From 937ac59277b655a7dbdea118ad12087cf7b9ad51 Mon Sep 17 00:00:00 2001 From: Robert Alessi Date: Mon, 7 Mar 2016 23:04:01 +0100 Subject: removed all of old \robustify commands --- arabluatex.dtx | 134 ++++++++++++++++++++++++++++++++++++++++----------------- 1 file changed, 95 insertions(+), 39 deletions(-) (limited to 'arabluatex.dtx') diff --git a/arabluatex.dtx b/arabluatex.dtx index 435e905..d0fe36f 100644 --- a/arabluatex.dtx +++ b/arabluatex.dtx @@ -149,36 +149,95 @@ % \url{http://ctan.org/pkg/arabtex}} Arab\LuaTeX\ is at present % nothing more than a modest piece of software. Hopefully---if I may % say so---it will eventually provide all of its valuable qualities to -% the \LuaLaTeX\ users. \package{arabtex} dates back to 1992. As far -% as I know, it was then the first and only way to typeset Arabic -% texts with \TeX\ and \LaTeX. To achieve this, \package{arabtex} -% provided---and still does---an Arabic font in \emph{Nasḫī} style and -% a macro package that defined its own input notation which was, as -% the author stated, \enquote{both machine, and human, readable, and -% suited for electronic transmission and e-mail -% communication}.\footcite[2]{Lagally2004} Even if the same can be -% said about Unicode, Arab\TeX\ \textsc{ASCII} input notation still -% surpasses Unicode input, in my opinion, when it comes to typesetting -% complex documents, such as scientific documents or critical editions -% where footnotes and other kind of annotations can be particulary -% abundant. It must also be said that most text editors have trouble -% in displaying Arabic script connected with preceding or following -% \LaTeX\ commands: it often happens that commands seem misplaced, not -% to mention punctuation marks, or opening or closing braces, brackets -% or parentheses that are unexpectedly displayed in the wrong -% direction. Of course, some text editors provide ways to get around -% such difficulties by inserting invisible Unicode characters, such as -% LEFT-TO-RIGHT or RIGHT-TO-LEFT MARKS (\texttt{U+200E}, -% \texttt{U+200F}), RTL/LTR \enquote{embed} characters -% (\texttt{U+202B}, \texttt{U+202A}) and RLO/LRO +% the \LuaLaTeX\ users. +% +% \package{arabtex} dates back to 1992. As far as I know, it was then +% the first and only way to typeset Arabic texts with \TeX\ and +% \LaTeX. To achieve this, \package{arabtex} provided---and still +% does---an Arabic font in \emph{Nasḫī} style and a macro package that +% defined its own input notation which was, as the author stated, +% \enquote{both machine, and human, readable, and suited for +% electronic transmission and e-mail +% communication}.\footnote{\textcite[2]{Lagally2004}.} Even if the +% same can be said about Unicode, Arab\TeX\ \textsc{ASCII} input +% notation still surpasses Unicode input, in my opinion, when it comes +% to typesetting complex documents, such as scientific documents or +% critical editions where footnotes and other kind of annotations can +% be particulary abundant. It must also be said that most text editors +% have trouble in displaying Arabic script connected with preceding or +% following \LaTeX\ commands: it often happens that commands seem +% misplaced, not to mention punctuation marks, or opening or closing +% braces, brackets or parentheses that are unexpectedly displayed in +% the wrong direction. Of course, some text editors provide ways to +% get around such difficulties by inserting invisible Unicode +% characters, such as LEFT-TO-RIGHT or RIGHT-TO-LEFT MARKS +% (\texttt{U+200E}, \texttt{U+200F}), RTL/LTR \enquote{embed} +% characters (\texttt{U+202B}, \texttt{U+202A}) and RLO/LRO % \enquote{bidi-override} characters (\texttt{U+202E}, -% \texttt{U+202D}).\footnote{Yudit probably has the best Unicode -% support. See \url{www.yudit.org}.} Nonetheless, it remains that -% inserting all the time these invisible characters in complex -% documents rapidly becomes confusing and cumbersome. +% \texttt{U+202D}).\footnote{Gáspár Sinai's Yudit probably has the +% best Unicode support. See \url{www.yudit.org}.} Nonetheless, it +% remains that inserting all the time these invisible characters in +% complex documents rapidly becomes confusing and cumbersome. +% +% The great advantage of Arab\TeX\ notation is that it is immune from +% all these difficulties, let alone its being clear and +% straightforward. One also must remember that computers are designed +% to process code. Arab\TeX\ notation is a way of encoding Arabic +% language, just as \TeX\ \enquote{mathematics mode} is a way of +% processing code to display mathematics. As such, not only does it +% allow greater control over typographical features, but it also can +% be processed in several different ways: so without going into +% details, depending on one's wishes, Arab\TeX\ input can be full +% vocalized Arabic (\emph{scriptio plena}), vocalized Arabic or +% non-vocalized Arabic (\emph{scriptio defectiva}); it further can be +% transliterated into whichever romanization standard the user may +% choose. +% +% But there may be more to be said on this point, as encoding Arabic +% also naturally encourages the coder to vocalize the texts---without +% compelling him to do so, of course. Accurate coding may even have +% other virtuous effects. For instance, hyphens may be used for tying +% particles or prefixes to words, or to mark inflectional endings, and +% so forth. In other words, accurate coding produces accurate texts +% that can stand to close grammatical scrutiny and to complex textual +% searches as well. +% +% Having that in mind, I started \package{arabluatex}. With the help +% of Lua, it will eventually interact with some other packages yet to +% come to produce from \verb|.tex| source files, in addition to +% printed books, \texttt{TEI xml} compliant critical editions and/or +% lexicons that can be searched, analyzed and correlated in various +% ways. +% +% +%\subsection{\package{arabluatex} is for \LuaLaTeX} +% It goes without saying that \package{arabluatex} requires +% \LuaLaTeX. \TeX\ and \LaTeX\ have \package{arabtex}, and \XeLaTeX\ +% has \package{arabxetex}. Both of them are much more advanced than +% \package{arabluatex}, as they can process a number of different +% languages,\footnote{To date, both packages support Arabic, Maghribi, +% Urdu, Pashto, Sindhi, Kashmiri, Uighuric and Old Malay; in addition +% to these, \package{arabtex} also has a Hebrew mode, including +% Judeo-Arabic and Yiddish.} whereas \package{arabluatex} can process +% only Arabic for the time being. More languages will be included in +% future releases of \package{arabluatex}. +% +% In comparison to \package{arabxetex}, \package{arabluatex} works in +% a very different way. The former relies on the +% \href{http://scripts.sil.org/TECkitIntro}{\texttt{TECkit}} engine +% which converts Arab\TeX\ input on the fly into Unicode Arabic +% script, whereas the latter passes Arab\TeX\ input on to a set of Lua +% functions. At first, \LaTeX\ commands are taken care of in different +% ways: some, as \cs{emph}, \cs{textbf} and the like are expected to +% have Arabic text as arguments, while others, as \cs{LR}, for +% \enquote{left-to-right text}, are not. Then, once what is Arabic is +% carefully separated form what is not, it is processed by other Lua +% functions which rely on different sets of correpondence tables to do +% the actual conversion in accordance with one's wishes. Finally, Lua +% returns to \TeX\ the converted strings---which may in turn contain +% some other Arab\TeX\ input yet to be processed---for further +% processing. % -% -% % \section{Usage} % \newcommand{\dmg}[1]{% % \SetTranslitConvention{dmg}% @@ -192,13 +251,16 @@ % & Arab\TeX\ notation \\ % & \texttt{dmg} & \texttt{loc} & \\ \midrule % \endhead \footnotetext{See below **} -% \arb[novoc]{a} & \dmg{A} & \loc{A} & \verb|a| \\ +% \arb[novoc]{a} & \dmg{A} & \loc{A} & \verb|A| \\ % \arb[novoc]{b} & \dmg{b} & \loc{b} & \verb|b| \\ % \arb[novoc]{t} & \dmg{t} & \loc{t} & \verb|t| \\ % \arb[novoc]{_t} & \dmg{_t} & \loc{_t} & \verb|_t| \\ % \arb[novoc]{j} & \dmg{j} & \loc{j} & \verb|^g| or \verb|j| \\ % \arb[novoc]{.h} & \dmg{.h} & \loc{.h} & \verb|.h| \\ % \arb[novoc]{x} & \dmg{x} & \loc{x} & \verb|_h| or \verb|x|\\ +% \arb[novoc]{d} & \dmg{d} & \loc{d} & \verb|d| \\ +% \arb[novoc]{_d} & \dmg{_d} & \loc{_d} & \verb|_d| \\ +% \arb[novoc]{r} & \dmg{r} & \loc{r} & \verb|r| \\ % \bottomrule % \caption{Standard notation for Arabic} % \end{longtable} @@ -221,10 +283,10 @@ % \package{arabluatex} requires \LuaLaTeX\ of course. % \begin{macrocode} \ifluatex\else - \PackageError{arabluatex}{lualatex needed}{% - Package `arabluatex' needs LuaTeX.\MessageBreak - So you should use `lualatex' to process you document!\MessageBreak - See documentation of `arabluatex' for further information.}% + \PackageError{arabluatex}{lualatex needed}{% + Package `arabluatex' needs LuaTeX.\MessageBreak + So you should use `lualatex' to process your document!\MessageBreak + See documentation of `arabluatex' for further information.}% \expandafter\expandafter\expandafter\csname endinput\endcsname \fi % \end{macrocode} @@ -319,17 +381,11 @@ \luastringO{\al@trans@convention}))}\egroup% \else \fi\fi\fi\fi}[\par] -% \robustify{\textbf} -% \robustify{\emph} -% \robustify{\footnote} -% \robustify{\marginpar} \DeclareDocumentCommand{\Marginpar}{m}{\marginpar{\textdir TLT #1}} \RenewDocumentCommand{\thefootnote}{}{\bgroup\textdir TLT\arabic{footnote}\egroup} \DeclareDocumentCommand{\LRfootnote}{m}{\bgroup\pardir TLT\LR{\footnote{#1}}\egroup} -% \AtBeginDocument{\ifdef{\abjad}{\robustify{\abjad}}{}} -% \AtBeginDocument{\ifdef{\Footnote}{\robustify{\Footnote}}{}} \endinput % \end{macrocode} % -- cgit v1.2.3