- tex2lyx: support for the polyglossia language variants

- XeTeX-polyglossia.tex: simplifications, added some more testcases

src/tex2lyx/Preamble.cpp

@@ -92,6 +92,41 @@ const char * const known_coded_languages[] = {"french", "afrikaans", "albanian",
 "uppersorbian", "uppersorbian", "english", "english", "vietnamese", "welsh",
 0};
 
+/**
+ * known polyglossia language names (including variants)
+ */
+const char * const polyglossia_languages[] = {
+"albanian", "croatian", "hebrew", "norsk", "swedish", "amharic", "czech", "hindi",
+"nynorsk", "syriac", "arabic", "danish", "icelandic", "occitan", "tamil",
+"armenian", "divehi", "interlingua", "polish", "telugu", "asturian", "dutch",
+"irish", "portuges", "thai", "bahasai", "english", "italian", "romanian", "turkish",
+"bahasam", "esperanto", "lao", "russian", "turkmen", "basque", "estonian", "latin",
+"samin", "ukrainian", "bengali", "farsi", "latvian", "sanskrit", "urdu", "brazil",
+"brazilian", "finnish", "lithuanian", "scottish", "usorbian", "breton", "french",
+"lsorbian", "serbian", "vietnamese", "bulgarian", "galician", "magyar", "slovak",
+"welsh", "catalan", "german", "malayalam", "slovenian", "coptic", "greek",
+"marathi", "spanish",
+"american", "ancient", "australian", "british", "monotonic", "newzealand",
+"polytonic", 0};
+
+/**
+ * the same as polyglossia_languages with .lyx names
+ * please keep this in sync with polyglossia_languages line by line!
+ */
+const char * const coded_polyglossia_languages[] = {
+"albanian", "croatian", "hebrew", "norsk", "swedish", "amharic", "czech", "hindi",
+"nynorsk", "syriac", "arabic_arabi", "danish", "icelandic", "occitan", "tamil",
+"armenian", "divehi", "interlingua", "polish", "telugu", "asturian", "dutch",
+"irish", "portuges", "thai", "bahasa", "english", "italian", "romanian", "turkish",
+"bahasam", "esperanto", "lao", "russian", "turkmen", "basque", "estonian", "latin",
+"samin", "ukrainian", "bengali", "farsi", "latvian", "sanskrit", "urdu", "brazilian",
+"brazilian", "finnish", "lithuanian", "scottish", "uppersorbian", "breton", "french",
+"lowersorbian", "serbian", "vietnamese", "bulgarian", "galician", "magyar", "slovak",
+"welsh", "catalan", "ngerman", "malayalam", "slovene", "coptic", "greek",
+"marathi", "spanish",
+"american", "ancientgreek", "australian", "british", "greek", "newzealand",
+"polutonikogreek", 0};
+
 /// languages with english quotes (.lyx names)
 const char * const known_english_quotes_languages[] = {"american", "australian",
 "bahasa", "bahasam", "brazilian", "canadian", "chinese-simplified", "english",
@@ -686,6 +721,8 @@ void Preamble::handle_package(Parser &p, string const & name,
 			// reasons for it.
 			h_preamble << "\\usepackage[" << opts << "]{babel}\n";
 			delete_opt(options, known_languages);
+			// finally translate the babel name to a LyX name
+			h_language = babel2lyx(h_language);
 		}
 		else
 			h_preamble << "\\usepackage{babel}\n";
@@ -824,9 +861,6 @@ void Preamble::handle_if(Parser & p, bool in_lyx_preamble)
 
 bool Preamble::writeLyXHeader(ostream & os, bool subdoc)
 {
-	// translate from babel to LyX names
-	h_language = babel2lyx(h_language);
-
 	// set the quote language
 	// LyX only knows the following quotes languages:
 	// english, swedish, german, polish, french and danish
@@ -1073,14 +1107,31 @@ void Preamble::parse(Parser & p, string const & forceclass,
 			h_paperpagestyle = p.verbatim_item();
 
 		else if (t.cs() == "setdefaultlanguage") {
-			// FIXME: we don't yet care about the option because LyX doesn't
+			xetex = true;
-			// support this yet, see bug #8214
+			// We don't yet care about non-language variant options
-			p.hasOpt() ? p.getOpt() : string();
+			// because LyX doesn't support this yet, see bug #8214
-			h_language = p.verbatim_item();
+			if (p.hasOpt()) {
+				string langopts = p.getOpt();
+				// check if the option contains a variant, if yes, extract it
+				string::size_type pos_var = langopts.find("variant");
+				string::size_type i = langopts.find(',', pos_var);
+				if (pos_var != string::npos){
+					string variant;
+					if (i == string::npos)
+						variant = langopts.substr(pos_var + 8, langopts.length() - pos_var - 9);
+					else
+						variant = langopts.substr(pos_var + 8, i - pos_var - 8);
+					h_language = variant;
+				}
+				p.verbatim_item();
+			} else
+				h_language = p.verbatim_item();
+			//finally translate the poyglossia name to a LyX name
+			h_language = polyglossia2lyx(h_language);
 		}
 
 		else if (t.cs() == "setotherlanguage") {
-			// FIXME: we don't yet care about the option because LyX doesn't
+			// We don't yet care about the option because LyX doesn't
 			// support this yet, see bug #8214
 			p.hasOpt() ? p.getOpt() : string();
 			p.verbatim_item();
@@ -1525,6 +1576,15 @@ string babel2lyx(string const & language)
 }
 
 
+string polyglossia2lyx(string const & language)
+{
+	char const * const * where = is_known(language, polyglossia_languages);
+	if (where)
+		return coded_polyglossia_languages[where - polyglossia_languages];
+	return language;
+}
+
+
 string rgbcolor2code(string const & name)
 {
 	char const * const * where = is_known(name, known_basic_colors);

src/tex2lyx/test/XeTeX-polyglossia.tex

@@ -1,20 +1,20 @@
-\documentclass[albanian]{article}
+\documentclass{article}
 \usepackage{fontspec}
-\setmainfont[Mapping=tex-text]{Aharoni}
+\setmainfont{Aharoni}
 \setsansfont[Mapping=tex-text]{AngsanaUPC}
 \setmonofont{Arial Black}
 \usepackage{xunicode}
 \usepackage{polyglossia}
-\setdefaultlanguage{coptic}
+\setdefaultlanguage[variant=british, ordinalmonthday = false]{english}
 \setotherlanguage{albanian}
-\setotherlanguage{divehi}
+\setotherlanguage[variant=polytonic]{greek}
-\setotherlanguage[variant=monotonic]{greek}
 \setotherlanguage{syriac}
 \begin{document}
 
 \part{df}
 
-1 Coptic \textgreek[variant=ancient]{ancient Greek} Coptic
+1 English \textgreek[variant=ancient]{ancient Greek} English
+\textgreek[numerals=arabic, variant=ancient]{ancient Greek with Arabic numerals} English
 
 \begin{albanian}%
 
@@ -23,26 +23,13 @@
 2 Albanian \textgreek{Greek} Albanian \textsc{smallcaps}
 
 \end{albanian}%
-\begin{divehi}%
-
-\part{df}
-
-3
-
-\end{divehi}%
 \begin{syriac}%
 
 \part{df}
 
-4
+3 Syriac
 
 \end{syriac}%
-\begin{divehi}%
-
-\part{df}
-
-5
-\end{divehi}%
 
 %empty language paragraph
 \begin{albanian}%

src/tex2lyx/tex2lyx.h

@@ -45,6 +45,8 @@ public:
 
 /// Translate babel language name to LyX language name
 extern std::string babel2lyx(std::string const & language);
+/// Translate polyglossia language name to LyX language name
+extern std::string polyglossia2lyx(std::string const & language);
 /// Translate basic color name or RGB color in LaTeX syntax to LyX color code
 extern std::string rgbcolor2code(std::string const & name);
 

src/tex2lyx/text.cpp

@@ -118,7 +118,7 @@ char const * const known_coded_ref_commands[] = { "ref", "pageref", "vref",
  "vpageref", "formatted", "eqref", 0 };
 
 /**
- * known polyglossia language names (inluding synomyms)
+ * known polyglossia language names (including variants)
  */
 const char * const polyglossia_languages[] = {
 "albanian", "croatian", "hebrew", "norsk", "swedish", "amharic", "czech", "hindi",
@@ -130,7 +130,9 @@ const char * const polyglossia_languages[] = {
 "brazilian", "finnish", "lithuanian", "scottish", "usorbian", "breton", "french",
 "lsorbian", "serbian", "vietnamese", "bulgarian", "galician", "magyar", "slovak",
 "welsh", "catalan", "german", "malayalam", "slovenian", "coptic", "greek",
-"marathi", "spanish", 0};
+"marathi", "spanish"
+"american", "ancient", "australian", "british", "monotonic", "newzealand",
+"polytonic", 0};
 
 /**
  * the same as polyglossia_languages with .lyx names
@@ -146,15 +148,9 @@ const char * const coded_polyglossia_languages[] = {
 "brazilian", "finnish", "lithuanian", "scottish", "uppersorbian", "breton", "french",
 "lowersorbian", "serbian", "vietnamese", "bulgarian", "galician", "magyar", "slovak",
 "welsh", "catalan", "ngerman", "malayalam", "slovene", "coptic", "greek",
-"marathi", "spanish", 0};
+"marathi", "spanish"
-
+"american", "ancientgreek", "australian", "british", "greek", "newzealand",
-string polyglossia2lyx(string const & language)
+"polutonikogreek", 0};
-{
-	char const * const * where = is_known(language, polyglossia_languages);
-	if (where)
-		return coded_polyglossia_languages[where - polyglossia_languages];
-	return language;
-}
 
 /*!
  * natbib commands.
@@ -3470,15 +3466,34 @@ void parse_text(Parser & p, ostream & os, unsigned flags, bool outer,
 		
 		else if (is_known(t.cs().substr(4, string::npos), polyglossia_languages)) {
 			// scheme is \textLANGUAGE{text} where LANGUAGE is in polyglossia_languages[]
-			string const lang = polyglossia2lyx(t.cs().substr(4, string::npos));
+			string lang;
-			// FIXME: we have to output the whole command if it has an option
+			// We have to output the whole command if it has an option
-			// because lyX doesn't support this yet, see bug #8214
+			// because LyX doesn't support this yet, see bug #8214,
-			if (p.hasOpt())
+			// only if there is a single option specifying a variant, we can handle it.
-				handle_ert(os, t.asInput() + p.getOpt(), context);
+			if (p.hasOpt()) {
-			else 
+				string langopts = p.getOpt();
+				// check if the option contains a variant, if yes, extract it
+				string::size_type pos_var = langopts.find("variant");
+				string::size_type i = langopts.find(',');
+				if (pos_var != string::npos){
+					string variant;
+					if (i == string::npos) {
+						variant = langopts.substr(pos_var + 8, langopts.length() - pos_var - 9);
+						lang = polyglossia2lyx(variant);
+						parse_text_attributes(p, os, FLAG_ITEM, outer,
+							                  context, "\\lang",
+							                  context.font.language, lang);
+					}
+					else
+						handle_ert(os, t.asInput() + langopts, context);
+				} else
+					handle_ert(os, t.asInput() + langopts, context);
+			} else {
+				lang = polyglossia2lyx(t.cs().substr(4, string::npos));
 				parse_text_attributes(p, os, FLAG_ITEM, outer,
-			                          context, "\\lang",
+					                  context, "\\lang",
-			                          context.font.language, lang);
+					                  context.font.language, lang);
+			}
 		}
 
 		else if (t.cs() == "inputencoding") {