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/FileName.h"
26 #include "support/filetools.h"
27 #include "support/lassert.h"
28 #include "support/lstrings.h"
29 #include "support/qstring_helpers.h"
30 #include "support/Messages.h"
36 using namespace lyx::support;
41 Language const * ignore_language = nullptr;
42 Language const * default_language = nullptr;
43 Language const * latex_language = nullptr;
44 Language const * reset_language = nullptr;
47 bool Language::isPolyglossiaExclusive() const
49 return babel().empty() && !polyglossia().empty() && required().empty();
53 bool Language::isBabelExclusive() const
55 return !babel().empty() && polyglossia().empty() && required().empty();
59 docstring const Language::translateLayout(string const & m) const
65 lyxerr << "Warning: not translating `" << m
66 << "' because it is not pure ASCII.\n";
70 TranslationMap::const_iterator it = layoutTranslations_.find(m);
71 if (it != layoutTranslations_.end())
74 docstring t = from_ascii(m);
80 string Language::fontenc(BufferParams const & params) const
82 // Don't use LaTeX fonts, so just return the language's preferred
83 // (although this is not used with nonTeXFonts anyway).
84 if (params.useNonTeXFonts)
85 return fontenc_.front() == "ASCII" ? "T1" : fontenc_.front();
87 // Determine optimal font encoding
88 // We check whether the used rm font supports an encoding our language supports
89 LaTeXFont const & lf =
90 theLaTeXFonts().getLaTeXFont(from_ascii(params.fontsRoman()));
91 vector<string> const lfe = lf.fontencs();
92 for (auto & fe : fontenc_) {
93 // ASCII means: support all T* encodings plus OT1
95 for (auto & afe : lfe) {
96 if (afe == "OT1" || prefixIs(afe, "T"))
97 // we found a suitable one; return that.
101 // For other encodings, just check whether the font supports it
102 if (lf.hasFontenc(fe))
105 // We did not find a suitable one; just take the first in the list,
106 // the priorized one (which is "T1" for ASCII).
107 return fontenc_.front() == "ASCII" ? "T1" : fontenc_.front();
111 string Language::dateFormat(size_t i) const
113 if (i > dateformats_.size())
115 return dateformats_.at(i);
119 docstring Language::decimalSeparator() const
121 if (lyxrc.default_decimal_sep == "locale") {
122 QLocale loc = QLocale(toqstr(code()));
123 return qstring_to_ucs4(QString(loc.decimalPoint()));
125 return from_utf8(lyxrc.default_decimal_sep);
129 bool Language::readLanguage(Lexer & lex)
145 LA_POSTBABELPREAMBLE,
155 // Keep these sorted alphabetically!
156 LexerKeyword languageTags[] = {
157 { "activechars", LA_ACTIVECHARS },
158 { "babelname", LA_BABELNAME },
159 { "dateformats", LA_DATEFORMATS },
160 { "encoding", LA_ENCODING },
162 { "fontencoding", LA_FONTENC },
163 { "guiname", LA_GUINAME },
164 { "hasguisupport", LA_HAS_GUI_SUPPORT },
165 { "internalencoding", LA_INTERNAL_ENC },
166 { "langcode", LA_LANG_CODE },
167 { "langvariety", LA_LANG_VARIETY },
168 { "polyglossianame", LA_POLYGLOSSIANAME },
169 { "polyglossiaopts", LA_POLYGLOSSIAOPTS },
170 { "postbabelpreamble", LA_POSTBABELPREAMBLE },
171 { "prebabelpreamble", LA_PREBABELPREAMBLE },
172 { "provides", LA_PROVIDES },
173 { "quotestyle", LA_QUOTESTYLE },
174 { "requires", LA_REQUIRES },
176 { "wordwrap", LA_WORDWRAP },
177 { "xindyname", LA_XINDYNAME }
181 bool finished = false;
182 lex.pushTable(languageTags);
183 // parse style section
184 while (!finished && lex.isOK() && !error) {
186 // See comment in LyXRC.cpp.
188 case Lexer::LEX_FEOF:
191 case Lexer::LEX_UNDEF: // parse error
192 lex.printError("Unknown language tag `$$Token'");
199 switch (static_cast<LanguageTags>(le)) {
200 case LA_END: // end of structure
206 case LA_POLYGLOSSIANAME:
207 lex >> polyglossia_name_;
209 case LA_POLYGLOSSIAOPTS:
210 lex >> polyglossia_opts_;
219 lex >> active_chars_;
226 vector<string> const fe =
227 getVectorFromString(lex.getString(true), "|");
228 fontenc_.insert(fontenc_.end(), fe.begin(), fe.end());
231 case LA_DATEFORMATS: {
233 vector<string> const df =
234 getVectorFromString(trim(lex.getString(true), "\""), "|");
235 dateformats_.insert(dateformats_.end(), df.begin(), df.end());
241 case LA_HAS_GUI_SUPPORT:
242 lex >> has_gui_support_;
244 case LA_INTERNAL_ENC:
245 lex >> internal_enc_;
250 case LA_LANG_VARIETY:
253 case LA_POSTBABELPREAMBLE:
254 babel_postsettings_ =
255 lex.getLongString(from_ascii("EndPostBabelPreamble"));
257 case LA_PREBABELPREAMBLE:
259 lex.getLongString(from_ascii("EndPreBabelPreamble"));
276 return finished && !error;
280 bool Language::read(Lexer & lex)
283 internal_enc_ = false;
284 rightToLeft_ = false;
287 lex.printError("No name given for language: `$$Token'.");
291 lang_ = lex.getString();
292 LYXERR(Debug::INFO, "Reading language " << lang_);
293 if (!readLanguage(lex)) {
294 LYXERR0("Error parsing language `" << lang_ << '\'');
298 encoding_ = encodings.fromLyXName(encodingStr_);
299 if (!encoding_ && !encodingStr_.empty()) {
300 encoding_ = encodings.fromLyXName("iso8859-1");
301 LYXERR0("Unknown encoding " << encodingStr_);
303 if (fontenc_.empty())
304 fontenc_.push_back("ASCII");
305 if (dateformats_.empty()) {
306 dateformats_.push_back("MMMM dd, yyyy");
307 dateformats_.push_back("MMM dd, yyyy");
308 dateformats_.push_back("M/d/yyyy");
314 void Language::readLayoutTranslations(Language::TranslationMap const & trans, bool replace)
316 for (auto const & t : trans) {
318 || layoutTranslations_.find(t.first) == layoutTranslations_.end())
319 layoutTranslations_[t.first] = t.second;
324 void Languages::read(FileName const & filename)
327 lex.setFile(filename);
328 lex.setContext("Languages::read");
332 case Lexer::LEX_FEOF:
338 if (lex.getString() != "Language") {
339 lex.printError("Unknown Language tag `$$Token'");
346 if (l.lang() == "latex") {
347 // Check if latex language was not already defined.
348 LASSERT(latex_language == nullptr, continue);
349 static const Language latex_lang = l;
350 latex_language = &latex_lang;
351 } else if (l.lang() == "ignore") {
352 // Check if ignore language was not already defined.
353 LASSERT(ignore_language == nullptr, continue);
354 static const Language ignore_lang = l;
355 ignore_language = &ignore_lang;
357 languagelist_[l.lang()] = l;
360 default_language = getLanguage("english");
361 if (!default_language) {
362 LYXERR0("Default language \"english\" not found!");
363 default_language = &(*languagelist_.begin()).second;
364 LYXERR0("Using \"" << default_language->lang() << "\" instead!");
367 // Read layout translations
368 FileName const path = libFileSearch(string(), "layouttranslations");
369 readLayoutTranslations(path);
375 bool readTranslations(Lexer & lex, Language::TranslationMap & trans)
378 if (lex.checkFor("End"))
382 string const key = lex.getString();
385 docstring const val = lex.getDocString();
395 VeryApproximateMatch,
400 Match match(string const & code, Language const & lang)
402 // we need to mimic gettext: code can be a two-letter code, which
403 // should match all variants, e.g. "de" should match "de_DE",
405 // special case for chinese:
406 // simplified => code == "zh_CN", langcode == "zh_CN"
407 // traditional => code == "zh_TW", langcode == "zh_CN"
408 string const variety = lang.variety();
409 string const langcode = variety.empty() ?
410 lang.code() : lang.code() + '_' + variety;
411 string const name = lang.lang();
412 if ((code == langcode && name != "chinese-traditional")
413 || (code == "zh_TW" && name == "chinese-traditional"))
415 if ((code.size() == 2) && (langcode.size() > 2)
416 && (code + '_' == langcode.substr(0, 3)))
417 return ApproximateMatch;
418 if (code.substr(0,2) == langcode.substr(0,2))
419 return VeryApproximateMatch;
427 Language const * Languages::getFromCode(string const & code) const
429 // 1/ exact match with any known language
430 for (auto const & l : languagelist_) {
431 if (match(code, l.second) == ExactMatch)
435 // 2/ approximate with any known language
436 for (auto const & l : languagelist_) {
437 if (match(code, l.second) == ApproximateMatch)
444 Language const * Languages::getFromCode(string const & code,
445 set<Language const *> const & tryfirst) const
447 // 1/ exact match with tryfirst list
448 for (auto const * lptr : tryfirst) {
449 if (match(code, *lptr) == ExactMatch)
453 // 2/ approximate match with tryfirst list
454 for (auto const * lptr : tryfirst) {
455 Match const m = match(code, *lptr);
456 if (m == ApproximateMatch || m == VeryApproximateMatch)
460 // 3/ stricter match in all languages
461 return getFromCode(code);
463 LYXERR0("Unknown language `" << code << "'");
468 void Languages::readLayoutTranslations(support::FileName const & filename)
471 lex.setFile(filename);
472 lex.setContext("Languages::read");
474 // 1) read all translations (exact and approximate matches) into trans
475 std::map<string, Language::TranslationMap> trans;
477 if (!lex.checkFor("Translation")) {
479 lex.printError("Unknown layout translation tag `$$Token'");
484 string const code = lex.getString();
485 bool found = getFromCode(code);
487 lex.printError("Unknown language `" + code + "'");
490 if (!readTranslations(lex, trans[code])) {
491 lex.printError("Could not read layout translations for language `"
497 // 2) merge all translations into the languages
498 // exact translations overwrite approximate ones
499 for (auto & tr : trans) {
500 for (auto & lang : languagelist_) {
501 Match const m = match(tr.first, lang.second);
504 lang.second.readLayoutTranslations(tr.second, m == ExactMatch);
511 Language const * Languages::getLanguage(string const & language) const
513 if (language == "reset")
514 return reset_language;
515 if (language == "ignore")
516 return ignore_language;
517 const_iterator it = languagelist_.find(language);
518 return it == languagelist_.end() ? reset_language : &it->second;