arabtex2utf: done programming functions exporting running paragraphs or Arabic verses

2 files changed, 165 insertions, 37 deletions

diff --git a/arabluatex.dtx b/arabluatex.dtx
index 3dcfc1f..d54290a 100644
--- a/arabluatex.dtx
+++ b/arabluatex.dtx
@@ -3481,9 +3481,15 @@ muhaddamaTaN mi'_danatu-hu}: \arb[trans]{ra'aytu
 %   'l-\uc{`AdAt-i}}}.\footnote{This is odd in Arabic script, but
 %   using such features as \cs{emph} or \cs{textbf} is a matter of
 %   personal taste.}
-%
-%   |\arb{\abjad{45} \framebox[1in][s]{kitAbu-hu fI 'l-`AdAti}}|\\
-%   \arb{\abjad{45} \framebox[1in][s]{kitAbu-hu fI 'l-`AdAti}}
+%   \NewDocumentCommand{\rlframebox}{o o m}{
+%    \IfNoValueTF{#2}{\IfNoValueTF{#1}{
+%    \framebox{\setRL#3}}{\framebox[#1]{\setRL#3}}
+% }{\framebox[#1][#2]{\setRL#3}}}
+%
+%   |\arb{\abjad{45} \rlframebox[1in][s]{kitAbu-hu fI 'l-`AdAti}}|\\
+%   \arb{\abjad{45} \rlframebox[1in][s]{kitAbu-hu fI
+%   'l-`AdAti}}\,\footnote{\cs{rlframefox} has been adapted from
+%   \cs{framebox} for insertions of right-to-left text.}
 % \end{quote}
 % 
 % The same applies to footnotes:---
@@ -4051,6 +4057,11 @@ wa-ya.sIru ta.hta 'l-jild-i
 \DeclareOptionX{fullvoc}{\def\al@mode{fullvoc}}
 \DeclareOptionX{novoc}{\def\al@mode{novoc}}
 \DeclareOptionX{trans}{\def\al@mode{trans}}
+\define@boolkey{arabluatex.sty}[@pkg@]{export}[true]{%
+  \AtBeginDocument{\luadirect{al_openstream()}%
+    \MkArbBreak{@al@ob, @al@cb}}
+  \AtEndDocument{\luadirect{al_closestream()}}
+}
 \ExecuteOptionsX{voc}
 \ProcessOptionsX\relax
 \def\al@mode@voc{voc}
@@ -4315,10 +4326,10 @@ wa-ya.sIru ta.hta 'l-jild-i
 % Lua functions to insert transliterated Arabic text. Therefore, it is
 % not documented.
 %    \begin{macrocode}
-\NewDocumentCommand{\txarb}{+m}{\bgroup\textdir
-  TRT\arabicfont#1\egroup}
-\NewDocumentCommand{\txtrans}{+m}{\bgroup\textdir
-  TLT\al@trans@font\al@trans@style#1\egroup}
+\NewDocumentCommand{\txarb}{+m}{%
+  \bgroup\textdir TRT\arabicfont#1\egroup}
+\NewDocumentCommand{\txtrans}{+m}{%
+  \bgroup\textdir TLT\al@trans@font\al@trans@style#1\egroup}
 %    \end{macrocode}
 % \end{macro}
 % \end{macro}
@@ -4451,6 +4462,7 @@ wa-ya.sIru ta.hta 'l-jild-i
 \define@key[al]{verse}{color}[]{\color{#1}}
 \define@boolkey[al]{verse}{utf}[true]{}
 \define@boolkey[al]{verse}{delim}[true]{}
+\define@boolkey[al]{verse}{export}[true]{}
 \define@choicekey[al]{verse}{mode}{fullvoc, voc, novoc,
   trans}{\def\al@mode{#1}}
 \presetkeys[al]{verse}{metre={}, utf=false,
@@ -4459,11 +4471,21 @@ wa-ya.sIru ta.hta 'l-jild-i
 % Then follows the environment itself:
 %    \begin{macrocode}
 \NewDocumentEnvironment{arabverse}{O{}}%
-{\par\centering\noindent\bgroup\setkeys[al]{verse}[metre]{#1}%
+{\bgroup\setkeys[al]{verse}[width,gutter,color,utf,delim,mode,metre]{#1}%
+  \if@pkg@export\ifal@verse@export%
+  \ArbOutFile{\begin{arabverse}}%
+    \luadirect{tooutfile(\luastringN{[#1]})}%
+    \else\fi\else\fi\egroup%
+  \par\centering\noindent\bgroup\setkeys[al]{verse}[metre]{#1}%
   \ifx\al@mode\al@mode@trans%
   \ifal@verse@utf\setRL\else\setLR\fi%
   \else\setRL\fi}%
-{\hfill\setkeys[al]{verse}[width,gutter,utf,mode]{#1}\egroup\par}
+%{\hfill\setkeys[al]{verse}[width,gutter,utf,mode]{#1}\egroup\par}
+{\hfill\setkeys[al]{verse}[width,gutter,color,utf,delim,mode,export]{#1}%
+  \egroup\par%
+  \bgroup\setkeys[al]{verse}[width,gutter,color,utf,delim,mode,metre]{#1}%
+  \if@pkg@export\ifal@verse@export%
+  \ArbOutFile{\end{arabverse}}\else\fi\else\fi\egroup}
 %    \end{macrocode}
 % \begin{macro}{\bayt}
 %   \changes{v1.6}{2016/12/17}{New macro \cs{bayt} for typesetting
@@ -4490,11 +4512,17 @@ wa-ya.sIru ta.hta 'l-jild-i
 \NewDocumentCommand{\SetHemistichDelim}{m}{\def\al@hemistich@delim{#1}}
 \def\al@verse@stroke{\leavevmode\xleaders\hbox{\arb{--}}\hfill\kern0pt}
 \NewDocumentCommand{\bayt}{m o m}{%
+  \if@pkg@export\ifal@verse@export%
+    \luadirect{al_doexport("arabverse")}\else\fi\else\fi%
   {^^^^200b}% quick fix to a bug (or a missing feature) in either
             % luacolor or adjustbox 
   \ifdefined\savenotes\savenotes\else\fi%
   \edef\al@tatweel{--}%
+  \if@pkg@export\ifal@verse@export%
+    \ArbOutFile{\bayt\@al@ob}\else\fi\else\fi%
   \adjustbox{width=\al@bayt@width, height=\Height}{\arb@utf{#1}}%
+  \if@pkg@export\ifal@verse@export%
+    \ArbOutFile{\@al@cb}\else\fi\else\fi%
   \IfNoValueTF{#2}{%
     \ifal@verse@delim\makebox[\al@gutter@width][c]{\al@hemistich@delim}%
     \else%
@@ -4510,12 +4538,26 @@ wa-ya.sIru ta.hta 'l-jild-i
     \fi%
     \else%
     \ifx\al@mode\al@mode@trans%
+    \if@pkg@export\ifal@verse@export%
+      \ArbOutFile{[}\else\fi\else\fi%
     \adjustbox{width=\al@gutter@width, height=\Height}{\arb@utf{#2}}%
+    \if@pkg@export\ifal@verse@export%
+      \ArbOutFile{]}\else\fi\else\fi%
     \else%
+    \if@pkg@export\ifal@verse@export%
+      \ArbOutFile{[}\else\fi\else\fi%
     \makebox[\al@gutter@width][s]{\arb@utf{#2}}%
+    \if@pkg@export\ifal@verse@export%
+      \ArbOutFile{]}\else\fi\else\fi%
     \fi\fi}%
+  \if@pkg@export\ifal@verse@export%
+    \ArbOutFile{\@al@ob}\else\fi\else\fi%
   \adjustbox{width=\al@bayt@width, height=\Height}{\arb@utf{#3}}%
+  \if@pkg@export\ifal@verse@export%
+    \ArbOutFile{\@al@cb\\ }\else\fi\else\fi%
   \ifdefined\spewnotes\spewnotes\else\fi%
+  \if@pkg@export\ifal@verse@export%
+    \luadirect{al_doexport("no")}\else\fi\else\fi%
 }
 %    \end{macrocode}
 % \end{macro} 
@@ -4599,6 +4641,64 @@ wa-ya.sIru ta.hta 'l-jild-i
   \AtBeginDocument{\deffootnote{2em}{1.6em}{\LR{\thefootnotemark}.\enskip}}}}
 %    \end{macrocode}
 % \end{macro}
+% 
+% \paragraph{Exporting Unicode Arabic to external file}
+% \begin{macro}{\SetArbOutSuffix}
+% \changes{v1.13}{2018/08/13}{Sets a suffix to be appended to the
+% filename of the external Unicode file.}By default, |_out| is the
+% suffix to be appended to the external file in which
+% \package{arabluatex} exports Unicode in place of
+% |arabtex| or |buckwalter| strings. Any other suffix may be
+% set with \cs{SetArbOutSuffix}\marg{suffix}.
+%    \begin{macrocode}
+\NewDocumentCommand{\SetArbOutSuffix}{m}{
+  \luadirect{al_utffilesuffix(\luastringN{#1})}}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\ArbOutFile}
+% \changes{v1.13}{2018/08/13}{Silently exports its argument in the selected
+% external file.} \cs{ArbOutFile}\oarg{newline}\marg{string} silently
+% exports \meta{string} to the external selected file. It may take
+% |newline| as an optional argument in which case a carriage return is
+% appended to |string|.
+%    \begin{macrocode}
+\NewDocumentCommand{\ArbOutFile}{O{no} +m}{%
+  \if@pkg@export%
+  \luadirect{tooutfile(\luastringN{#2}, "#1")}\else\fi}
+%    \end{macrocode}
+% \end{macro}
+% \begin{environment}{arabexport}
+%   \changes{v1.13}{2018/08/13}{Processes and print its argument in
+%   the current file and exports it in full Unicode in the external
+%   selected \texttt{.tex} file.} The |arabexport| environment
+%   processes and prints its argument unchanged to the current |.pdf|
+%   file. Additionally, if \package{arabluatex} is loaded with the
+%   |export| option, this argument is exported to the external
+%   selected |.tex| file with Unicode in place of the original
+%   |arabtex| or |buckwalter| strings.
+%    \begin{macrocode}
+\NewEnviron{arabexport}{%
+  \if@pkg@export%
+  \par
+  \BODY
+  \luadirect{al_doexport("yes")}
+  \luadirect{tex.sprint(arbtoutf(\luastringO{\BODY}))}
+  \else\par\BODY\fi
+  }[\if@pkg@export\par\luadirect{al_doexport("no")}\else\par\fi]
+%    \end{macrocode}
+% \end{environment}
+% \begin{macro}{\arbpardir}
+% \changes{v1.13}{2018/08/13}{Sets the direction of Arabic paragraphs
+% once they are converted to Unicode.} \cs{arbpardir} is automatically
+% inserted by \package{arabluatex} at the beginning of Arabic
+% paragraphs converted to Unicode so that they are printed in the
+% right direction.
+%    \begin{macrocode}
+\NewDocumentCommand{\arbpardir}{}{%
+  \ifx\al@mode\al@mode@trans\setLR\else\setRL\fi}
+%    \end{macrocode}
+% \end{macro}
+% 
 % \subsection*{Errors and Warnings}
 %    \begin{macrocode}
 \newcommand{\al@warning}[1]{\PackageWarning{arabluatex}{#1}}
diff --git a/arabluatex.lua b/arabluatex.lua
index c0d9022..ac6a251 100644
--- a/arabluatex.lua
+++ b/arabluatex.lua
@@ -49,7 +49,7 @@ local cmdstar = lpeg.Cs(spce * lpeg.P("*"))
 local bsqbracketsii = lpeg.Cs(bsqbrackets^-2)
 local bcbracesii = lpeg.Cs(bcbraces^-2)
 local cmd = lpeg.Cs(dblbkslash * ascii^1 * cmdstar^-1)
-local cmdargs = lpeg.Cs(spce * bsqbracketsii * bcbracesii * bsqbrackets^-1)
+local cmdargs = lpeg.Cs(spce^-1 * bsqbracketsii * bcbracesii * bsqbrackets^-1)
 local arbargs = lpeg.Cs(spce^-1 * bsqbrackets^-1 * bcbraces)
 local baytargs = lpeg.Cs(spce * bcbraces * bsqbrackets^-1 * bcbraces)
 
@@ -642,7 +642,9 @@ local function processbuckw(str)
 return str
 end
 
---- ***
+-- The following functions produce a copy of the original .tex source
+-- file in which all arabtex strings are replaced with Unicode
+-- equivalents
 local utffilesuffix = "_out"
 local export_utf = "no"
 
@@ -671,14 +673,21 @@ function al_openstream()
 end
 
 local function processarbtoutf(str)
-   str = "\n\\begin{arbexport}\n"..str
+   str = "\\begin{arabexport}"..str
    str = string.gsub(str, "(\\prname)%s?(%b{})", function(tag, body)
                         body = string.sub(body, 2, -2)
                         return string.format("\\arb[trans]{\\uc{%s}}", body)
    end)
-   str = string.gsub(str, "(\\begin%s?{arab})(%b[])", "\\par\\bgroup\\arbpardir\\arb%2{")
+   str = string.gsub(str, "(\\begin%s?{arab})(%b[])", function(tag, body)
+                        if string.find(body, "trans") then
+                           return string.format("\\par\\bgroup\\setLR\\arb%s{", body)
+                        else
+                           return string.format("\\par\\bgroup\\setRL\\arb%s{", body)
+                        end
+   end)
    str = string.gsub(str, "(\\begin%s?{arab})", "\\par\\bgroup\\arbpardir\\arb{")
    str = string.gsub(str, "\\end%s?{arab}", "}\\egroup\\par")
+   -- This does not work, while the following two do. Look into this later.
    -- str = gsub(str, lpeg.Cs("\\arb") * spcenc * bsqbrackets^-1 * bcbraces, function(tag, opt, body)
    --            body = string.sub(body, 2, -2)
    --            return string.format("%s%s\\@al@pr@ob%s\\@al@pr@cb", tag, opt, body)
@@ -699,50 +708,50 @@ local function processarbtoutf(str)
                         body = string.sub(body, 2, -2)
                         return string.format("%s\\@al@pr@ob%s\\@al@pr@cb", tag, body)
    end)
-   str = string.gsub(str, "{", "\\@al@ob ")
-   str = string.gsub(str, "}", "\\@al@cb ")
+   str = string.gsub(str, "{", "\\@al@ob")
+   str = string.gsub(str, "}", "\\@al@cb")
    str = string.gsub(str, "\\@al@pr@ob", "{")
    str = string.gsub(str, "\\@al@pr@cb", "}")
    str = string.gsub(str, "(%b{})", function(body)
                         body = string.sub(body, 2, -2)
-                        body = string.gsub(body, "\\@al@ob%s?", "{")
-                        body = string.gsub(body, "\\@al@cb%s?", "}")
+                        body = string.gsub(body, "(%s?)(\\@al@ob)", "%1{")
+                        body = string.gsub(body, "(\\@al@cb)(%s?)", "}%2")
                         return string.format("{%s}", body)
    end)
-   str = str.."\n\\end{arbexport}"
+   str = str.."\\end{arabexport}"
    return str
 end
 
 function arbtoutf(str)
    str = processarbtoutf(str)
-   str = "\\ToOutFile{"..str.."}"
-   str = string.gsub(str, "(\\ToOutFile)%s?(%b{})", function(tag, body)
+   str = "\\ArbOutFile{"..str.."}"
+   str = string.gsub(str, "(\\ArbOutFile)%s?(%b{})", function(tag, body)
                         body = string.sub(body, 2, -2)
-                        body = gsub(body, lpeg.Cs("\\arb") * arbargs, "}%1%2\\ToOutFile{")
+                        body = gsub(body, lpeg.Cs("\\arb") * arbargs, "}%1%2\\ArbOutFile{")
                         return string.format("%s{%s}", tag, body)
    end)
-   str = string.gsub(str, "(\\ToOutFile)%s?(%b{})", function(tag, body)
+   str = string.gsub(str, "(\\ArbOutFile)%s?(%b{})", function(tag, body)
                         body = string.sub(body, 2, -2)
-                        body = string.gsub(body, "(\\abjad)%s?(%b{})", "}%1%2\\ToOutFile{")
+                        body = string.gsub(body, "(\\abjad)%s?(%b{})", "}%1%2\\ArbOutFile{")
                         return string.format("%s{%s}", tag, body)
    end)
-   str = string.gsub(str, "(\\ToOutFile)%s?(%b{})", function(tag, body)
+   str = string.gsub(str, "(\\ArbOutFile)%s?(%b{})", function(tag, body)
                         body = string.sub(body, 2, -2)
-                        body = string.gsub(body, "(\\arbmark)%s?(%b{})", "}%1%2\\ToOutFile{")
+                        body = string.gsub(body, "(\\arbmark)%s?(%b{})", "}%1%2\\ArbOutFile{")
                         return string.format("%s{%s}", tag, body)
    end)
-   -- str = string.gsub(str, "(\\arb%s?)(%b[])(%b{})", "}%1%2%3\\ToOutFile{")
-   -- str = string.gsub(str, "(\\arb)%s?(%b{})", "}%1%2\\ToOutFile{")
-   -- str = gsub(str, lpeg.Cs("\\bayt") * baytargs, "}\\ToOutFile{\\bayt}%1%2\\ToOutFile{")
---   str = gsub(str, lpeg.Cs("\\arb") * arbargs, "}%1%2\\ToOutFile{")
---   str = string.gsub(str, "(\\abjad)%s?(%b{})", "}%1%2\\ToOutFile{")
---   str = string.gsub(str, "(\\abjad)(%s?)(%b{})", "}%1%3\\tooutfile{")
+   -- str = string.gsub(str, "(\\arb%s?)(%b[])(%b{})", "}%1%2%3\\ArbOutFile{")
+   -- str = string.gsub(str, "(\\arb)%s?(%b{})", "}%1%2\\ArbOutFile{")
+   -- str = gsub(str, lpeg.Cs("\\bayt") * baytargs, "}\\ArbOutFile{\\bayt}%1%2\\ArbOutFile{")
+--   str = gsub(str, lpeg.Cs("\\arb") * arbargs, "}%1%2\\ArbOutFile{")
+--   str = string.gsub(str, "(\\abjad)%s?(%b{})", "}%1%2\\ArbOutFile{")
+--   str = string.gsub(str, "(\\abjad)(%s?)(%b{})", "}%1%3\\ArbOutFile{")
 --   str = string.gsub(str, "(\\begin%s?{arab})", "}%1")
---   str = string.gsub(str, "(\\end%s?{arab})", "%1\\ToOutFile{")
+--   str = string.gsub(str, "(\\end%s?{arab})", "%1\\ArbOutFile{")
 --   str = string.gsub(str, "(\\begin)", "\n%1")
 --   str = string.gsub(str, "(\\end)", "\n%1")
 --   str = string.gsub(str, "(\\bayt)", "\n%1")
---   str = string.gsub(str, "(\\\\)", "%1\n")
+   --   str = string.gsub(str, "(\\\\)", "%1\n")
    return str
 end
 
@@ -784,8 +793,11 @@ function al_closestream()
    --]]
    t = string.gsub(t, "\\arabicfont{}", "")
    t = string.gsub(t, "\\par ", "\n\n")
-   t = string.gsub(t, "\\@al@ob%s?", "{")
-   t = string.gsub(t, "\\@al@cb", "}")
+   t = string.gsub(t, "(%s?)(\\@al@ob%s?)", "{")
+   t = string.gsub(t, "(\\@al@cb)(%s?)", "}%2")
+   t = gsub(t, lpeg.Cs("\\begin") * spcenc^-1 * bcbraces * cmdargs, "\n%1%2%3\n")
+   t = string.gsub(t, "(\\\\)", "%1\n")
+   t = string.gsub(t, "(\\end%s?)(%b{})", "\n%1%2")
    t = string.gsub(t, "\n\n\n", "\n\n")
    if string.find(t, "\\begin%s?{document}.-\\arb%s?[%[%{]") then      
       tex.print([[\unexpanded{\PackageWarningNoLine{arabluatex}{There are still 'arabtex' strings to be converted. Open ]]..tex.jobname..utffilesuffix..".tex"..[[ and compile it one more time}}]])
@@ -800,7 +812,7 @@ function al_closestream()
    return true
 end
 
-
+-- Process standard arabluatex modes:
 function processvoc(str, rules, scheme)
    str = takeoutarb(str)
    str = processarbnull(str, scheme)
@@ -820,6 +832,10 @@ function processvoc(str, rules, scheme)
    if export_utf == "yes" then
       tofile = "\\txarb{"..str.."}"
       tooutfile(tofile)
+   elseif export_utf == "arabverse" then
+      tofile = "\\txarb{"..str.."}"
+      tooutfile(tofile)
+      return str
    else
       return str
    end
@@ -847,6 +863,10 @@ function processfullvoc(str, rules, scheme)
    if export_utf == "yes" then
       tofile = "\\txarb{"..str.."}"
       tooutfile(tofile)
+   elseif export_utf == "arabverse" then
+      tofile = "\\txarb{"..str.."}"
+      tooutfile(tofile)
+      return str
    else
       return str
    end
@@ -872,6 +892,10 @@ function processnovoc(str, rules, scheme)
    if export_utf == "yes" then
       tofile = "\\txarb{"..str.."}"
       tooutfile(tofile)
+   elseif export_utf == "arabverse" then
+      tofile = "\\txarb{"..str.."}"
+      tooutfile(tofile)
+      return str
    else
       return str
    end
@@ -899,7 +923,11 @@ function processtrans(str, mode, rules, scheme)
    str = unprotectarb(str)
    if export_utf == "yes" then
       tofile = str
-   tooutfile(str)
+      tooutfile(str)
+   elseif export_utf == "arabverse" then
+      tofile = str
+      tooutfile(tofile)
+      return str      
    else
       return str
    end