* This file is part of LyX, the document processor.
* Licence details can be found in the file COPYING.
*
- * \author André Pönitz
- * \author Uwe Stöhr
+ * \author André Pönitz
+ * \author Uwe Stöhr
*
* Full author contact details are available in file CREDITS.
*/
#include "support/filetools.h"
#include "support/lstrings.h"
+#include <boost/regex.hpp>
+
#include <algorithm>
#include <iostream>
#include <sstream>
using namespace std;
using namespace lyx::support;
+using boost::regex;
+using boost::smatch;
namespace lyx {
// needed to handle encodings with babel
bool one_language = true;
-// to avoid that the babel options overwrite the documentclass options
-bool documentclass_language;
-
namespace {
+//add these to known_languages when updating to lyxformat 268:
+//"chinese-simplified", "chinese-traditional", "japanese", "korean"
const char * const known_languages[] = { "afrikaans", "american", "arabic",
-"austrian", "bahasa", "basque", "belarusian", "brazil", "breton", "british",
-"bulgarian", "canadian", "canadien", "catalan", "croatian", "czech", "danish",
-"dutch", "english", "esperanto", "estonian", "finnish", "francais", "french",
-"frenchb", "frenchle", "frenchpro", "galician", "german", "germanb", "greek",
-"hebrew", "icelandic", "irish", "italian", "lsorbian", "magyar", "naustrian",
-"ngerman", "ngermanb", "norsk", "nynorsk", "polish", "portuges", "romanian",
-"russian", "russianb", "scottish", "serbian", "slovak", "slovene", "spanish",
-"swedish", "thai", "turkish", "ukraineb", "ukrainian", "usorbian", "welsh", 0};
-
-//note this when updating to lyxformat 305:
+"austrian", "bahasa", "basque", "belarusian", "brazil", "brazilian", "breton",
+"british", "bulgarian", "canadian", "canadien", "catalan", "croatian", "czech",
+"danish", "dutch", "english", "esperanto", "estonian", "finnish", "francais",
+"french", "frenchb", "frenchle", "frenchpro", "galician", "german", "germanb",
+"greek", "hebrew", "icelandic", "irish", "italian", "kazakh", "lsorbian", "magyar",
+"naustrian", "ngerman", "ngermanb", "norsk", "nynorsk", "polish", "portuges",
+"portuguese", "romanian", "russian", "russianb", "scottish", "serbian", "slovak",
+"slovene", "spanish", "swedish", "thai", "turkish", "ukraineb", "ukrainian",
+"usorbian", "welsh", 0};
+
+//add this when updating to lyxformat 305:
//bahasai, indonesian, and indon = equal to bahasa
-//malay, and meyalu = equal to bahasam
-
+//malay and meyalu = equal to bahasam
+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"
+const char * const known_english_quotes_languages[] = {"american", "canadian",
+"english", "esperanto", "hebrew", "irish", "scottish", "thai", 0};
+
+//add this to known_french_quotes_languages when updating to lyxformat 327:
+//"spanish-mexico"
+const char * const known_french_quotes_languages[] = {"albanian", "arabic",
+"basque", "canadien", "catalan", "galician", "greek", "italian", "norsk",
+"nynorsk", "spanish", "turkish", 0};
+
+const char * const known_german_quotes_languages[] = {"austrian", "bulgarian",
+"czech", "icelandic", "lithuanian", "lsorbian", "naustrian", "serbian",
+"serbian-latin", "slovak", "slovene", "usorbian", 0};
+
+const char * const known_polish_quotes_languages[] = {"afrikaans", "croatian",
+"dutch", "estonian", "magyar", "polish", "romanian", 0};
+
+const char * const known_swedish_quotes_languages[] = {"bahasa", "finnish",
+"swedish", 0};
+
char const * const known_fontsizes[] = { "10pt", "11pt", "12pt", 0 };
const char * const known_roman_fonts[] = { "ae", "bookman", "charter",
string h_spacing = "single";
string h_papersize = "default";
string h_use_geometry = "false";
-string h_use_amsmath = "0";
+string h_use_amsmath = "1";
+string h_use_esint = "1";
string h_cite_engine = "basic";
string h_use_bibtopic = "false";
string h_paperorientation = "portrait";
string h_output_changes = "false";
string h_margins = "";
-
-void handle_opt(vector<string> & opts, char const * const * what, string & target)
+// returns true if at least one of the options in what has been found
+bool handle_opt(vector<string> & opts, char const * const * what, string & target)
{
if (opts.empty())
- return;
+ return false;
+ bool found = false;
// the last language option is the document language (for babel and LyX)
// the last size option is the document font size
vector<string>::iterator it;
for (; *what; ++what) {
it = find(opts.begin(), opts.end(), *what);
if (it != opts.end()) {
- documentclass_language = true;
if (it >= position) {
+ found = true;
target = *what;
position = it;
}
}
}
+ return found;
}
return;
// remove found options from the list
- // do this after handle_opt to avoid potential memory leaks and to be able
- // to find in every case the last language option
+ // do this after handle_opt to avoid potential memory leaks
vector<string>::iterator it;
for (; *what; ++what) {
it = find(opts.begin(), opts.end(), *what);
}
-void handle_package(string const & name, string const & opts)
+void handle_package(Parser &p, string const & name, string const & opts,
+ bool in_lyx_preamble)
{
vector<string> options = split_options(opts);
add_package(name, options);
string scale;
// roman fonts
- if (is_known(name, known_roman_fonts))
+ if (is_known(name, known_roman_fonts)) {
h_font_roman = name;
+ p.skip_spaces();
+ }
if (name == "fourier") {
h_font_roman = "utopia";
if (opts == "expert")
h_font_sc = "true";
}
+
if (name == "mathpazo")
h_font_roman = "palatino";
h_font_sf_scale = scale_as_percentage(scale);
}
}
+
// typewriter fonts
if (is_known(name, known_typewriter_fonts)) {
h_font_typewriter = name;
h_font_tt_scale = scale_as_percentage(scale);
}
}
+
// font uses old-style figure
if (name == "eco")
h_font_osf = "true";
else if (name == "amsmath" || name == "amssymb")
- h_use_amsmath = "1";
+ h_use_amsmath = "2";
+
+ else if (name == "esint")
+ h_use_esint = "2";
else if (name == "babel" && !opts.empty()) {
// check if more than one option was used - used later for inputenc
one_language = false;
h_inputencoding = "auto";
}
- // only set the document language when there was not already one set
- // via the documentclass options
- // babel takes the the last language given in the documentclass options
- // as document language. If there is no such language option, the last
- // option of its \usepackage call is used.
- if (documentclass_language == false) {
- handle_opt(options, known_languages, h_language);
- delete_opt(options, known_languages);
- 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_russian_languages))
- h_language = "russian";
- else if (is_known(h_language, known_ukrainian_languages))
- h_language = "ukrainian";
- h_quotes_language = h_language;
- }
+ // babel takes the last language of the option of its \usepackage
+ // call as document language. If there is no such language option, the
+ // last language in the documentclass options is used.
+ handle_opt(options, known_languages, h_language);
+ delete_opt(options, known_languages);
}
+
else if (name == "fontenc")
- ; // ignore this
+ ;// ignore this
else if (name == "inputenc") {
- // only set when there is not more than one inputenc option
- // therefore check for the "," character
- // also only set when there is not more then one babel language option
+ // h_inputencoding is only set when there is not more than one
+ // inputenc option because otherwise h_inputencoding must be
+ // set to "auto" (the default encoding of the document language)
+ // Therefore check for the "," character.
+ // It is also only set when there is not more then one babel
+ // language option but this is handled in the routine for babel.
if (opts.find(",") == string::npos && one_language == true)
- if (opts == "ascii")
- //change ascii to auto to be in the unicode range, see
- //http://bugzilla.lyx.org/show_bug.cgi?id=4719
- h_inputencoding = "auto";
- else
- h_inputencoding = opts;
+ h_inputencoding = opts;
+ if (!options.empty())
+ p.setEncoding(options.back());
options.clear();
}
+
else if (name == "makeidx")
; // ignore this
+ else if (name == "prettyref")
+ ; // ignore this
+
+ else if (name == "varioref")
+ ; // ignore this
+
else if (name == "verbatim")
; // ignore this
- else if (name == "color")
+ else if (name == "nomencl")
+ ; // ignore this
+
+ else if (name == "textcomp")
+ ; // ignore this
+
+ else if (name == "url")
+ ; // ignore this
+
+ else if (name == "color") {
// with the following command this package is only loaded when needed for
// undefined colors, since we only support the predefined colors
h_preamble << "\\@ifundefined{definecolor}\n {\\usepackage{color}}{}\n";
+ }
else if (name == "graphicx")
; // ignore this
; // Ignore this, the geometry settings are made by the \geometry
// command. This command is handled below.
- else if (is_known(name, known_languages)) {
- if (is_known(name, known_french_languages))
- h_language = "french";
- else if (is_known(name, known_german_languages))
- h_language = "german";
- else if (is_known(name, known_ngerman_languages))
- h_language = "ngerman";
- else if (is_known(name, known_russian_languages))
- h_language = "russian";
- else if (is_known(name, known_ukrainian_languages))
- h_language = "ukrainian";
- else
- h_language = name;
- h_quotes_language = h_language;
- }
+ else if (is_known(name, known_languages))
+ h_language = name;
+
else if (name == "natbib") {
h_cite_engine = "natbib_authoryear";
vector<string>::iterator it =
}
}
}
+
else if (name == "jurabib")
h_cite_engine = "jurabib";
- else if (options.empty())
- h_preamble << "\\usepackage{" << name << "}\n";
- else {
- h_preamble << "\\usepackage[" << opts << "]{" << name << "}\n";
- options.clear();
+ else if (!in_lyx_preamble) {
+ if (options.empty())
+ h_preamble << "\\usepackage{" << name << "}";
+ else {
+ h_preamble << "\\usepackage[" << opts << "]{"
+ << name << "}";
+ options.clear();
+ }
}
+
// We need to do something with the options...
if (!options.empty())
cerr << "Ignoring options '" << join(options, ",")
<< "' of package " << name << '.' << endl;
+
+ // remove the whitespace
+ p.skip_spaces();
}
void end_preamble(ostream & os, TextClass const & /*textclass*/)
{
+ // merge synonym languages
+ 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";
+
+ // set the quote language
+ // LyX only knows the following quotes languages:
+ // english, swedish, german, polish, french and danish
+ // (quotes for "japanese" and "chinese-traditional" are missing because
+ // they wouldn't be useful: http://www.lyx.org/trac/ticket/6383)
+ // conversion list taken from
+ // http://en.wikipedia.org/wiki/Quotation_mark,_non-English_usage
+ // (quotes for kazakh and interlingua are unknown)
+ // danish
+ 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))
+ 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))
+ h_quotes_language = "german";
+ // polish
+ else if (is_known(h_language, known_polish_quotes_languages))
+ h_quotes_language = "polish";
+ // swedish
+ 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_brazilian_languages)
+ || is_known(h_language, known_portuguese_languages))
+ h_quotes_language = "english";
+
+ // output the LyX file settings
os << "#LyX file created by tex2lyx " << PACKAGE_VERSION << "\n"
- << "\\lyxformat 247\n"
+ << "\\lyxformat 264\n"
<< "\\begin_document\n"
<< "\\begin_header\n"
<< "\\textclass " << h_textclass << "\n";
<< "\\papersize " << h_papersize << "\n"
<< "\\use_geometry " << h_use_geometry << "\n"
<< "\\use_amsmath " << h_use_amsmath << "\n"
+ << "\\use_esint " << h_use_esint << "\n"
<< "\\cite_engine " << h_cite_engine << "\n"
<< "\\use_bibtopic " << h_use_bibtopic << "\n"
<< "\\paperorientation " << h_paperorientation << "\n"
// initialize fixed types
special_columns['D'] = 3;
bool is_full_document = false;
+ bool is_lyx_file = false;
+ bool in_lyx_preamble = false;
// determine whether this is a full document or a fragment for inclusion
while (p.good()) {
//
// cat codes
//
- if (t.cat() == catLetter ||
- t.cat() == catSuper ||
- t.cat() == catSub ||
- t.cat() == catOther ||
- t.cat() == catMath ||
- t.cat() == catActive ||
- t.cat() == catBegin ||
- t.cat() == catEnd ||
- t.cat() == catAlign ||
- t.cat() == catParameter)
- h_preamble << t.character();
-
- else if (t.cat() == catSpace || t.cat() == catNewline)
+ if (!in_lyx_preamble &&
+ (t.cat() == catLetter ||
+ t.cat() == catSuper ||
+ t.cat() == catSub ||
+ t.cat() == catOther ||
+ t.cat() == catMath ||
+ t.cat() == catActive ||
+ t.cat() == catBegin ||
+ t.cat() == catEnd ||
+ t.cat() == catAlign ||
+ t.cat() == catParameter))
+ h_preamble << t.cs();
+
+ else if (!in_lyx_preamble &&
+ (t.cat() == catSpace || t.cat() == catNewline))
h_preamble << t.asInput();
- else if (t.cat() == catComment)
- h_preamble << t.asInput();
+ else if (t.cat() == catComment) {
+ // regex to parse comments (currently not used)
+ static regex const islyxfile("%% LyX .* created this file");
+ static regex const usercommands("User specified LaTeX commands");
+
+ string const comment = t.asInput();
+
+ // magically switch encoding default if it looks like XeLaTeX
+ static string const magicXeLaTeX =
+ "% This document must be compiled with XeLaTeX ";
+ if (comment.size() > magicXeLaTeX.size()
+ && comment.substr(0, magicXeLaTeX.size()) == magicXeLaTeX
+ && h_inputencoding == "auto") {
+ cerr << "XeLaTeX comment found, switching to UTF8\n";
+ h_inputencoding = "utf8";
+ }
+ smatch sub;
+ if (regex_search(comment, sub, islyxfile)) {
+ is_lyx_file = true;
+ in_lyx_preamble = true;
+ } else if (is_lyx_file
+ && regex_search(comment, sub, usercommands))
+ in_lyx_preamble = false;
+ else if (!in_lyx_preamble)
+ h_preamble << t.asInput();
+ }
else if (t.cs() == "pagestyle")
h_paperpagestyle = p.verbatim_item();
else if (t.cs() == "makeatletter") {
+ // LyX takes care of this
p.setCatCode('@', catLetter);
}
else if (t.cs() == "makeatother") {
+ // LyX takes care of this
p.setCatCode('@', catOther);
}
else if (t.cs() == "newcommand" || t.cs() == "renewcommand"
- || t.cs() == "providecommand") {
+ || t.cs() == "providecommand"
+ || t.cs() == "DeclareRobustCommand"
+ || t.cs() == "ProvideTextCommandDefault"
+ || t.cs() == "DeclareMathAccent") {
bool star = false;
if (p.next_token().character() == '*') {
p.get_token();
if (name == "\\rmdefault")
if (is_known(body, known_roman_fonts))
h_font_roman = body;
-
if (name == "\\sfdefault")
if (is_known(body, known_sans_fonts))
h_font_sans = body;
-
if (name == "\\ttdefault")
if (is_known(body, known_typewriter_fonts))
h_font_typewriter = body;
-
if (name == "\\familydefault") {
string family = body;
// remove leading "\"
h_font_default_family = family.erase(0,1);
}
// only non-lyxspecific stuff
- if ( name != "\\noun"
- && name != "\\tabularnewline"
- && name != "\\LyX"
- && name != "\\lyxline"
- && name != "\\lyxaddress"
- && name != "\\lyxrightaddress"
- && name != "\\lyxdot"
- && name != "\\boldsymbol"
- && name != "\\lyxarrow"
- && name != "\\rmdefault"
- && name != "\\sfdefault"
- && name != "\\ttdefault"
- && name != "\\familydefault") {
+ if (!in_lyx_preamble) {
ostringstream ss;
ss << '\\' << t.cs();
if (star)
string::size_type i = h_paperfontsize.find("pt");
if (i != string::npos)
h_paperfontsize.erase(i);
- // to avoid that the babel options overwrite the documentclass options
- documentclass_language = false;
+ // The documentclass options are always parsed before the options
+ // of the babel call so that a language cannot overwrite the babel
+ // options.
handle_opt(opts, known_languages, h_language);
delete_opt(opts, known_languages);
- 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_russian_languages))
- h_language = "russian";
- else if (is_known(h_language, known_ukrainian_languages))
- h_language = "ukrainian";
- h_quotes_language = h_language;
+
// paper orientation
if ((it = find(opts.begin(), opts.end(), "landscape")) != opts.end()) {
h_paperorientation = "landscape";
opts.erase(it);
}
// paper sides
+ if ((it = find(opts.begin(), opts.end(), "oneside"))
+ != opts.end()) {
+ h_papersides = "1";
+ opts.erase(it);
+ }
if ((it = find(opts.begin(), opts.end(), "twoside"))
!= opts.end()) {
h_papersides = "2";
opts.erase(it);
}
// paper columns
+ if ((it = find(opts.begin(), opts.end(), "onecolumn"))
+ != opts.end()) {
+ h_papercolumns = "1";
+ opts.erase(it);
+ }
if ((it = find(opts.begin(), opts.end(), "twocolumn"))
!= opts.end()) {
h_papercolumns = "2";
else if (t.cs() == "usepackage") {
string const options = p.getArg('[', ']');
string const name = p.getArg('{', '}');
- if (options.empty() && name.find(',')) {
- vector<string> vecnames;
- split(name, vecnames, ',');
- vector<string>::const_iterator it = vecnames.begin();
- vector<string>::const_iterator end = vecnames.end();
- for (; it != end; ++it)
- handle_package(trim(*it), string());
- } else {
- handle_package(name, options);
- }
+ vector<string> vecnames;
+ split(name, vecnames, ',');
+ vector<string>::const_iterator it = vecnames.begin();
+ vector<string>::const_iterator end = vecnames.end();
+ for (; it != end; ++it)
+ handle_package(p, trim(*it), options,
+ in_lyx_preamble);
+ }
+
+ else if (t.cs() == "inputencoding") {
+ string const encoding = p.getArg('{','}');
+ h_inputencoding = encoding;
+ p.setEncoding(encoding);
}
else if (t.cs() == "newenvironment") {
ss << p.getOpt();
ss << '{' << p.verbatim_item() << '}';
ss << '{' << p.verbatim_item() << '}';
- if (name != "lyxcode" && name != "lyxlist" &&
- name != "lyxrightadress" &&
- name != "lyxaddress" && name != "lyxgreyedout")
+ if (!in_lyx_preamble)
h_preamble << ss.str();
}
else if (t.cs() == "def") {
string name = p.get_token().cs();
while (p.next_token().cat() != catBegin)
- name += p.get_token().asString();
- h_preamble << "\\def\\" << name << '{'
- << p.verbatim_item() << "}";
+ name += p.get_token().cs();
+ if (!in_lyx_preamble)
+ h_preamble << "\\def\\" << name << '{'
+ << p.verbatim_item() << "}";
}
else if (t.cs() == "newcolumntype") {
for (; *margin; ++margin) {
k += 1;
// search for the "=" in e.g. "lmargin=2cm" to get the value
- for(int i = 0; i < opts.size(); i++) {
+ for(size_t i = 0; i != opts.size(); i++) {
if (opts.at(i).find(*margin) != string::npos) {
string::size_type pos = opts.at(i).find("=");
string value = opts.at(i).substr(pos + 1);
}
}
- else if (!t.cs().empty())
+ else if (!t.cs().empty() && !in_lyx_preamble)
h_preamble << '\\' << t.cs();
+
+ // remove the whitespace
+ p.skip_spaces();
}
+
+ // remove the whitespace
p.skip_spaces();
// Force textclass if the user wanted it