3 * This file is part of LyX, the document processor.
4 * Licence details can be found in the file COPYING.
6 * \author Lars Gullik Bjønnes
7 * \author Jean-Marc Lasgouttes
8 * \author Jürgen Spitzmüller
11 * Full author contact details are available in file CREDITS.
18 #include "BufferParams.h"
20 #include "LaTeXFonts.h"
24 #include "support/debug.h"
25 #include "support/docstring.h"
26 #include "support/FileName.h"
27 #include "support/filetools.h"
28 #include "support/lassert.h"
29 #include "support/lstrings.h"
30 #include "support/qstring_helpers.h"
31 #include "support/Messages.h"
37 using namespace lyx::support;
42 Language const * ignore_language = nullptr;
43 Language const * default_language = nullptr;
44 Language const * latex_language = nullptr;
45 Language const * reset_language = nullptr;
48 bool Language::isPolyglossiaExclusive() const
50 return babel().empty() && !polyglossia().empty() && required().empty();
54 bool Language::isBabelExclusive() const
56 return !babel().empty() && polyglossia().empty() && required().empty();
60 docstring const Language::translateLayout(string const & msg) const
66 lyxerr << "Warning: not translating `" << msg
67 << "' because it is not pure ASCII.\n";
68 return from_utf8(msg);
71 TranslationMap::const_iterator it = layoutTranslations_.find(msg);
72 if (it != layoutTranslations_.end())
75 docstring t = from_ascii(msg);
81 string Language::fontenc(BufferParams const & params) const
83 // Don't use LaTeX fonts, so just return the language's preferred
84 // (although this is not used with nonTeXFonts anyway).
85 if (params.useNonTeXFonts)
86 return fontenc_.front() == "ASCII" ? "T1" : fontenc_.front();
88 // Determine optimal font encoding
89 // We check whether the used rm font supports an encoding our language supports
90 LaTeXFont const & lf =
91 theLaTeXFonts().getLaTeXFont(from_ascii(params.fontsRoman()));
92 vector<string> const & lfe = lf.fontencs();
93 for (auto & fe : fontenc_) {
94 // ASCII means: support all T* encodings plus OT1
96 for (auto & afe : lfe) {
97 if (afe == "OT1" || prefixIs(afe, "T"))
98 // we found a suitable one; return that.
102 // For other encodings, just check whether the font supports it
103 if (lf.hasFontenc(fe))
106 // We did not find a suitable one; just take the first in the list,
107 // the priorized one (which is "T1" for ASCII).
108 return fontenc_.front() == "ASCII" ? "T1" : fontenc_.front();
112 string Language::dateFormat(size_t i) const
114 if (i > dateformats_.size())
116 return dateformats_.at(i);
120 docstring Language::decimalSeparator() const
122 if (lyxrc.default_decimal_sep == "locale") {
123 QLocale loc = QLocale(toqstr(code()));
124 return qstring_to_ucs4(QString(loc.decimalPoint()));
126 return from_utf8(lyxrc.default_decimal_sep);
130 bool Language::readLanguage(Lexer & lex)
146 LA_POSTBABELPREAMBLE,
156 // Keep these sorted alphabetically!
157 LexerKeyword languageTags[] = {
158 { "activechars", LA_ACTIVECHARS },
159 { "babelname", LA_BABELNAME },
160 { "dateformats", LA_DATEFORMATS },
161 { "encoding", LA_ENCODING },
163 { "fontencoding", LA_FONTENC },
164 { "guiname", LA_GUINAME },
165 { "hasguisupport", LA_HAS_GUI_SUPPORT },
166 { "internalencoding", LA_INTERNAL_ENC },
167 { "langcode", LA_LANG_CODE },
168 { "langvariety", LA_LANG_VARIETY },
169 { "polyglossianame", LA_POLYGLOSSIANAME },
170 { "polyglossiaopts", LA_POLYGLOSSIAOPTS },
171 { "postbabelpreamble", LA_POSTBABELPREAMBLE },
172 { "prebabelpreamble", LA_PREBABELPREAMBLE },
173 { "provides", LA_PROVIDES },
174 { "quotestyle", LA_QUOTESTYLE },
175 { "requires", LA_REQUIRES },
177 { "wordwrap", LA_WORDWRAP },
178 { "xindyname", LA_XINDYNAME }
182 bool finished = false;
183 lex.pushTable(languageTags);
184 // parse style section
185 while (!finished && lex.isOK() && !error) {
187 // See comment in LyXRC.cpp.
189 case Lexer::LEX_FEOF:
192 case Lexer::LEX_UNDEF: // parse error
193 lex.printError("Unknown language tag `$$Token'");
200 switch (static_cast<LanguageTags>(le)) {
201 case LA_END: // end of structure
207 case LA_POLYGLOSSIANAME:
208 lex >> polyglossia_name_;
210 case LA_POLYGLOSSIAOPTS:
211 lex >> polyglossia_opts_;
220 lex >> active_chars_;
227 vector<string> const fe =
228 getVectorFromString(lex.getString(true), "|");
229 fontenc_.insert(fontenc_.end(), fe.begin(), fe.end());
232 case LA_DATEFORMATS: {
234 vector<string> const df =
235 getVectorFromString(trim(lex.getString(true), "\""), "|");
236 dateformats_.insert(dateformats_.end(), df.begin(), df.end());
242 case LA_HAS_GUI_SUPPORT:
243 lex >> has_gui_support_;
245 case LA_INTERNAL_ENC:
246 lex >> internal_enc_;
251 case LA_LANG_VARIETY:
254 case LA_POSTBABELPREAMBLE:
255 babel_postsettings_ =
256 lex.getLongString(from_ascii("EndPostBabelPreamble"));
258 case LA_PREBABELPREAMBLE:
260 lex.getLongString(from_ascii("EndPreBabelPreamble"));
277 return finished && !error;
281 bool Language::read(Lexer & lex)
284 internal_enc_ = false;
285 rightToLeft_ = false;
288 lex.printError("No name given for language: `$$Token'.");
292 lang_ = lex.getString();
293 LYXERR(Debug::INFO, "Reading language " << lang_);
294 if (!readLanguage(lex)) {
295 LYXERR0("Error parsing language `" << lang_ << '\'');
299 encoding_ = encodings.fromLyXName(encodingStr_);
300 if (!encoding_ && !encodingStr_.empty()) {
301 encoding_ = encodings.fromLyXName("iso8859-1");
302 LYXERR0("Unknown encoding " << encodingStr_);
304 if (fontenc_.empty())
305 fontenc_.push_back("ASCII");
306 if (dateformats_.empty()) {
307 dateformats_.push_back("MMMM dd, yyyy");
308 dateformats_.push_back("MMM dd, yyyy");
309 dateformats_.push_back("M/d/yyyy");
315 void Language::readLayoutTranslations(Language::TranslationMap const & trans, bool replace)
317 for (auto const & t : trans) {
319 || layoutTranslations_.find(t.first) == layoutTranslations_.end())
320 layoutTranslations_[t.first] = t.second;
325 void Languages::read(FileName const & filename)
328 lex.setFile(filename);
329 lex.setContext("Languages::read");
333 case Lexer::LEX_FEOF:
339 if (lex.getString() != "Language") {
340 lex.printError("Unknown Language tag `$$Token'");
347 if (l.lang() == "latex") {
348 // Check if latex language was not already defined.
349 LASSERT(latex_language == nullptr, continue);
350 static const Language latex_lang = l;
351 latex_language = &latex_lang;
352 } else if (l.lang() == "ignore") {
353 // Check if ignore language was not already defined.
354 LASSERT(ignore_language == nullptr, continue);
355 static const Language ignore_lang = l;
356 ignore_language = &ignore_lang;
358 languagelist_[l.lang()] = l;
361 default_language = getLanguage("english");
362 if (!default_language) {
363 LYXERR0("Default language \"english\" not found!");
364 default_language = &(*languagelist_.begin()).second;
365 LYXERR0("Using \"" << default_language->lang() << "\" instead!");
368 // Read layout translations
369 FileName const path = libFileSearch(string(), "layouttranslations");
370 readLayoutTranslations(path);
376 bool readTranslations(Lexer & lex, Language::TranslationMap & trans)
379 if (lex.checkFor("End"))
383 string const key = lex.getString();
386 docstring const val = lex.getDocString();
396 VeryApproximateMatch,
401 Match match(string const & code, Language const & lang)
403 // we need to mimic gettext: code can be a two-letter code, which
404 // should match all variants, e.g. "de" should match "de_DE",
406 // special case for chinese:
407 // simplified => code == "zh_CN", langcode == "zh_CN"
408 // traditional => code == "zh_TW", langcode == "zh_CN"
409 string const variety = lang.variety();
410 string const langcode = variety.empty() ?
411 lang.code() : lang.code() + '_' + variety;
412 string const name = lang.lang();
413 if ((code == langcode && name != "chinese-traditional")
414 || (code == "zh_TW" && name == "chinese-traditional"))
416 if ((code.size() == 2) && (langcode.size() > 2)
417 && (code + '_' == langcode.substr(0, 3)))
418 return ApproximateMatch;
419 if (code.substr(0,2) == langcode.substr(0,2))
420 return VeryApproximateMatch;
428 Language const * Languages::getFromCode(string const & code) const
430 // 1/ exact match with any known language
431 for (auto const & l : languagelist_) {
432 if (match(code, l.second) == ExactMatch)
436 // 2/ approximate with any known language
437 for (auto const & l : languagelist_) {
438 if (match(code, l.second) == ApproximateMatch)
445 Language const * Languages::getFromCode(string const & code,
446 set<Language const *> const & tryfirst) const
448 // 1/ exact match with tryfirst list
449 for (auto const * lptr : tryfirst) {
450 if (match(code, *lptr) == ExactMatch)
454 // 2/ approximate match with tryfirst list
455 for (auto const * lptr : tryfirst) {
456 Match const m = match(code, *lptr);
457 if (m == ApproximateMatch || m == VeryApproximateMatch)
461 // 3/ stricter match in all languages
462 return getFromCode(code);
464 LYXERR0("Unknown language `" << code << "'");
469 void Languages::readLayoutTranslations(support::FileName const & filename)
472 lex.setFile(filename);
473 lex.setContext("Languages::read");
475 // 1) read all translations (exact and approximate matches) into trans
476 std::map<string, Language::TranslationMap> trans;
478 if (!lex.checkFor("Translation")) {
480 lex.printError("Unknown layout translation tag `$$Token'");
485 string const code = lex.getString();
486 bool found = getFromCode(code);
488 lex.printError("Unknown language `" + code + "'");
491 if (!readTranslations(lex, trans[code])) {
492 lex.printError("Could not read layout translations for language `"
498 // 2) merge all translations into the languages
499 // exact translations overwrite approximate ones
500 for (auto & tr : trans) {
501 for (auto & lang : languagelist_) {
502 Match const m = match(tr.first, lang.second);
505 lang.second.readLayoutTranslations(tr.second, m == ExactMatch);
512 Language const * Languages::getLanguage(string const & language) const
514 if (language == "reset")
515 return reset_language;
516 if (language == "ignore")
517 return ignore_language;
518 const_iterator it = languagelist_.find(language);
519 return it == languagelist_.end() ? reset_language : &it->second;