct} is used to separate multiple
%    cites; \cs{citemid} is used to separate a cite number from
%    additional information such as \qq{Theorem 4.9}.
%    \begin{macrocode}
\def\citeleft{[}
\def\citeright{]}
\def\citemid{,\penalty9999 \space}
\def\citepunct{,\penalty9999 \hskip.13em plus.1em minus.05em\relax}
\def\citeform#1{}
%    \end{macrocode}
%    The following definition provides some indirection that helps to
%    deal with author-year object cites.
%    \begin{macrocode}
\def\@citeleft{\citeleft}
%    \end{macrocode}
%
%    The information used by \cs{cite} for key \texttt{moo} is stored in
%    \cs{b@moo} in the form
%\begin{verbatim}
%\citesel{status1}{status2}{number}{year}{backref-info}
%\end{verbatim}%
%
%    The first status flag is 1 if this key has already been cited
%    earlier in the same document; 0 if not yet. This is used in some
%    bibliography schemes to print a full list of author names for the first
%    citation and an abbreviated author list for subsequent citations.
%
%    The second status flag is 1 if this key has already been used by a
%    define-cite command (such as \cs{bib}); 0 otherwise. This makes it
%    possible to issue a warning message as soon as the conflict is seen,
%    on the first \lat/ run, instead of on a subsequent run during the
%    processing of the \fn{.aux} file.
%
%    When an author/year citation scheme is in use, args 3 and 4 hold
%    resp author names and year. Otherwise arg 3 simply holds a cite
%    number.
%
%    And finally, arg 5 holds a list of backref pointers indicating the
%    locations in the document where this entry has been cited.
%    \begin{macrocode}
\def\citesel@author#1#2#3#4#5{\PrintCiteNames{#3}}
\def\citesel@authoryear#1#2#3#4#5{\PrintCNY{#3}{#4}}
\def\citesel@object#1#2#3#4#5{\PrintCiteNames{#3} \citeleft#4}
\def\citesel@year#1#2#3#4#5{#4}
\def\citesel@number#1#2#3#4#5{#3}
\let\citesel\citesel@number
%    \end{macrocode}
%
%    \begin{macrocode}
\def\citesel@update#1#2#3#4#5#6{%
  \gdef#6{\citesel 1#2{#3}{#4}{#5}}%
  \if 0#1\relax \g@addto@macro\all@cites{#6}\fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\citesel@backrefs#1#2#3#4#5{%
  \@for\@tempc:=#5\do{\@xp\backref\@xp{\@tempc}}%
}
\def\backref#1{\toks@{#1}\typeout{backref=[\the\toks@]}}
%
%    \end{macrocode}
%
%    Here is the difference between the various optional forms of
%    \cs{cite}:
%\begin{verbatim}
% \cite{xyz} -> \cite@a\citesel{xyz}{}{}
%            -> \cite@bc\b@xyz\citesel{}
%
% \cite{xyz}*{blub} -> \cite@a\citesel{xyz}{}{blub}
%                   -> \cite@bc\b@xyz\citesel{blub}
%
% \cite[blub]{xyz} -> \cite@a\citesel{xyz}{blub}{}
%                  -> \cite@bc\b@xyz\citesel{blub}
%\end{verbatim}
%
%    Need to handle the standard [...] option for compatibility's sake.
%    \begin{macrocode}
\newcommand{\InnerCite}[1]{\star@{\cite@a\citesel{#1}}{}}
\renewcommand{\cite}[2][]{%
  \if\cite@single#2,\@gobble \else\MultipleCiteKeyWarning{#2}{#1}\fi
  \@ifempty{#1}{%
    \cites@o{#2}%
  }{%
    \ObsoleteCiteOptionWarning
    \cites@a{*{#1}}{#2}%
  }%
}
\edef\cite@single#1,#2{\iffalse{\fi\iffalse{\fi\string}#2.\string}}
%    \end{macrocode}
%
%    A list of simple cites. Make it robust in case used inside a figure
%    caption. (But then also, by the way, listoffigures should provide
%    special handling.)
%    \begin{macrocode}
\DeclareRobustCommand{\cites}{\cites@a{}}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cites@o#1{\star@{\cites@oo{#1}}{}}
\def\cites@oo#1#2{\@ifempty{#2}{\cites@a{}{#1}}{\cites@a{*{#2}}{#1}}}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cites@a#1#2{%
  \begingroup
  \toks@{\endgroup \cites@b{#1}}%
  \vdef\@tempa{#2}%
  \edef\@tempa{\the\toks@ \@firstofone{\@xp\zap@space\@tempa} \@empty}%
  \@tempa,\@empty
  \edef\@tempa{\endgroup\@nx\citelist{\the\toks@}}%
  \@tempa
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cites@b#1#2,#3{%
  \begingroup
  \toks@{\InnerCite{#2}#1}%
  \ifx\@empty#3\@xp\@gobble\fi
  \cites@c#3%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cites@c#1,#2{%
  \toks@\@xp{\the\toks@ \InnerCite{#1}}%
  \ifx\@empty#2\@xp\@gobble\fi
  \cites@c#2%
}
%    \end{macrocode}
%
%    Grouping that encloses an entire cite block (a single cite or a
%    list of cites).
%    \begin{macrocode}
\def\cite@begingroup{\begingroup\let\cite@begingroup\relax}
\let\cite@endgroup\endgroup
%    \end{macrocode}
%
%    The job of \cs{cite@a} is to convert the cite key to all catcode-12
%    characters before passing it on to \cs{cite@b}.
%    \begin{macrocode}
%% #1<-\citesel@year #2<-key
\def\cite@a#1#2{%
  \cite@begingroup
  \cites@init
  \let\citesel#1\relax
  \ifx\citesel\citesel@author \let\citeleft\@empty \let\citeright\@empty\fi
  \begingroup
  \toks@{\endgroup \cite@b}%
  \vdef\@tempa{#2}%
  \edef\@tempa{\the\toks@{\@firstofone{\@xp\zap@space\@tempa} \@empty}}%
  \@tempa
}
%    \end{macrocode}
%
%    \begin{macrocode}
% #1<-key #2<-star-optional-arg
\def\cite@b#1#2{%
  \@xp\cite@bc\csname b@#1\@xp\endcsname {#1}{#2}%
}
%    \end{macrocode}
%
%    If it's uninitialized, plug in an empty cite structure.
%    \cs{cite@bc} should be executed only once for a given instance of a
%    cite key. All further processing should go through \cs{cite@cj}.
%    \begin{macrocode}
\def\cite@bc#1#2{%
  \def\CurrentCite{#2}%
  \ifx#1\@@undefined \global\let#1\relax\fi
  \ifx#1\relax \global\let#1\empty@cite \fi
  \@xp\cite@nobib@test#1{}{}{}{}\@nil#1%
  \cite@cj#1%
}
\def\empty@cite{\citesel 00{}{}{}}
%    \end{macrocode}
%
%    If arg 4 is empty, it means there wasn't any \cs{bib} command that
%    defined a valid number.
%    \begin{macrocode}
% #1 \citesel #2 0 #3 0 #4 number #5 etc
\def\cite@nobib@test#1#2#3#4#5\@nil#6{%
  \@ifempty{#4}{%
    \G@refundefinedtrue
    \UndefinedCiteWarning
    \xdef#6{\@nx\citesel #2#3{%
      \@nx\CitePrintUndefined{\CurrentCite}}{}{}}%
  }{}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\DeclareRobustCommand{\CitePrintUndefined}[1]{%
  \begingroup\fontshape{n}\fontseries\mddefault \ttfamily ?#1\endgroup
}
%    \end{macrocode}
%    This has to be a \cs{let}, not a \cs{def}.
%    \begin{macrocode}
\let\CPU@normal\CitePrintUndefined
%
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@multiple#1,#2#3\@nil{\ifx\relax#2F\else T\fi T}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\MultipleCiteKeyWarning#1#2{%
  \amsrefs@warning{%
    Use of \string\cites\space is recommended instead of %
    \string\cite\space\MessageBreak
    for multiple cites '#1'}%
  \@ifnotempty{#2}{%
    \amsrefs@warning{Star option requires \string\citelist\space here}%
  }%
  \global\let\MultipleCiteKeyWarning\@gobbletwo
}
%    \end{macrocode}
%
%\begin{verbatim}
%\citesel{cite-already?}{bib-already?}{number}{year}{backrefs}
%\end{verbatim}
%
%    \begin{macrocode}
\def\ObsoleteCiteOptionWarning{%
  \amsrefs@warning{%
    The form \string\cite{...}*{...} is recommended\MessageBreak
    instead of \string\cite[...]{...}}%
  \global\let\ObsoleteCiteOptionWarning\@empty
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\ar@hyperlink#1{\hyper@@link [cite]{}{cite.\CurrentCite}{#1}}
%    \end{macrocode}
%
%    \begin{macrocode}
%% #1<-\b@foo #2<-star-optional-arg
\def\cite@cj#1#2{%
  \leavevmode
  \begingroup
  \cite@cb#1% write info to aux file
  \@citeleft
  \ar@hyperlink{#1}%
  \@ifnotempty{#2}{\citemid{#2}}%
  \citeright \endgroup
  \ignorespaces % when called inside \citelist, a space might follow
  \cite@endgroup
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\extr@cite{\@xp\@gobblethree\string}
%    \end{macrocode}
%
%    This is a copy of the standard warning from \cs{@citex}.
%    \begin{macrocode}
\def\UndefinedCiteWarning{%
  \@latex@warning{%
    Citation `\CurrentCite ' on page \thepage\space undefined}%
}
%    \end{macrocode}
%
%    Here is the original code for \cs{nocite}. [mjd,2001-12-04]
%    \begin{macrocode}
%\def\nocite#1{\@bsphack
%  \@for\@citeb:=#1\do{%
%    \edef\@citeb{\expandafter\@firstofone\@citeb}%
%    \if@filesw\immediate\write\@auxout{\string\citation{\@citeb}}\fi
%    \@ifundefined{b@\@citeb}{\G@refundefinedtrue
%        \@latex@warning{Citation `\@citeb' undefined}}{}}%
%  \@esphack}
%\expandafter\let\csname b@*\endcsname\@empty
%    \end{macrocode}
%
%    For this special case, we want the entry to be defined but not
%    included in the \cs{all@cites} list.
%    \begin{macrocode}
\@namedef{b@*}{\citesel 11{*}{*}{*}}
%    \end{macrocode}
%
%    \begin{macrocode}
\renewcommand{\nocite}[1]{\othercites{#1}}
%    \end{macrocode}
%
%
%    \begin{macrocode}
\newcommand{\othercites}[1]{%
  \cite@begingroup \let\cite@endgroup\@empty
  \def\citelist{\othercitelist}%
  \cites{#1}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newcommand{\othercitelist}[1]{%
  \cite@begingroup \let\cite@endgroup\@empty
  \cites@init \let\citeleft\relax \let\citeright\ignorespaces
  \def\InnerCite{\OtherCite}%
  \def\cite@cj ##1##2{%
    \begingroup \@xp\citesel##1%
    \cite@cb ##1\endgroup \ignorespaces \cite@endgroup
  }%
  #1\relax
  \endgroup
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\citesel@other#1#2#3#4#5#6{}
\def\OtherCite#1{\cite@a\citesel@other{#1}{}{}}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@cb#1{%
  \if@filesw
    \immediate\write\@auxout{\string\citation{\CurrentCite}}%
  \fi
%    \end{macrocode}
%    Define \cs{citesel} to make \cs{b@whatever} update itself and
%    add itself to the \cs{all@cites} list if it's not already there.
%    \begin{macrocode}
  \begingroup \let\citesel\citesel@update #1#1\endgroup
}
%    \end{macrocode}
%
%    When processing the aux file at begin-document, this is what
%    \cs{bibcite} will do:
%    \begin{macrocode}
\def\amsRefs@bibcite#1{\@xp\bibcite@a\csname b@#1\endcsname}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\amsRefs#1{\csname amsRefs@\@xp\@gobble\string#1\endcsname}
\AtBeginDocument{%
  \if@filesw
    \immediate\write\@mainaux{%
      \string\providecommand\string\amsRefs{}%
    }%
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
%% #1<-\b@foo #2<-{number}{} or {author}{year} info
\def\bibcite@a#1#2{%
%    \end{macrocode}
%    Most of the time arg 1 will already be defined, by an earlier
%    \cs{citedest} command in the aux file. Then we just need to change
%    the number.
%    \begin{macrocode}
  \ifx\relax#1\gdef#1{\citesel 00#2{}}%
  \else
    \begingroup
    \@xp\bibcite@b\@xp#1#1{#2}%
    \endgroup
  \fi
}
%    \end{macrocode}
%
%    \begin{macrocode}
% #1       #2         #3#4 #5  #6    #7         #8
% \b@foo ->\citesel 00{num}{year}{backrefs} {{...}{...}}
\def\bibcite@b#1#2#3#4#5#6#7#8{\gdef#1{\citesel#3#4#8{#7}}}
%    \end{macrocode}
%
%    The \cs{citedest} command goes into the aux file to provide
%     back-reference support.
%    \begin{macrocode}
\newcommand{\citedest}[1]{\@xp\cite@dest\csname b@#1\endcsname}
\def\cite@dest#1{%
  \ifx\relax#1\gdef#1{\citesel 00{}{}{}}\fi
  \@xp\cite@dest@b\@xp#1#1}
%    \end{macrocode}
%
%    \begin{macrocode}
% #1       #2         #3#4 #5  #6    #7         #8
% \b@foo ->\citesel 00{num}{year}{backrefs} {more-backref-info}
\def\cite@dest@b#1#2#3#4#5#6#7#8{%
  \@ifempty{#7}{%
    \def#1{\citesel #3#4{#5}{#6}{{#8}}}%
  }{%
    \gdef#1{\citesel #3#4{#5}{#6}{#7,{#8}}}%
  }%
}
%    \end{macrocode}
%
%    Use the hyperref methods instead of this [mjd,2001-12-28]
%    \begin{macrocode}
\def\BackRefData{section={\thesection},page={\thepage}}
\let\BackRefData\@empty
%    \end{macrocode}
%
%    \begin{macrocode}
\let\all@cites\@empty
%    \end{macrocode}
%
%    \begin{macrocode}
\DeclareRobustCommand{\citelist}{\citelist@sorted}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\NonNumericCiteWarning{%
%%  \amsrefs@warning{%
%%    Unable to confirm that cite keys are numeric: not sorting%
%%  }%
}
%    \end{macrocode}
%
%\begin{verbatim}
%\citelist@sorted{\cite{aaa} \cite{bbb}*{Theorem 4.9} \cite{ccc}}
%-> \citeleft
%   \cite@cj\b@aaa{}\cite@cj\b@bbb{Theorem 4.9}\cite@cj\b@ccc{}
%   \citeright
%\end{verbatim}
%
%    \begin{macrocode}
\def\citelist@sorted#1{%
  \leavevmode
  \citeleft\nopunct % suppress first \citepunct
  \cite@begingroup \let\cite@endgroup\@empty
  \cites@init
  \toks@\@xp{\citepunct}%
  \edef\restore@citepunct{\def\@nx\citepunct{\the\toks@}}%
  \def\citeleft{\@addpunct{\citepunct}}%
  \let\citeright\ignorespaces
  \def\cite{\InnerCite}%
  \ifx\citesel\citesel@number
    \cite@sorted@s #1\cite@sorted@e
  \else \NonNumericCiteWarning #1\relax
  \fi
  \endgroup
  \citeright
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\citelist@unsorted#1{%
  \leavevmode
  \citeleft\nopunct % suppress first \citepunct
  \cite@begingroup \let\cite@endgroup\@empty
  \cites@init
  \def\citeleft{\@addpunct{\citepunct}}%
  \let\citeright\ignorespaces
  \def\cite{\InnerCite}%
  #1\relax
  \endgroup
  \citeright
}
%    \end{macrocode}
%
%    By defining this as TeX's maxint, undefined cites migrate to the
%    end of a sorted list.
%    \begin{macrocode}
\def\CPU@sort#1{2147483647}
%    \end{macrocode}
%
%    \begin{macrocode}
\let\cite@dash\@empty
%    \end{macrocode}
%
%    Need to reiterate the definition of \cs{CurrentCite} because I
%    haven't gotten the grouping complications sorted out yet.
%    [mjd,2001-12-20]
%    \begin{macrocode}
\def\cite@print#1#2{%
  \begingroup
  \let\CitePrintUndefined\CPU@normal
  \edef\CurrentCite{\extr@cite#1}%
  \cite@cjs#1{#2}%
  \BackCite
  \endgroup
  \restore@citepunct
}
%    \end{macrocode}
%
%    \begin{macrocode}
\let\BackCite\@empty
%    \end{macrocode}
%
%    \begin{macrocode}
\@ifundefined{ifBR@verbose}{\let\ifBR@verbose\iffalse \let\fi\fi}{}%
\def\back@cite{%
  \ifBR@verbose
    \PackageInfo{backref}{back cite \string `\CurrentCite\string '}%
  \fi
  \Hy@backout{\CurrentCite}%
}
%    \end{macrocode}
%
%    In an AMS-style bibliography, the backref info might follow the
%    final period of the reference, or it might follow some Math Reviews
%    info, without a period.
%    \begin{macrocode}
\def\print@backrefs#1{\SentenceSpace$\uparrow$\csname br@#1\endcsname}
\let\PrintBackRefs\@gobble
%    \end{macrocode}
%
%    \begin{macrocode}
\def\load@backrefs{\@starttoc{brf}{}}
\def\LoadBackRefs{}
%    \end{macrocode}
%
%    \begin{macrocode}
\IfOption{backrefs}{%
  \let\PrintBackRefs\print@backrefs
  \let\LoadBackRefs\load@backrefs
  \let\BackCite\back@cite
  \AtBeginDocument{%
    \@ifundefined{Hy@backout}{%
\amsrefs@warning{backref option requires hyperref package}%
      \let\BackCite\@empty
    }{}%
  }
}{%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@sorted@s{%
  \begingroup
    \let\CitePrintUndefined\CPU@sort
    \let\cite@cjs\cite@cj \let\cite@cj\cite@compress
    \begingroup
    \cite@toks\@emptytoks
    \let\cite@cj\cite@sort@prep
    \ignorespaces
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@sort@prep#1#2{%
  \cite@toks\@xp{\the\cite@toks \cite@cj#1{#2}}%
  \ignorespaces
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@sorted@e{%
  \edef\sort@cite@temp{\the\cite@toks}%
  \let\cite@cj\cite@sorted
  \sort@cite@temp
  \edef\7{\endgroup\the\cite@toks}\7\last@cite \endgroup
}
%    \end{macrocode}
%
%    \begin{macrocode}
\newtoks\cite@toks
%    \end{macrocode}
%
%    \begin{macrocode}
% #1 \b@foo #2 empty?
\def\cite@sorted#1#2{%
  \afterassignment\@nilgobble
  \@tempcnta 0#1\relax\@nil % highest number so far
  \cite@toks{\cite@cj#1{#2}}%
  \edef\sort@cite@temp{\the\cite@toks}%
  \let\cite@cj\sort@cite@b
  \ignorespaces
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\sort@cite@b#1#2{%
  \afterassignment\@nilgobble
  \@tempcntb 0#1\relax\@nil
  \ifnum\@tempcntb>\@tempcnta
    \cite@toks\@xp{\the\cite@toks \cite@cj#1{#2}}%
    \@tempcnta\@tempcntb
  \else
    \let\cite@cj\sort@cite@c \cite@toks\@emptytoks
    \def\@tempb{\cite@toks\@xp{\the\cite@toks \cite@cj#1{#2}}}%
    \sort@cite@temp \@tempb \let\cite@cj\sort@cite@b
  \fi
  \edef\sort@cite@temp{\the\cite@toks}%
  \ignorespaces
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\sort@cite@c#1#2{%
  \ifnum\@tempcntb<0#1\relax
    \@tempb \let\@tempb\@empty
  \fi
  \cite@toks\@xp{\the\cite@toks \cite@cj#1{#2}}%
  \ignorespaces
}
%    \end{macrocode}
%
%    When the time comes to apply compression, we have at our disposal a
%    list of internal cite calls that looks like this:
%\begin{verbatim}
%\cite@cj\b@xxx{}\cite@cj\b@yyy{}...
%\end{verbatim}
%    Now we bind \cs{cite@cj} to a suitable function and execute the
%    list. For compression purposes, we just need to compare the
%    previous number with the next number and keep track of the current
%    state.
% \begin{enumerate}
% \item[0.] First number. Just print it, set the state to 1, and set
%    last-cite-number to current-number.
%
% \item[1.] If current-number = last-cite-number + 1, then the current
%    number might be a middle number in a range. Save it as "last-cite";
%    set last-cite-number to current-number;
%    set the state to 2.
%
% \item[2a.] If current-number = last-cite-number + 1, we are definitely
%    now in the middle (or last number) of a range. Define
%    \cs{cite@dash} to print a dash; save current-number as "last-cite";
%    set last-cite-number to current-number; state remains 2.
%
% \item[2b.] If current-number is NOT equal to last-cite-number + 1,
%    then the previous number was either the second of two, or the end
%    of a three-plus range. Print \cs{cite@dash} (which in the first
%    case will just be empty), then print last@cite. Then do everything
%    for state 0, and set the state to 1, because we might be at the
%    beginning of a new range.
%
% \item[end.] Print last-cite. If it happens that there was only a
%    single cite, last-cite will be empty and nothing will happen.
%    Otherwise this will finish off the current range or sequence of 2.
%
% \end{enumerate}
%
%    \begin{macrocode}
\let\last@cite\@empty
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@compress#1#2{%
  \cite@print#1{#2}%
  \@ifempty{#2}{%
    \@tempcnta 0#1\relax \let\cite@cj\cite@compress@a
  }{%
    \let\citepunct\citeAltPunct
  }%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@dash@cb{%
  \begingroup
  \def\cite@print##1##2{\edef\CurrentCite{\extr@cite##1}\cite@cb##1}%
  \last@cite
  \endgroup
  \let\cite@dash\cite@dash@cb
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@compress@a#1#2{%
  \@ifempty{#2}{%
    \advance\@tempcnta\@ne \@tempcntb 0#1\relax
    \ifnum\@tempcnta=\@tempcntb
      \def\cite@dash{\textendash \nopunct \cite@dash@cb}%
      \def\last@cite{\cite@print#1{#2}}%
      \let\cite@cj\cite@compress@b
    \else
      \cite@print#1{}\@tempcnta\@tempcntb
    \fi
  }{%
    \cite@print#1{#2}\let\cite@cj\cite@compress
    \let\citepunct\citeAltPunct
  }%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\cite@compress@b#1#2{%
  \@ifempty{#2}{%
    \advance\@tempcnta\@ne \@tempcntb 0#1\relax
    \ifnum\@tempcnta=\@tempcntb
      \cite@dash
      \def\last@cite{\cite@print#1{#2}}%
    \else
      \last@cite \let\last@cite\@empty
      \cite@print#1{#2}\@tempcnta\@tempcntb
      \let\cite@cj\cite@compress@a
    \fi
  }{%
    \cite@print#1{#2}\let\cite@cj\cite@compress
    \let\citepunct\citeAltPunct
  }%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\citeAltPunct{;\ }
%    \end{macrocode}
%
%    \begin{macrocode}
\IfOption{non-sorted-cites}{%
  \DeclareRobustCommand{\citelist}{\citelist@unsorted}%
}{}
\IfOption{alphabetic}{%
  \DeclareRobustCommand{\citelist}{\citelist@unsorted}%
}{}
%    \end{macrocode}
%
%    \begin{macrocode}
\DeclareRobustCommand{\ycite}[1]{%
  \star@{\cite@a\citesel@year{#1}}{}%
}
\DeclareRobustCommand{\ycites}{%
  \amsrefs@warning{Not implemented yet: \protect\ycites}%
  \cites
}
\DeclareRobustCommand{\ocite}[1]{%
  \star@{\cite@a\citesel@object{#1}}{}%
}
\DeclareRobustCommand{\ocites}{%
  \amsrefs@warning{Not implemented yet: \protect\ocites}%
  \cites
}
%    \end{macrocode}
%
%    \begin{macrocode}
\DeclareRobustCommand{\citeauthor}[1]{%
  \star@{\cite@a\citesel@author{#1}}{}%
}
\DeclareRobustCommand{\citeauthory}[1]{%
  \citeauthor{#1} \ycite{#1}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\DeclareRobustCommand{\fullcite}{%
  \amsrefs@warning{Not implemented yet: \protect\fullcite}%
  \cite
}
\DeclareRobustCommand{\fullocite}{%
  \amsrefs@warning{Not implemented yet: \protect\fullocite}%
  \ocite
}
%    \end{macrocode}
%
%    \begin{macrocode}
\IfOption{author-year}{%
  \let\citesel\citesel@authoryear
  \def\citeleft{(}\def\citeright{)}%
  \def\@citeleft{\ifx\citesel\citesel@object\else\citeleft\fi}%
  \def\citepunct{; }%
  \let\BibLabel\@empty
}{%
  \let\ycite\cite \let\ocite\cite
  \let\ycites\cites \let\ocites\cites
}
%    \end{macrocode}
%
%    Turn off hyper stuff if the hyperref package does not seem to be
%    loaded.
%    \begin{macrocode}
\AtBeginDocument{%
  \@ifundefined{hyper@@link}{%
    \let\ar@hyperlink\@firstofone
    \let\hyper@anchor@thebib\@firstofone
    \let\hyper@anchorstart\@gobble \let\hyper@anchorend\relax
    \let\BackCite\@empty
  }{}%
}
%    \end{macrocode}
%
%    \begin{macrocode}
\IfOption{jpa}{%
  \amsrefs@warning{The 'jpa' option is obsolete}%
  \typeout{Trying \string\usepackage{amsjpa} instead ...}%
  \RequirePackage{amsjpa}[2000/02/02]
}{}
%    \end{macrocode}
%
%    Here ends the \pkg{amsrefs} package, unless the author-year option
%    is in effect; then we want to use some different bibspecs.
%    \begin{macrocode}
\IfOption{author-year}{}{\endinput}
\renewcommand{\AuthorList}[1]{\PrintNames{author}{}{}{#1}}%
%    \end{macrocode}
%
%    Invert the first author's name.
%    \begin{macrocode}
\BibSpec{firstauthor}{
 +{}{}{surname}
 +{,}{ }{given}
 +{,}{ }{jr}
}
%    \end{macrocode}
%    No parens around the year.
%    \begin{macrocode}
\renewcommand{\PrintDate}[1]{\print@date}
\def\print@date{%
  \IfEmptyBibField{date}%
    {\IfEmptyBibField{year}{\BibField{status}}{\bib@year}}%
    {\bib@year}%
}
\BibSpec{article}{%
  +{}{\PrintAuthors}  {author}
  +{.}{ \PrintDate}   {date}
  +{.}{ \textit}       {title}
  +{.}{ }             {part}
  +{,}{ }             {journal}
%  +{,}{ ISSN }        {ISSN}
  +{} { \textbf}      {volume}
%    \end{macrocode}
%    The date form is tricky depending on presence or absence of DOI.
%    \begin{macrocode}
  +{,}{ \PrintDOI}    {doi}
  +{,}{ \eprint}      {eprint}
  +{} { \parenthesize}{status}
  +{,}{ \eprintpages} {pages}
  +{,}{ }             {note}
  +{.}{}              {transition}
  +{} {\SentenceSpace \ReviewList}   {review}
  +{}{ \parenthesize}{language}
}
\BibSpec{book}{%
  +{}{\PrintAuthors}{author}
  +{}{\PrintEditorsA}{editor}
  +{.}{ \PrintDate}{date}
  +{.}{ \textit}{title}
  +{,}{ }{type}
  +{,}{ \EnglishInitialCaps}{booktitle}
  +{,}{ \PrintEdition}{edition}
  +{,}{ }{series}
  +{,}{ vol.~}{volume}
  +{,}{ part~}{part}
  +{,}{ }{publisher}
  +{,}{ }{organization}
  +{,}{ }{place}
  +{,}{ \ISBNList}{isbn}
  +{}{ \parenthesize}{status}
  +{,}{ }{note}
  +{.}{}{transition}
  +{}{\SentenceSpace \ReviewList}{review}
  +{}{ \parenthesize}{language}
}
\BibSpec{collection.article}{%
  +{}{\PrintAuthors}{author}
  +{.}{ \PrintDate}{date}
  +{.}{ \textit}{title}
  +{,}{ }{type}
  +{,}{ \EnglishInitialCaps}{booktitle}
  +{,}{ \PrintEdition}{edition}
  +{}{ \PrintEditorsB}{editor}
  +{,}{ Proc.\ }{conference}
  +{,}{ }{series}
  +{,}{ vol.~}{volume}
  +{,}{ part~}{part}
  +{,}{ }{publisher}
  +{,}{ }{organization}
  +{,}{ }{place}
  +{,}{ \DashPages}{pages}
  +{,}{ \ISBNList}{isbn}
  +{}{ \parenthesize}{status}
  +{,}{ }{note}
  +{.}{}{transition}
  +{}{\SentenceSpace \ReviewList}{review}
  +{}{ \parenthesize}{language}
}
\BibSpec{report}{%
  +{}{\PrintAuthors}{author}
  +{}{\PrintEditorsA}{editor}
  +{.}{ \PrintDate}{date}
  +{.}{ \textit}{title}
  +{,}{ }{type}
  +{,}{ \EnglishInitialCaps}{booktitle}
  +{,}{ Technical Report }{number}
  +{,}{ }{series}
  +{,}{ vol.~}{volume}
  +{,}{ part~}{part}
  +{,}{ }{publisher}
  +{,}{ }{organization}
  +{,}{ }{institution}
  +{,}{ }{place}
  +{,}{ \ISBNList}{isbn}
  +{,}{ \eprint}      {eprint}
  +{}{ \parenthesize}{status}
  +{,}{ }{note}
  +{.}{}{transition}
  +{}{\SentenceSpace \ReviewList}{review}
  +{}{ \parenthesize}{language}
}
\BibSpec{thesis}{%
  +{}{\PrintAuthors}{author}
  +{}{\PrintEditorsA}{editor}
  +{.}{ \PrintDate}{date}
  +{.}{ \textit}{title}
  +{,}{ \PrintThesisType}{type}
  +{,}{ part~}{part}
  +{,}{ }{organization}
  +{,}{ }{institution}
  +{,}{ }{place}
  +{}{ \parenthesize}{status}
  +{,}{ }{note}
  +{.}{}{transition}
  +{}{\SentenceSpace \ReviewList}{review}
  +{}{ \parenthesize}{language}
}
%</pkg>
%    \end{macrocode}
%
%    The usual \cs{endinput} to ensure that random garbage at the end of
%    the file doesn't get copied by \fn{docstrip}.
%    \begin{macrocode}
\endinput
%    \end{macrocode}
%
% \CheckSum{3542}
% \Finale