Fix language handling in tex2lyx:

- Add missing babel synonyms (taken from a standard TeXLive 2010 install).
  Some synonyms might still be missing (if they are defined not by babel
  itself, but additional packages). Adding these is now simple: Only
  known_languages and known_coded_languages need to be touched.
- Translate babel language name arguments of \foreignlanguage and
  \selectlanguage to LyX names.
- Translate second argument of \foreignlanguage to LyX.
- Cleanup mixed language names in the preamble parser: Now all checks are
  done with LyX names, not a mix of babel and LyX names. This sets correct
  quotes even if the language as given as an alias.

git-svn-id: svn://svn.lyx.org/lyx/lyx-devel/trunk@37188 a592a061-630c-0410-9148-cb99ea01b6c8

diff --git a/lib/languages b/lib/languages
index 04f5a479ce25e826e0865e5a453b82fc7e4a9157..0fa452eafd3caf9e6e94e3e3b830e5222cafda57 100644 (file)
--- a/lib/languages
+++ b/lib/languages
@@ -517,6 +517,13 @@ Language korean
        LangCode        ko
 End
 
+# Currently not supported (file format change!)
+# Language kurmanji
+#      GuiName          "Kurmanji"
+#      BabelName        kurmanji
+#      LangCode         kmr
+# End
+
 # Currently not supported (file format change!)
 # Language lao
 #      GuiName          "Lao"

diff --git a/src/tex2lyx/preamble.cpp b/src/tex2lyx/preamble.cpp
index 22ffcd976cb3c1026ad8d0a6b739d825326103d9..1b432a2bde74e83704e6c49f86cd9c5dd7a3c6fc 100644 (file)
--- a/src/tex2lyx/preamble.cpp
+++ b/src/tex2lyx/preamble.cpp
@@ -57,53 +57,69 @@ namespace {
 // Both changes require first that support for non-babel languages (CJK,
 // armtex) is added.
 // add turkmen for lyxformat 383
-const char * const known_languages[] = { "afrikaans", "albanian", "american",
-"arabic", "arabtex", "austrian", "bahasa", "bahasai", "bahasam", "basque",
-"belarusian", "brazil", "brazilian", "breton", "british", "bulgarian",
+/**
+ * known babel language names (including synonyms)
+ * not in standard babel: arabic, arabtex, belarusian, serbian-latin, thai
+ * not yet supported by LyX: kurmanji
+ * please keep this in sync with known_coded_languages line by line!
+ */
+const char * const known_languages[] = {"acadian", "afrikaans", "albanian",
+"american", "arabic", "arabtex", "austrian", "bahasa", "bahasai", "bahasam",
+"basque", "belarusian", "brazil", "brazilian", "breton", "british", "bulgarian",
 "canadian", "canadien", "catalan", "croatian", "czech", "danish", "dutch",
-"english", "esperanto", "estonian", "finnish", "francais", "french",
+"english", "esperanto", "estonian", "farsi", "finnish", "francais", "french",
 "frenchb", "frenchle", "frenchpro", "galician", "german", "germanb", "greek",
-"hebrew", "icelandic", "indon", "indonesian", "interlingua", "irish",
-"italian", "kazakh", "latin", "latvian", "lithuanian", "lsorbian", "magyar",
-"malay", "meyalu", "mongolian", "naustrian", "ngerman", "ngermanb", "norsk",
-"nynorsk", "polutonikogreek", "polish", "portuges", "portuguese", "romanian",
-"russian", "russianb", "samin", "scottish", "serbian", "serbian-latin",
-"slovak", "slovene", "spanish", "swedish", "thai", "turkish", "ukraineb",
-"ukrainian", "usorbian", "vietnam", "welsh", 0};
-
-const char * const known_bahasa_languages[] = {"bahasa", "bahasai",
-                                               "indon", "indonesian", 0};
-const char * const known_bahasam_languages[] = {"bahasam", "malay",
-                                               "meyalu", 0};
-const char * const known_brazilian_languages[] = {"brazil", "brazilian", 0};
-const char * const known_french_languages[] = {"french", "frenchb", "francais",
-                                               "frenchle", "frenchpro", 0};
-const char * const known_german_languages[] = {"german", "germanb", 0};
-const char * const known_ngerman_languages[] = {"ngerman", "ngermanb", 0};
-const char * const known_portuguese_languages[] = {"portuges", "portuguese", 0};
-const char * const known_russian_languages[] = {"russian", "russianb", 0};
-const char * const known_ukrainian_languages[] = {"ukrainian", "ukraineb", 0};
-
-//add these to known_english_quotes_languages when updating to lyxformat 268:
-//"chinese-simplified", "korean"
-// This requires first that support for non-babel languages (CJK) is added.
-const char * const known_english_quotes_languages[] = {"american", "canadian",
-"english", "esperanto", "hebrew", "irish", "scottish", "thai", 0};
+"hebrew", "hungarian", "icelandic", "indon", "indonesian", "interlingua",
+"irish", "italian", "kazakh", "latin", "latvian", "lithuanian", "lowersorbian",
+"lsorbian", "magyar", "malay", "meyalu", "mongolian", "naustrian", "newzealand",
+"ngerman", "ngermanb", "norsk", "nynorsk", "polutonikogreek", "polish",
+"portuges", "portuguese", "romanian", "russian", "russianb", "samin",
+"scottish", "serbian", "serbian-latin", "slovak", "slovene", "spanish",
+"swedish", "thai", "turkish", "ukraineb", "ukrainian", "uppersorbian",
+"UKenglish", "USenglish", "usorbian", "vietnam", "welsh", 0};
+
+/**
+ * the same as known_languages with .lyx names
+ * please keep this in sync with known_languages line by line!
+ */
+const char * const known_coded_languages[] = {"french", "afrikaans", "albanian",
+"american", "arabic_arabi", "arabic_arabtex", "austrian", "bahasa", "bahasa", "bahasam",
+"basque", "belarusian", "brazilian", "brazilian", "breton", "british", "bulgarian",
+"canadian", "canadien", "catalan", "croatian", "czech", "danish", "dutch",
+"english", "esperanto", "estonian", "farsi", "finnish", "french", "french",
+"french", "french", "french", "galician", "german", "german", "greek",
+"hebrew", "magyar", "icelandic", "bahasa", "bahasa", "interlingua",
+"irish", "italian", "kazakh", "latin", "latvian", "lithuanian", "lowersorbian",
+"lowersorbian", "magyar", "bahasam", "bahasam", "mongolian", "naustrian", "english",
+"ngerman", "ngerman", "norsk", "nynorsk", "polutonikogreek", "polish",
+"portuguese", "portuguese", "romanian", "russian", "russian", "samin",
+"scottish", "serbian", "serbian-latin", "slovak", "slovene", "spanish",
+"swedish", "thai", "turkish", "ukrainian", "ukrainian", "uppersorbian",
+"uppersorbian", "english", "english", "vietnamese", "welsh", 0};
+
+/// languages with english quotes (.lyx names)
+const char * const known_english_quotes_languages[] = {"american", "bahasa",
+"bahasam", "brazilian", "canadian", "chinese-simplified", "english",
+"esperanto", "hebrew", "irish", "korean", "portuguese", "scottish", "thai", 0};
 
 //add this to known_french_quotes_languages when updating to
 //lyxformat 383: "turkmen"
-const char * const known_french_quotes_languages[] = {"albanian", "arabic",
-"basque", "canadien", "catalan", "galician", "greek", "italian", "norsk",
-"nynorsk", "polutonikogreek", "spanish", "spanish-mexico", "turkish",
-"vietnam", 0};
+/// languages with french quotes (.lyx names)
+const char * const known_french_quotes_languages[] = {"albanian",
+"arabic_arabi", "arabic_arabtex", "basque", "canadien", "catalan", "french",
+"galician", "greek", "italian", "norsk", "nynorsk", "polutonikogreek",
+"russian", "spanish", "spanish-mexico", "turkish", "ukrainian", "vietnamese", 0};
 
+/// languages with german quotes (.lyx names)
 const char * const known_german_quotes_languages[] = {"austrian", "bulgarian",
-"czech", "icelandic", "lithuanian", "lsorbian", "naustrian", "serbian",
-"serbian-latin", "slovak", "slovene", "usorbian",  0};
+"czech", "german", "icelandic", "lithuanian", "lowersorbian", "naustrian",
+"ngerman", "serbian", "serbian-latin", "slovak", "slovene", "uppersorbian", 0};
 
+/// languages with polish quotes (.lyx names)
 const char * const known_polish_quotes_languages[] = {"afrikaans", "croatian",
 "dutch", "estonian", "magyar", "polish", "romanian", 0};
 
+/// languages with swedish quotes (.lyx names)
 const char * const known_swedish_quotes_languages[] = {"finnish",
 "swedish", 0};
 
@@ -137,7 +153,7 @@ const char * const known_coded_paper_margins[] = { "leftmargin", "topmargin",
 ostringstream h_preamble;
 string h_textclass               = "article";
 string h_use_default_options     = "false";
-string h_options                 = string();
+string h_options;
 string h_language                = "english";
 string h_inputencoding           = "auto";
 string h_font_roman              = "default";
@@ -182,27 +198,12 @@ string h_defskip                 = "medskip";
 string h_paragraph_indentation   = "default";
 string h_quotes_language         = "english";
 string h_papercolumns            = "1";
-string h_papersides              = string();
+string h_papersides;
 string h_paperpagestyle          = "default";
 string h_listings_params;
 string h_tracking_changes        = "false";
 string h_output_changes          = "false";
-string h_margins                 = "";
-
-
-/// translates a babel language name to a LyX language name
-string babel2lyx(string language)
-{
-       if (language == "arabtex")
-               return "arabic_arabtex";
-       if (language == "arabic")
-               return "arabic_arabi";
-       if (language == "lsorbian")
-               return "lowersorbian";
-       if (language == "usorbian")
-               return "uppersorbian";
-       return language;
-}
+string h_margins;
 
 
 // returns true if at least one of the options in what has been found
@@ -606,25 +607,8 @@ void handle_package(Parser &p, string const & name, string const & opts,
 
 void end_preamble(ostream & os, TextClass const & /*textclass*/)
 {
-       // merge synonym languages
-       if (is_known(h_language, known_bahasa_languages))
-               h_language = "bahasa";
-       else if (is_known(h_language, known_bahasam_languages))
-               h_language = "bahasam";
-       else if (is_known(h_language, known_brazilian_languages))
-               h_language = "brazilian";
-       else if (is_known(h_language, known_french_languages))
-               h_language = "french";
-       else if (is_known(h_language, known_german_languages))
-               h_language = "german";
-       else if (is_known(h_language, known_ngerman_languages))
-               h_language = "ngerman";
-       else if (is_known(h_language, known_portuguese_languages))
-               h_language = "portuguese";
-       else if (is_known(h_language, known_russian_languages))
-               h_language = "russian";
-       else if (is_known(h_language, known_ukrainian_languages))
-               h_language = "ukrainian";
+       // translate from babel to LyX names
+       h_language = babel2lyx(h_language);
 
        // set the quote language
        // LyX only knows the following quotes languages:
@@ -638,15 +622,10 @@ void end_preamble(ostream & os, TextClass const & /*textclass*/)
        if (h_language == "danish")
                h_quotes_language = "danish";
        // french
-       else if (is_known(h_language, known_french_quotes_languages)
-               || is_known(h_language, known_french_languages)
-               || is_known(h_language, known_russian_languages)
-               || is_known(h_language, known_ukrainian_languages))
+       else if (is_known(h_language, known_french_quotes_languages))
                h_quotes_language = "french";
        // german
-       else if (is_known(h_language, known_german_quotes_languages)
-               || is_known(h_language, known_german_languages)
-               || is_known(h_language, known_ngerman_languages))
+       else if (is_known(h_language, known_german_quotes_languages))
                h_quotes_language = "german";
        // polish
        else if (is_known(h_language, known_polish_quotes_languages))
@@ -655,15 +634,9 @@ void end_preamble(ostream & os, TextClass const & /*textclass*/)
        else if (is_known(h_language, known_swedish_quotes_languages))
                h_quotes_language = "swedish";
        //english
-       else if (is_known(h_language, known_english_quotes_languages)
-               || is_known(h_language, known_bahasa_languages)
-               || is_known(h_language, known_bahasam_languages)
-               || is_known(h_language, known_brazilian_languages)
-               || is_known(h_language, known_portuguese_languages))
+       else if (is_known(h_language, known_english_quotes_languages))
                h_quotes_language = "english";
 
-       h_language = babel2lyx(h_language);
-
        // output the LyX file settings
        os << "#LyX file created by tex2lyx " << PACKAGE_VERSION << "\n"
           << "\\lyxformat " << LYX_FORMAT << '\n'
@@ -745,6 +718,7 @@ void end_preamble(ostream & os, TextClass const & /*textclass*/)
 
 } // anonymous namespace
 
+
 void parse_preamble(Parser & p, ostream & os, 
        string const & forceclass, TeX2LyXDocClass & tc)
 {
@@ -1130,6 +1104,16 @@ void parse_preamble(Parser & p, ostream & os,
        end_preamble(os, tc);
 }
 
+
+/// translates a babel language name to a LyX language name
+string babel2lyx(string const & language)
+{
+       char const * const * where = is_known(language, known_languages);
+       if (where)
+               return known_coded_languages[where - known_languages];
+       return language;
+}
+
 // }])
 
 

diff --git a/src/tex2lyx/test/test-structure.tex b/src/tex2lyx/test/test-structure.tex
index 68589858a57968dc53f714df022056933d10410b..95d5be3f8e8516e673fa0c0c75d542a082ebb0f5 100644 (file)
--- a/src/tex2lyx/test/test-structure.tex
+++ b/src/tex2lyx/test/test-structure.tex
@@ -1,9 +1,10 @@
 %% LyX trick_preamble_code_into_believing_that_this_was_created_by_lyx created this file.  For more info, see http://www.lyx.org/.
 %% Do not edit unless you really know what you are doing.
-\documentclass[legalpaper]{article}
+\documentclass[legalpaper,francais,german,newzealand]{article}
 
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LyX specific LaTeX commands.
 
+\usepackage{babel}
 \newcommand{\noun}[1]{\textsc{#1}}
 
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Textclass specific LaTeX commands.
@@ -159,6 +160,15 @@ and bibliography:
 
 \noindent blabla bla bla
 
+switch to german:\selectlanguage{german}
+Hallo!
+\foreignlanguage{french}{some \emph{french}}
+back to english:\selectlanguage{newzealand}
+and some nested \foreignlanguage{francais}{french \foreignlanguage{german}{nested
+\emph{german}} french} english stuff.
+Note that we both used \texttt{french} and the \texttt{francais} alias for the
+french text, but for some reason this does not work with the
+\texttt{newzealand} alias and \texttt{english} for english text.
 
 \section{Another Appendix section}
 

diff --git a/src/tex2lyx/tex2lyx.h b/src/tex2lyx/tex2lyx.h
index f6fae1e9b4cb40ee0d9e5e1305fb1b4559badc0a..fabed2bc45a5ff7f6029117dc6032d69877b8a7c 100644 (file)
--- a/src/tex2lyx/tex2lyx.h
+++ b/src/tex2lyx/tex2lyx.h
@@ -43,6 +43,8 @@ class TeX2LyXDocClass : public DocumentClass
 /// in preamble.cpp
 void parse_preamble(Parser & p, std::ostream & os, 
        std::string const & forceclass, TeX2LyXDocClass & tc);
+/// Translate babel language name to LyX language name
+extern std::string babel2lyx(std::string const & language);
 
 /// used packages with options
 extern std::map<std::string, std::vector<std::string> > used_packages;

diff --git a/src/tex2lyx/text.cpp b/src/tex2lyx/text.cpp
index 113d9c547314bcbb6b7f3688a59fdf6491632dae..ff1dbbf91f7da971d13aa8b9801ed0c1d8dfb2f0 100644 (file)
--- a/src/tex2lyx/text.cpp
+++ b/src/tex2lyx/text.cpp
@@ -2441,21 +2441,15 @@ void parse_text(Parser & p, ostream & os, unsigned flags, bool outer,
                        // save the language for the case that a
                        // \foreignlanguage is used 
 
-                       // FIXME: \lang needs a LyX name, but we set a LaTeX name
-                       context.font.language = subst(p.verbatim_item(), "\n", " ");
-                       os << "\\lang " << context.font.language << "\n";
+                       context.font.language = babel2lyx(p.verbatim_item());
+                       os << "\n\\lang " << context.font.language << "\n";
                }
 
                else if (t.cs() == "foreignlanguage") {
-                       context.check_layout(os);
-                       // FIXME: \lang needs a LyX name, but we set a LaTeX name
-                       os << "\n\\lang " << subst(p.verbatim_item(), "\n", " ") << "\n";
-                       os << subst(p.verbatim_item(), "\n", " ");
-                       // FIXME: the second argument of selectlanguage
-                       // has to be parsed (like for \textsf, for
-                       // example). 
-                       // set back to last selectlanguage
-                       os << "\n\\lang " << context.font.language << "\n";
+                       string const lang = babel2lyx(p.verbatim_item());
+                       parse_text_attributes(p, os, FLAG_ITEM, outer,
+                                             context, "\\lang",
+                                             context.font.language, lang);
                }
 
                else if (t.cs() == "inputencoding") {