% \iffalse 
%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%
%%                          ogonek package
%%
%% Purpose: to provide best possible ogonek in every situation
%%
%% The copyright holder of this file is Janusz S. Bie\'n
%% (JSBien@plearn.edu.pl)
%%
%% The code consists of parts written by J. S. Bie\'n, L. Holenderski,
%% D. H\"ubel and Igor Moo, who finally assembled this file. Any bugs
%% can (and should) be attributed to him.
%%
%% Note: you are not allowed to modify this file.
%% Note: I don't make any changes in T1 definitions so aAeE(iIuU)
%%       checking does not work there.
%%
% \fi
% \iffalse
% Since I hate the presence of section `Printing the documentation'
% in every DOCed package my driver is put into a metacomment.
% To print documentation LaTeX this file. If you need to change
% the driver run DOCSTRIP and ask it to make ogonek.drv by
% stripping ogonek.dtx with option `driver'. Do *not* modify this file.
%<*driver>
\documentclass{ltxdoc}
\DisableCrossrefs
% \OnlyDescription
% \EnableCrossrefs
% \CodelinesIndex
\begin{document}
   \DocInput{ogonek.dtx}
\end{document}
%</driver>
% \fi
\def\fileversion{v0.51}
\def\filedate{95/07/17}
\def\docdate {94/12/21}
% \CheckSum{236}
% \CharacterTable
%  {Upper-case    \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
%   Lower-case    \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
%   Digits        \0\1\2\3\4\5\6\7\8\9
%   Exclamation   \!     Double quote  \"     Hash (number) \#
%   Dollar        \$     Percent       \%     Ampersand     \&
%   Acute accent  \'     Left paren    \(     Right paren   \)
%   Asterisk      \*     Plus          \+     Comma         \,
%   Minus         \-     Point         \.     Solidus       \/
%   Colon         \:     Semicolon     \;     Less than     \<
%   Equals        \=     Greater than  \>     Question mark \?
%   Commercial at \@     Left bracket  \[     Backslash     \\
%   Right bracket \]     Circumflex    \^     Underscore    \_
%   Grave accent  \`     Left brace    \{     Vertical bar  \|
%   Right brace   \}     Tilde         \~}
%
% \def\ogon/{\texttt{ogonek}}
%
% \title{The \ogon/ package\thanks{Version \fileversion\ dated \filedate.}}
% \author{Janusz Stanis\l aw Bie\'n}
% \date{\docdate}
%
% \maketitle
% \begin{abstract}
% This \LaTeXe\ package provides a command to typeset letters with the
% ogonek diacritic mark; they are used in Polish and Lithuanian. The
% command is named |\k| in accordance with the recommendation of the
% Technical Working Group on Multiple Language Coordination of the \TeX\
% Users Group. The principal purpose of the command is to provide the
% high quality ogonek with CM fonts, although for Polish the best
% results are obtained with the special Polish PL fonts; the command
% can be also used with DC fonts.
% \end{abstract}
%
% \section{Introduction} 
% The ogonek diacritic mark (|\k|) is absent in the original Computer
% Modern font (\cite{CM}), probably because it was not needed for
% Donald Knuth's Art of Computer Programming. The
% ogonek was included in the extended \TeX\ layout agreed in 1990 at the
% \TeX\ conference in Cork in Ireland and therefore often called simply
% the Cork layout; however, there was still no standard command to
% typeset it.  This was remedied in 1992, when the \TeX\ Users Group
% Technical Working Group on Multiple Language Coordination
% WG-92-03\footnote {The group was described in \cite{Council}}
% recommended a set of \TeX\ conventions concerning languages (cf.
% \cite{Conv}).  In particular, the command names were proposed for
% typesetting letters and accents introduced in the extended layout;
% the command |\k| was assigned to the ogonek and the name justified as
% the last letter of the word {\it ogonek}\footnote {Actually J\"org
% Knappen wrote in \cite{summary} that \texttt{\string\k} stands also
% for the first letter
% of the Scandinavian {kvist}. It can be viewed also as the first
% letter of the German word {Krummhaken}} 
% 
% In \cite{Conv} WG-92-3 proposed also a set of two-letter names for
% the language-switching macros. We use the two names from this list
% (but without the preceeding backslash) as the option names in our
% package: PL for Polish and LT for Lithuanian.
% 
% The lack of a standard way to typeset ogonek with Computer Modern
% fonts and its predecessors (including AM, i.e. Almost Computern
% Modern fonts) was from the very beginning a very serious obstacle for
% high quality typesetting of Polish texts. Several various techniques
% were developed independently to circumvent this problem; in the
% present package we use the method developed at the Faculty of
% Mathematics, Informatics and Mechanics of the Warsaw University and
% used in \LaTeX\ styles plfonts and plhb.\footnote {Thanks to the
% contribution of Piotr Filip Sawicki, the support of these styles is a
% standard feature of AUC \TeX, a sophisticated (La)\TeX\ environment for
% Emacs, since the release 9.0 of May 1994.} 
% 
% The primary problem was to find a CM character which bears sufficient
% ressemblance to ogonek. Several characters (including e.g. comma)
% were tried till 1988, when Jerzy Ryll suggested to use |\lhook| (left
% hook) symbol available in Plain \TeX\ as a part of the |\hookrightarrow|
% ($\hookrightarrow$) symbol; this is the character \texttt{'54} in math italics
% fonts. Ryll's idea was described in the note \cite{TeXline} and
% Janusz S. Bie\'n's \texttt{pl.sty} style using this technique was sent to the
% \TeX line editor to be included in the Aston \TeX\ archive;
% unfortunately, it seems that it never managed to get there. The
% idea was also presented in a paper written in Polish in 1988, which
% however appeared much later (\cite{cttex90}).
% 
% The remaining problem was to achieve proper positioning of the left
% hook character with the appropriate letters for every fonts size and
% shape; as ogonek accompanies such different letters as a, A, e and E,
% this was not an easy task. At first it was done simply by hand, as in
% Janusz S.~Bie\'n's \texttt{plfonts.tex} file loaded during the \LaTeX\ format
% generation.  The credit for solving this problem is due to Leszek
% Holenderski, who in 1989 created his \texttt{plfonts.sty}, which patched the
% standard \LaTeX\ font switching mechanism with the code for adjusting
% the placing of ogonek. We use his code here without any substantial
% changes.\footnote {Bie\'n's notes say that he started to use Ryll's
% technique on 22nd June 1988 and created a version of Holenderski's
% style on 17th October 1999 (the version was called \texttt{plhb.sty}, 
% where \texttt{hb}
% stands for {Holenderski's style as modified by Bie\'n} and \texttt{pl} stands
% both for Polish and the earlier \texttt{pl} style; it used a different input
% convention than the original Holenderski's style)}
% 
% In the extended \TeX\ layout used at present practically only in
% Norbert Schwarz's DC fonts (cf. \cite{DC1}, \cite{DC2}) but envisaged
% as the future \TeX\ standard and therefore recommended for \LaTeXe\
% users the slots are assigned for both Polish letters with ogonek and
% the ogonek itself; typeseting all Polish letters and some Lithuanian
% ones causes therefore no problem and requires only referencing the
% appropriate characters; the remaining Lithuanian characters have to
% be typeset using by the composition of the appropriate chcaracters
% (the |\accent| primitive cannot be used for this purpose because it
% placed the accent {over} the letter).
% 
% The primary problem with the extended \TeX\ font layout was (at to some
% measure still is) its incompatibility with the standard CM layouts,
% which for many users makes the migration to the new layout
% prohibitively difficult. For many applications a good solution was a
% mixed layout, with the lower part (character codes from 0 to 127)
% fully compatible with CM fonts and the higher part more or less
% compatible with the Cork layout. We will call this layout
% {Cork-extended CM layout}\footnote {It seems to be little known that
% the layout should be coded in the TFM and PK files by means of the
% Metafont \textbf{font\_coding\_scheme} command; to the best of our knowledge,
% the only program which takes advantage of this fact is Eberhard
% Mattes' dvispell}.
% 
% The PL fonts, developed by Bogus{\l}aw Jackowski and Marek Ry\'cko with
% some advice of a professional typographer Roman Tomaszewski and
% included in the MeX distribution\footnote{available e.g. from
% Comprehensive \TeX\ Archive Network in the directory
% tex-archive/languages/polish }, are a special
% case of Cork-like extended Computer Modern fonts---in the higher
% part they contain Polish letters with ogonek placed in the same slots
% as in the Cork layout; however, they contain also the Polish double
% opening quote moved from its Cork position in the lower part to the
% higher part of the font. This layout can be called PL-extended CM
% fonts\footnote {At present (i.e. in all MeX releases including 1.5) a
% PL font have the \textbf{font\_coding\_scheme} identical with the CM font it is
% compatible with. For example, both plr10 and cmr10 have the coding
% scheme {TeX text}, pltt10 and cmtt10 {TeX typewriter text} etc.
% Dvispell users would appreciate very much if the PL fonts were
% distinguishable from CM fonts by the coding scheme field, which can
% be asigned such values as {PL-extended TeX text}, {PL-extended TeX
% typewriter text} etc.}
% 
% The PL fonts provide the best quality for Polish texts; however, for
% those Lithuanian letters with ogonek which do not coincide with Polish
% ones it is necessary to use the same technique as for CM fonts. In
% consequence, for Lithuanian texts the use of DC fonts is probably an
% optimal solution.
% 
% 
% \section{Usage} 
% The package is loaded in the standard way with the
% |\usepackage{ogonek}| command.
% 
% As the fonts called by us the PL-extended CM fonts are not widely
% used, they do not have also a generally accepted symbol for their
% layout.  Mariusz Olko in his preliminary version of polski package
% referes to them as OT1P, while W{\l}odzimierz Bzyl in his LaMeXe uses
% the OT4 symbol.
% In consequence \ogon/ works with the following
% font encodings: 
% OT1 (standard meaning)
% OT1P (PL fonts with Olko's package)
% OT4 (PL fonts with LaMeX2e and later versions of Olko's package)
% T1 (standard meaning)
%
% The package accepts two language options:
%
% \begin{tabular}{cl}
% PL & only Polish letters with ogonek\\
% LT & Lithuanian letters --- which subsume the Polish ones --- with ogonek
% \end{tabular}
%
% Omitting the language option allows to use any letter with ogonek.
% 
% \section{Hyphenation of words with ogonek accent} 
% The full and correct hyphenation of words with ogonek (and other Polish
% letters) is possible with DC and PL fonts; details to be written later.
% 
% \StopEventually{
% \begin{thebibliography}{8}
% \bibitem{TeXline}
% Janusz S. Bie\'n. Polish Language and {\TeX}. \TeX line 8, January 1989,
% p.~2.
% 
% \bibitem{cttex90} Janusz S. Bie\'n. Co to jest {\TeX}. Available by
% anonymous FTP from ftp.mimuw.edu.pl in pub/users/jsbien/tex as
% cttex90.tar.Z or from LISTSERV@PLEARN.edu.pl as CTTEX90 PACKAGE.
% 
% \bibitem{Cork}
% Michael J. Ferguson. Report on Multilingual Activities.
% TUGboat Vol. 11, No 4, November 1990, pp 514--516
% 
% \bibitem{Council}
% Michael J. Ferguson. The Technical Council.
% {\it \TeX\ and TUG News} Vol. 1, No. 3, November 1992, pp  5--8.  
% 
% \bibitem{Conv}
% Yannis Haralambous. \TeX\ Convention Concerning Languages.
% {\it \TeX\ and TUG News} Vol. 1, No. 4, December 1992, pp  3--10.  
% 
% \bibitem{DC1}
% Yannis Haralambous. DC fonts---questions and answers (I). 
% {\it \TeX\ and TUG News} Vol. 1, No. 4, December 1992, pp 15--17.
% 
% \bibitem{DC2}
% Yannis Haralambous. DC fonts---questions and answers (II). 
% {\it \TeX\ and TUG News} Vol. 2, No. 1, February 1993, pp 10--12.
% 
% \bibitem{summary}
% J\"org Knappen. Summary on Polish (La)\TeX. Info-TeX@SHSU.edu and
% TeX-Euro@vm.urz.uni-heidelberg.de, 20th Novemebr 1992.
% 
% \bibitem{CM}
% Donald E. Knuth. Computer Modern Typefaces. Computers and Typesetting
% Vol. E. Addison-Wesley, Reading, Massachusetts 1986.
% \end{thebibliography}
% }
%
% 
% \section{Implementation}
% {\small Beware: comments in this section were written by Igor Moo.}
% \subsection{Identification}
% We start the code with standard identification of the package
%    \begin{macrocode}
%<*style>
\NeedsTeXFormat{LaTeX2e}[95/06/01]

\ProvidesPackage{ogonek}[\filedate\space\fileversion\space
		     Provides macro `\string\k' for ogonek]
%    \end{macrocode}
% \subsection{Processing options}
% \subsubsection{Encoding selection options}
% \begin{macro}{\ogonek@obsolete}
% In previous versions of \ogon/ options were present for selection of
% font encoding(s) used in a document. Now they are no longer needed since
% we try to guess what encodings are really used.
%    \begin{macrocode}
\newcommand\ogonek@obsolete[1]{%
  \PackageWarningNoLine{ogonek}{Option #1 is now obsolete \MessageBreak 
     due to dynamic encodings testing}
  }
\DeclareOption{T1}{\ogonek@obsolete{T1}}
\DeclareOption{OT1}{\ogonek@obsolete{OT1}}
\DeclareOption{OT1P}{\ogonek@obsolete{OT1P}}
\DeclareOption{OT4}{\ogonek@obsolete{OT4}}
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Language selection options}
% \begin{macro}{\@testogonekletter}
% Here we define macro that will be used below to test if a ogonked 
% letter is `legal'. Primarily we define it just to gobble it's 
% argument.
%
% If option \texttt{PL} is specified the macro is redefined to accept 
% only Polish ogonked letters. Option \texttt{LT} redefines it to allow
% only Lithuanian letters.
% 
% If both options were specified all aAeEiIuU letters will be allowed 
% since in \LaTeXe\ options are processed in order of
% declaration and LT overwrites PL.
%    \begin{macrocode}
\let\@testogonekletter\@gobble
\DeclareOption{PL}{
  \def\@testogonekletter#1{%
    \ifx a#1\else    \ifx A#1\else
    \ifx e#1\else    \ifx E#1\else
      \PackageWarning{ogonek}%
	{Unusual Polish letter #1 with ogonek}\fi
	\fi \fi \fi
  }
}
\DeclareOption{LT}{
  \def\@testogonekletter#1{%
    \ifx a#1\else    \ifx A#1\else
    \ifx e#1\else    \ifx E#1\else
    \ifx i#1\else    \ifx I#1\else
    \ifx u#1\else    \ifx U#1\else
	\PackageWarning{ogonek}%
       {Unusual Lithuanian letter #1 with ogonek}\fi 
	\fi \fi \fi \fi \fi \fi \fi
    }
}
%    \end{macrocode}
% \end{macro}
% Now we're ready to process the options
%    \begin{macrocode}
\ProcessOptions
%    \end{macrocode}
% \subsection{Positioning of ogonek in old fonts}
% This comes from L.~Holenderski's \texttt{plfonts.sty}.
% Positionig of ogonek for specific letters is tuned for 300dpi
% Computer Modern fonts, but works reasonably well with other
% resolutions. 
% \begin{macro}{\sob}
% Macro |\sob| positioning ogonek under a letter.
%    \begin{macrocode}
\dimendef\pl@left=0 \dimendef\pl@down=1 
\dimendef\pl@right=2 \dimendef\pl@temp=3

% typeset `ogonek' box
\def\sob#1#2#3#4#5{% parameters: letter and fractions hl,ho,vl,vo
\setbox0\hbox{#1}\setbox1\hbox{$_\mathchar'454$}\setbox2\hbox{p}%
\pl@right=#2\wd0 \advance\pl@right by-#3\wd1
\pl@down=#5\ht1 \advance\pl@down by-#4\ht0
\pl@left=\pl@right \advance\pl@left by\wd1
\pl@temp=-\pl@down \advance\pl@temp by\dp2 \dp1=\pl@temp
\leavevmode\kern\pl@right\lower\pl@down\box1\kern-\pl@left #1}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\aob}\begin{macro}{\Aob}\begin{macro}{\eob}
% \begin{macro}{\Eob}\begin{macro}{\iob}\begin{macro}{\Iob}
% \begin{macro}{\uob}\begin{macro}{\Uob}
% Special positioning of ogonek for some letters
%    \begin{macrocode}
\def\aob{\sob a{.66}{.20}{0}{.90}}
\def\Aob{\sob A{.80}{.50}{0}{.90}}
\def\eob{\sob e{.50}{.35}{0}{.93}}
\def\Eob{\sob E{.60}{.35}{0}{.90}}
\def\iob{\sob i{.66}{.20}{0}{.90}}
\def\Iob{\sob I{.80}{.50}{0}{.90}}
\def\uob{\sob u{.66}{.20}{0}{.90}}
\def\Uob{\sob U{.60}{.35}{0}{.90}}
%    \end{macrocode}
% \end{macro}\end{macro}\end{macro}\end{macro}
% \end{macro}\end{macro}\end{macro}\end{macro}
%
% \begin{macro}{\@iIuUogonek}
% \begin{macro}{\@oldfontsogonek}
% Below we define macros producing ogonek in encodings OT4 (OT1P) (this 
% needs special positioning of ogonek only for iIuU since for aAeE we
% have composities) and OT1. This could be done in a more \LaTeX y way
% if we had something like \verb+\DeclareTextComposite+ allowing
% replacement to be macro not a single character. But we haven't.
%    \begin{macrocode}
\def\@iIuUogonek#1{%
	\ifx i#1\iob\else
	\ifx I#1\Iob\else
	\ifx u#1\uob\else
	\ifx U#1\Uob\else\sob {#1}{.50}{.35}{0}{.90}\fi 
	\fi \fi \fi 
	}
\def\@oldfontsogonek#1{%
	\ifx a#1\aob\else
	\ifx A#1\Aob\else
	\ifx e#1\eob\else
	\ifx E#1\Eob\else
	\@iIuUogonek{#1}
	\fi \fi \fi \fi 
   }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \subsection{Testing of encodings used in a document}
% This testing must be carried off when the document begins,
% since only then all used encodings are already known.
% We use |\AtBeginDocument| hook for that purpose.
% This will work unless some package loaded after \ogon/ 
% has an idea to declare encodings `at begin document' (I cannot think 
% of any reason for that).
%
% First my favourite hack for operations on names 
% constructed with |\csname|:
%    \begin{macrocode}
\newcommand\n@me[2]{\expandafter#1\csname#2\endcsname}
%    \end{macrocode}
% You can not only say |\n@me\ifx{T@T1}\sth| but even |\n@me\show{a name}|
% or |\n@me\newcommand{and another}{...}| (sic!).
%
% The testing really starts here. If an encoding \texttt{XXX} is known
% a macro with name |\T@XXX| is defined. In that way we check
% what encodings are in use.
%
% For every encoding found we define |\k| to 
% test if accentee is legal and put appropriate kind of ogonek.
%    \begin{macrocode}
\AtBeginDocument{%
%    \end{macrocode}
% We don't make any changes for T1, since all we need is defined by 
% default.
%    \begin{macrocode}
  \n@me\ifx{T@T1}\relax \else \PackageInfo{ogonek}{T1 is known} \fi
%    \end{macrocode}
% For OT1 encoding we define ogonek as |\@oldfontsogonek|
%    \begin{macrocode}
  \n@me\ifx{T@OT1}\relax 
  \else \PackageInfo{ogonek}{Defining ogonek for OT1} 
    \DeclareTextCommand\k{OT1}[1]{%
       \@testogonekletter{#1}\@oldfontsogonek{#1}}
    \fi
%    \end{macrocode}
% For OT4 or OT1P |\k| won't know how to put ogonek under aAeE,
% but we have composities for that cases.
%
% We are lucky that ogonek is always allowed under aAeE. Otherwise
% we would have to invent a way to incorporate test into composities.
%    \begin{macrocode}
  \n@me\ifx{T@OT4}\relax 
  \else \PackageInfo{ogonek}{Defining ogonek for OT4} 
    \DeclareTextCommand\k{OT4}[1]{%
       \@testogonekletter{#1}\@iIuUogonek{#1}}
    \DeclareTextComposite\k{OT4}{a}{"A1}
    \DeclareTextComposite\k{OT4}{A}{"81}
    \DeclareTextComposite\k{OT4}{e}{"A6}
    \DeclareTextComposite\k{OT4}{E}{"86}
    \fi
  \n@me\ifx{T@OT1P}\relax 
  \else \PackageInfo{ogonek}{Defining ogonek for OT1P} 
    \DeclareTextCommand\k{OT1P}[1]{%
       \@testogonekletter{#1}\@iIuUogonek{#1}}
    \DeclareTextComposite\k{OT1P}{a}{"A1}
    \DeclareTextComposite\k{OT1P}{A}{"81}
    \DeclareTextComposite\k{OT1P}{e}{"A6}
    \DeclareTextComposite\k{OT1P}{E}{"86}
    \fi
  }
%    \end{macrocode}
% And that's all.
%    \begin{macrocode}
\endinput
%</style>
%    \end{macrocode}
%
% \iffalse metaquestion
% this definition intended for T1 encoding differs from
% `official' lxiie version. why?  IM
%\def\@gonekaccent #1{%
%        \setbox\z@\hbox {#1}\ifdim\ht\z@=1ex\accent12 #1%
%       \else {\ooalign{\unhbox\z@\crcr\hidewidth\char12\hidewidth}}\fi}
% \fi
%
% \Finale
% \iffalse
%% 
%% End of file `ogonek.sty'.
% \fi