X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FEncoding.cpp;h=d0610efb2e2af646dd60f2355dab5d65744bfa42;hb=9e645a5cfc9d6c3e66d627002e68384c92c4d0b8;hp=a6f6a60ea3162144c5711fed6c7853658502c9d6;hpb=3013ce3716603e751152e162a5b155cd455c745e;p=lyx.git diff --git a/src/Encoding.cpp b/src/Encoding.cpp index a6f6a60ea3..d0610efb2e 100644 --- a/src/Encoding.cpp +++ b/src/Encoding.cpp @@ -17,15 +17,17 @@ #include "Lexer.h" #include "support/debug.h" +#include "support/docstring.h" #include "support/gettext.h" #include "support/lstrings.h" +#include "support/mutex.h" #include "support/textutils.h" #include "support/unicode.h" -#include - -#include #include +#include +#include +#include using namespace std; using namespace lyx::support; @@ -42,197 +44,13 @@ Encodings::MathSymbolSet Encodings::mathsym; namespace { -char_type arabic_table[172][4] = { - {0xfe80, 0xfe80, 0xfe80, 0xfe80}, // 0x0621 = hamza - {0xfe81, 0xfe82, 0xfe81, 0xfe82}, // 0x0622 = ligature madda on alef - {0xfe83, 0xfe84, 0xfe83, 0xfe84}, // 0x0623 = ligature hamza on alef - {0xfe85, 0xfe86, 0xfe85, 0xfe86}, // 0x0624 = ligature hamza on waw - {0xfe87, 0xfe88, 0xfe87, 0xfe88}, // 0x0625 = ligature hamza under alef - {0xfe89, 0xfe8a, 0xfe8b, 0xfe8c}, // 0x0626 = ligature hamza on ya - {0xfe8d, 0xfe8e, 0xfe8d, 0xfe8e}, // 0x0627 = alef - {0xfe8f, 0xfe90, 0xfe91, 0xfe92}, // 0x0628 = baa - {0xfe93, 0xfe94, 0xfe93, 0xfe94}, // 0x0629 = taa marbuta - {0xfe95, 0xfe96, 0xfe97, 0xfe98}, // 0x062a = taa - {0xfe99, 0xfe9a, 0xfe9b, 0xfe9c}, // 0x062b = thaa - {0xfe9d, 0xfe9e, 0xfe9f, 0xfea0}, // 0x062c = jeem - {0xfea1, 0xfea2, 0xfea3, 0xfea4}, // 0x062d = haa - {0xfea5, 0xfea6, 0xfea7, 0xfea8}, // 0x062e = khaa - {0xfea9, 0xfeaa, 0xfea9, 0xfeaa}, // 0x062f = dal - - {0xfeab, 0xfeac, 0xfeab, 0xfeac}, // 0x0630 = thal - {0xfead, 0xfeae, 0xfead, 0xfeae}, // 0x0631 = ra - {0xfeaf, 0xfeb0, 0xfeaf, 0xfeb0}, // 0x0632 = zain - {0xfeb1, 0xfeb2, 0xfeb3, 0xfeb4}, // 0x0633 = seen - {0xfeb5, 0xfeb6, 0xfeb7, 0xfeb8}, // 0x0634 = sheen - {0xfeb9, 0xfeba, 0xfebb, 0xfebc}, // 0x0635 = sad - {0xfebd, 0xfebe, 0xfebf, 0xfec0}, // 0x0636 = dad - {0xfec1, 0xfec2, 0xfec3, 0xfec4}, // 0x0637 = tah - {0xfec5, 0xfec6, 0xfec7, 0xfec8}, // 0x0638 = zah - {0xfec9, 0xfeca, 0xfecb, 0xfecc}, // 0x0639 = ain - {0xfecd, 0xfece, 0xfecf, 0xfed0}, // 0x063a = ghain - {0, 0, 0, 0}, // 0x063b - {0, 0, 0, 0}, // 0x063c - {0, 0, 0, 0}, // 0x063d - {0, 0, 0, 0}, // 0x063e - {0, 0, 0, 0}, // 0x063f - - {0, 0, 0, 0}, // 0x0640 - {0xfed1, 0xfed2, 0xfed3, 0xfed4}, // 0x0641 = fa - {0xfed5, 0xfed6, 0xfed7, 0xfed8}, // 0x0642 = qaf - {0xfed9, 0xfeda, 0xfedb, 0xfedc}, // 0x0643 = kaf - {0xfedd, 0xfede, 0xfedf, 0xfee0}, // 0x0644 = lam - {0xfee1, 0xfee2, 0xfee3, 0xfee4}, // 0x0645 = meem - {0xfee5, 0xfee6, 0xfee7, 0xfee8}, // 0x0646 = noon - {0xfee9, 0xfeea, 0xfeeb, 0xfeec}, // 0x0647 = ha - {0xfeed, 0xfeee, 0xfeed, 0xfeee}, // 0x0648 = waw - {0xfeef, 0xfef0, 0xfeef, 0xfef0}, // 0x0649 = alef maksura - {0xfef1, 0xfef2, 0xfef3, 0xfef4}, // 0x064a = ya - {0x065b, 0x065b, 0x065b, 0x065b}, // 0x064b = fathatan - {0x065c, 0x065c, 0x065c, 0x065c}, // 0x064c = dammatan - {0x064d, 0x064d, 0x064d, 0x064d}, // 0x064d = kasratan - {0x064e, 0x064e, 0x064e, 0x064e}, // 0x064e = fatha - {0x064f, 0x064f, 0x064f, 0x064f}, // 0x064f = damma - - {0x0650, 0x0650, 0x0650, 0x0650}, // 0x0650 = kasra - {0x0651, 0x0651, 0x0651, 0x0651}, // 0x0651 = shadda - {0x0652, 0x0652, 0x0652, 0x0652}, // 0x0652 = sukun - - {0, 0, 0, 0}, // 0x0653 - {0, 0, 0, 0}, // 0x0654 - {0, 0, 0, 0}, // 0x0655 - {0, 0, 0, 0}, // 0x0656 - {0, 0, 0, 0}, // 0x0657 - {0, 0, 0, 0}, // 0x0658 - {0, 0, 0, 0}, // 0x0659 - {0, 0, 0, 0}, // 0x065a - {0, 0, 0, 0}, // 0x065b - {0, 0, 0, 0}, // 0x065c - {0, 0, 0, 0}, // 0x065d - {0, 0, 0, 0}, // 0x065e - {0, 0, 0, 0}, // 0x065f - {0, 0, 0, 0}, // 0x0660 - {0, 0, 0, 0}, // 0x0661 - {0, 0, 0, 0}, // 0x0662 - {0, 0, 0, 0}, // 0x0663 - {0, 0, 0, 0}, // 0x0664 - {0, 0, 0, 0}, // 0x0665 - {0, 0, 0, 0}, // 0x0666 - {0, 0, 0, 0}, // 0x0667 - {0, 0, 0, 0}, // 0x0668 - {0, 0, 0, 0}, // 0x0669 - {0, 0, 0, 0}, // 0x066a - {0, 0, 0, 0}, // 0x066b - {0, 0, 0, 0}, // 0x066c - {0, 0, 0, 0}, // 0x066d - {0, 0, 0, 0}, // 0x066e - {0, 0, 0, 0}, // 0x066f - {0, 0, 0, 0}, // 0x0670 - {0, 0, 0, 0}, // 0x0671 - {0, 0, 0, 0}, // 0x0672 - {0, 0, 0, 0}, // 0x0673 - {0, 0, 0, 0}, // 0x0674 - {0, 0, 0, 0}, // 0x0675 - {0, 0, 0, 0}, // 0x0676 - {0, 0, 0, 0}, // 0x0677 - {0, 0, 0, 0}, // 0x0678 - {0, 0, 0, 0}, // 0x0679 - {0, 0, 0, 0}, // 0x067a - {0, 0, 0, 0}, // 0x067b - {0, 0, 0, 0}, // 0x067c - {0, 0, 0, 0}, // 0x067d - {0xfb56, 0xfb57, 0xfb58, 0xfb59}, // 0x067e = peh - {0, 0, 0, 0}, // 0x067f - {0, 0, 0, 0}, // 0x0680 - {0, 0, 0, 0}, // 0x0681 - {0, 0, 0, 0}, // 0x0682 - {0, 0, 0, 0}, // 0x0683 - {0, 0, 0, 0}, // 0x0684 - {0, 0, 0, 0}, // 0x0685 - {0xfb7a, 0xfb7b, 0xfb7c, 0xfb7d}, // 0x0686 = tcheh - {0, 0, 0, 0}, // 0x0687 - {0, 0, 0, 0}, // 0x0688 - {0, 0, 0, 0}, // 0x0689 - {0, 0, 0, 0}, // 0x068a - {0, 0, 0, 0}, // 0x068b - {0, 0, 0, 0}, // 0x068c - {0, 0, 0, 0}, // 0x068d - {0, 0, 0, 0}, // 0x068e - {0, 0, 0, 0}, // 0x068f - {0, 0, 0, 0}, // 0x0690 - {0, 0, 0, 0}, // 0x0691 - {0, 0, 0, 0}, // 0x0692 - {0, 0, 0, 0}, // 0x0693 - {0, 0, 0, 0}, // 0x0694 - {0, 0, 0, 0}, // 0x0695 - {0, 0, 0, 0}, // 0x0696 - {0, 0, 0, 0}, // 0x0697 - {0xfb8a, 0xfb8b, 0xfb8a, 0xfb8b}, // 0x0698 = jeh - {0, 0, 0, 0}, // 0x0699 - {0, 0, 0, 0}, // 0x069a - {0, 0, 0, 0}, // 0x069b - {0, 0, 0, 0}, // 0x069c - {0, 0, 0, 0}, // 0x069d - {0, 0, 0, 0}, // 0x069e - {0, 0, 0, 0}, // 0x069f - {0, 0, 0, 0}, // 0x06a0 - {0, 0, 0, 0}, // 0x06a1 - {0, 0, 0, 0}, // 0x06a2 - {0, 0, 0, 0}, // 0x06a3 - {0, 0, 0, 0}, // 0x06a4 - {0, 0, 0, 0}, // 0x06a5 - {0, 0, 0, 0}, // 0x06a6 - {0, 0, 0, 0}, // 0x06a7 - {0, 0, 0, 0}, // 0x06a8 - {0xfb8e, 0xfb8f, 0xfb90, 0xfb91}, // 0x06a9 = farsi kaf - {0, 0, 0, 0}, // 0x06aa - {0, 0, 0, 0}, // 0x06ab - {0, 0, 0, 0}, // 0x06ac - {0, 0, 0, 0}, // 0x06ad - {0, 0, 0, 0}, // 0x06ae - {0xfb92, 0xfb93, 0xfb94, 0xfb95}, // 0x06af = gaf - {0, 0, 0, 0}, // 0x06b0 - {0, 0, 0, 0}, // 0x06b1 - {0, 0, 0, 0}, // 0x06b2 - {0, 0, 0, 0}, // 0x06b3 - {0, 0, 0, 0}, // 0x06b4 - {0, 0, 0, 0}, // 0x06b5 - {0, 0, 0, 0}, // 0x06b6 - {0, 0, 0, 0}, // 0x06b7 - {0, 0, 0, 0}, // 0x06b8 - {0, 0, 0, 0}, // 0x06b9 - {0, 0, 0, 0}, // 0x06ba - {0, 0, 0, 0}, // 0x06bb - {0, 0, 0, 0}, // 0x06bc - {0, 0, 0, 0}, // 0x06bd - {0, 0, 0, 0}, // 0x06be - {0, 0, 0, 0}, // 0x06bf - {0, 0, 0, 0}, // 0x06c0 - {0, 0, 0, 0}, // 0x06c1 - {0, 0, 0, 0}, // 0x06c2 - {0, 0, 0, 0}, // 0x06c3 - {0, 0, 0, 0}, // 0x06c4 - {0, 0, 0, 0}, // 0x06c5 - {0, 0, 0, 0}, // 0x06c6 - {0, 0, 0, 0}, // 0x06c7 - {0, 0, 0, 0}, // 0x06c8 - {0, 0, 0, 0}, // 0x06c9 - {0, 0, 0, 0}, // 0x06ca - {0, 0, 0, 0}, // 0x06cb - {0xfbfc, 0xfbfd, 0xfbfe, 0xfbff} // 0x06cc = farsi yeh -}; - - -char_type const arabic_start = 0x0621; -char_type const arabic_end = 0x06cc; - - typedef map CharInfoMap; CharInfoMap unicodesymbols; typedef set CharSet; typedef map CharSetMap; CharSet forced; -CharSetMap forcedselected; +CharSetMap forcedSelected; typedef set MathAlphaSet; MathAlphaSet mathalpha; @@ -241,7 +59,7 @@ MathAlphaSet mathalpha; /// The highest code point in UCS4 encoding (1<<20 + 1<<16) char_type const max_ucs4 = 0x110000; -} // namespace anon +} // namespace EncodingException::EncodingException(char_type c) @@ -250,26 +68,38 @@ EncodingException::EncodingException(char_type c) } -const char * EncodingException::what() const throw() +const char * EncodingException::what() const noexcept { return "Could not find LaTeX command for a character"; } CharInfo::CharInfo( - docstring const textcommand, docstring const mathcommand, - std::string const textpreamble, std::string const mathpreamble, - std::string const tipashortcut, unsigned int flags) - : textcommand_(textcommand), mathcommand_(mathcommand), - textpreamble_(textpreamble), mathpreamble_(mathpreamble), - tipashortcut_(tipashortcut), flags_(flags) + docstring const & text_command, docstring const & math_command, + std::string const & text_preamble, std::string const & math_preamble, + std::string const & tipa_shortcut, unsigned int flags) + : text_commands_({text_command}), math_commands_({math_command}), + text_preamble_(text_preamble), math_preamble_(math_preamble), + tipa_shortcut_(tipa_shortcut), flags_(flags) { } + +CharInfo::CharInfo( + std::vector const & text_commands, std::vector const & math_commands, + std::string const & text_preamble, std::string const & math_preamble, + std::string const & tipa_shortcut, unsigned int flags) + : text_commands_(text_commands), math_commands_(math_commands), + text_preamble_(text_preamble), math_preamble_(math_preamble), + tipa_shortcut_(tipa_shortcut), flags_(flags) +{ +} + + Encoding::Encoding(string const & n, string const & l, string const & g, string const & i, bool f, bool u, Encoding::Package p) : name_(n), latexName_(l), guiName_(g), iconvName_(i), fixedwidth_(f), - unsafe_(u), forced_(&forcedselected[n]), package_(p) + unsafe_(u), forced_(&forcedSelected[n]), package_(p) { if (n == "ascii") { // ASCII can encode 128 code points and nothing else @@ -288,10 +118,28 @@ Encoding::Encoding(string const & n, string const & l, string const & g, void Encoding::init() const { + // Since the the constructor is the only method which sets complete_ + // to false the test for complete_ is thread-safe without mutex. if (complete_) return; - start_encodable_ = 0; + static Mutex mutex; + Mutex::Locker lock(&mutex); + + // We need to test again for complete_, since another thread could + // have set it to true while we were waiting for the lock and we must + // not modify an encoding which is already complete. + if (complete_) + return; + + // We do not make any member mutable so that it can be easily verified + // that all const methods are thread-safe: init() is the only const + // method which changes complete_, encodable_ and start_encodable_, and + // it uses a mutex to ensure thread-safety. + CharSet & encodable = const_cast(this)->encodable_; + char_type & start_encodable = const_cast(this)->start_encodable_; + + start_encodable = 0; // temporarily switch off lyxerr, since we will generate iconv errors lyxerr.disable(); if (fixedwidth_) { @@ -305,10 +153,10 @@ void Encoding::init() const char_type const uc = ucs4[0]; CharInfoMap::const_iterator const it = unicodesymbols.find(uc); if (it == unicodesymbols.end()) - encodable_.insert(uc); + encodable.insert(uc); else if (!it->second.force()) { if (forced_->empty() || forced_->find(uc) == forced_->end()) - encodable_.insert(uc); + encodable.insert(uc); } } } else { @@ -321,22 +169,22 @@ void Encoding::init() const if (!eightbit.empty()) { CharInfoMap::const_iterator const it = unicodesymbols.find(c); if (it == unicodesymbols.end()) - encodable_.insert(c); + encodable.insert(c); else if (!it->second.force()) { if (forced_->empty() || forced_->find(c) == forced_->end()) - encodable_.insert(c); + encodable.insert(c); } } } } lyxerr.enable(); - CharSet::iterator it = encodable_.find(start_encodable_); - while (it != encodable_.end()) { - encodable_.erase(it); - ++start_encodable_; - it = encodable_.find(start_encodable_); + CharSet::iterator it = encodable.find(start_encodable); + while (it != encodable.end()) { + encodable.erase(it); + ++start_encodable; + it = encodable.find(start_encodable); } - complete_ = true; + const_cast(this)->complete_ = true; } @@ -352,9 +200,16 @@ bool Encoding::encodable(char_type c) const { // assure the used encoding is properly initialized init(); - if (iconvName_ == "UTF-8" && package_ == none) return true; + // platex does not load inputenc: force conversion of supported characters + if (package_ == Encoding::japanese + && ((0xb7 <= c && c <= 0x05ff) // Latin-1 Supplement ... Hebrew + || (0x1d00 <= c && c <= 0x218f) // Phonetic Extensions ... Number Forms + || (0x2193 <= c && c <= 0x2aff) // Arrows ... Supplemental Mathematical Operators + || (0xfb00 <= c && c <= 0xfb4f) // Alphabetic Presentation Forms + || (0x1d400 <= c && c <= 0x1d7ff))) // Mathematical Alphanumeric Symbols + return false; if (c < start_encodable_ && !isForced(c)) return true; if (encodable_.find(c) != encodable_.end()) @@ -372,22 +227,21 @@ pair Encoding::latexChar(char_type c) const CharInfoMap::const_iterator const it = unicodesymbols.find(c); if (it == unicodesymbols.end()) throw EncodingException(c); - // at least one of mathcommand and textcommand is nonempty - if (it->second.textcommand().empty()) + // at least one of mathCommand and textCommand is nonempty + if (it->second.textCommand().empty()) return make_pair( - "\\ensuremath{" + it->second.mathcommand() + '}', false); - return make_pair(it->second.textcommand(), !it->second.textnotermination()); + "\\ensuremath{" + it->second.mathCommand() + '}', false); + return make_pair(it->second.textCommand(), !it->second.textNoTermination()); } -pair Encoding::latexString(docstring const input, bool dryrun) const +pair Encoding::latexString(docstring const & input, bool dryrun) const { docstring result; docstring uncodable; bool terminate = false; - for (size_t n = 0; n < input.size(); ++n) { + for (char_type const c : input) { try { - char_type const c = input[n]; pair latex_char = latexChar(c); docstring const latex = latex_char.first; if (terminate && !prefixIs(latex, '\\') @@ -404,14 +258,15 @@ pair Encoding::latexString(docstring const input, bool dry result += latex; terminate = latex_char.second; } catch (EncodingException & /* e */) { - LYXERR0("Uncodable character in latexString!"); + LYXERR0("Uncodable character <" << docstring(1, c) + << "> in latexString!"); if (dryrun) { result += "<" + _("LyX Warning: ") + _("uncodable character") + " '"; - result += docstring(1, input[n]); + result += docstring(1, c); result += "'>"; } else - uncodable += input[n]; + uncodable += c; } } return make_pair(result, uncodable); @@ -423,16 +278,19 @@ vector Encoding::symbolsList() const // assure the used encoding is properly initialized init(); - // first all encodable characters - vector symbols(encodable_.begin(), encodable_.end()); - // add those below start_encodable_ + // first all those below start_encodable_ + vector symbols; for (char_type c = 0; c < start_encodable_; ++c) symbols.push_back(c); - // now the ones from the unicodesymbols file - CharInfoMap::const_iterator const end = unicodesymbols.end(); - CharInfoMap::const_iterator it = unicodesymbols.begin(); - for (; it != end; ++it) - symbols.push_back(it->first); + // add all encodable characters + copy(encodable_.begin(), encodable_.end(), back_inserter(symbols)); + // now the ones from the unicodesymbols file that are not already there + for (auto const & elem : unicodesymbols) { + if (find(symbols.begin(), symbols.end(), elem.first) == symbols.end()) + symbols.push_back(elem.first); + } + // finally, sort the vector + sort(symbols.begin(), symbols.end()); return symbols; } @@ -455,21 +313,22 @@ bool Encodings::latexMathChar(char_type c, bool mathmode, addMathSym(c); return false; } - // at least one of mathcommand and textcommand is nonempty - bool use_math = (mathmode && !it->second.mathcommand().empty()) || - (!mathmode && it->second.textcommand().empty()); + // at least one of mathCommand and textCommand is nonempty + bool use_math = (mathmode && !it->second.mathCommand().empty()) || + (!mathmode && it->second.textCommand().empty()); if (use_math) { - command = it->second.mathcommand(); - needsTermination = !it->second.mathnotermination(); + command = it->second.mathCommand(); + needsTermination = !it->second.mathNoTermination(); addMathCmd(c); } else { if (!encoding || command.empty()) { - command = it->second.textcommand(); - needsTermination = !it->second.textnotermination(); - addTextCmd(c); + command = it->second.textCommand(); + needsTermination = !it->second.textNoTermination(); } if (mathmode) addMathSym(c); + else + addTextCmd(c); } return use_math; } @@ -481,23 +340,33 @@ char_type Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, CharInfoMap::const_iterator const end = unicodesymbols.end(); CharInfoMap::const_iterator it = unicodesymbols.begin(); for (combining = false; it != end; ++it) { - docstring const math = it->second.mathcommand(); - docstring const text = it->second.textcommand(); - if ((cmdtype & MATH_CMD) && math == cmd) { - combining = it->second.combining(); - needsTermination = !it->second.mathnotermination(); - if (req && it->second.mathfeature() && - !it->second.mathpreamble().empty()) - req->insert(it->second.mathpreamble()); - return it->first; + if (it->second.deprecated()) + continue; + + if (cmdtype & MATH_CMD) { + for (const docstring& math : it->second.mathCommands()) { + if ((cmdtype & MATH_CMD) && math == cmd) { + combining = it->second.combining(); + needsTermination = !it->second.mathNoTermination(); + if (req && it->second.mathFeature() && + !it->second.mathPreamble().empty()) + req->insert(it->second.mathPreamble()); + return it->first; + } + } } - if ((cmdtype & TEXT_CMD) && text == cmd) { - combining = it->second.combining(); - needsTermination = !it->second.textnotermination(); - if (req && it->second.textfeature() && - !it->second.textpreamble().empty()) - req->insert(it->second.textpreamble()); - return it->first; + + if (cmdtype & TEXT_CMD) { + for (const docstring& text : it->second.textCommands()) { + if (text == cmd) { + combining = it->second.combining(); + needsTermination = !it->second.textNoTermination(); + if (req && it->second.textFeature() && + !it->second.textPreamble().empty()) + req->insert(it->second.textPreamble()); + return it->first; + } + } } } needsTermination = false; @@ -512,6 +381,38 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, rem = empty_docstring(); bool const mathmode = cmdtype & MATH_CMD; bool const textmode = cmdtype & TEXT_CMD; + + // Easy case: the command is a complete entry of unicodesymbols. + for (const auto & unicodeSymbol : unicodesymbols) { + if (mathmode) { + for (const auto & command : unicodeSymbol.second.mathCommands()) { + if (command == cmd) { + docstring value; + value += unicodeSymbol.first; + needsTermination = !unicodeSymbol.second.mathNoTermination(); + if (req && unicodeSymbol.second.mathFeature() + && !unicodeSymbol.second.mathPreamble().empty()) + req->insert(unicodeSymbol.second.mathPreamble()); + return value; + } + } + } + if (textmode) { + for (const auto & command : unicodeSymbol.second.textCommands()) { + if (command == cmd) { + docstring value; + value += unicodeSymbol.first; + needsTermination = !unicodeSymbol.second.textNoTermination(); + if (req && unicodeSymbol.second.textFeature() + && !unicodeSymbol.second.textPreamble().empty()) + req->insert(unicodeSymbol.second.textPreamble()); + return value; + } + } + } + } + + // Otherwise, try to map as many commands as possible, matching prefixes of the command. docstring symbols; size_t const cmdend = cmd.size(); size_t prefix = 0; @@ -566,9 +467,11 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, size_t unicmd_size = 0; char_type c = 0; for (; it != uniend; ++it) { - docstring const math = mathmode ? it->second.mathcommand() + if (it->second.deprecated()) + continue; + docstring const math = mathmode ? it->second.mathCommand() : docstring(); - docstring const text = textmode ? it->second.textcommand() + docstring const text = textmode ? it->second.textCommand() : docstring(); if (!combcmd.empty() && it->second.combining() && (math == combcmd || text == combcmd)) @@ -635,7 +538,7 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, || (tmp.size() == prefix + 1 && !isAlphaASCII(tmp[1]) && (prefix == 1 || !isAlphaASCII(tmp[2]))) - || k == cmdend + || k == cmdend || !isAlphaASCII(cmd[k]) || tmp[tmp.size() - 1] == '}' ) { @@ -644,16 +547,16 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, i = j + 1; unicmd_size = cur_size; if (math == tmp) - needsTermination = !it->second.mathnotermination(); + needsTermination = !it->second.mathNoTermination(); else - needsTermination = !it->second.textnotermination(); + needsTermination = !it->second.textNoTermination(); if (req) { - if (math == tmp && it->second.mathfeature() && - !it->second.mathpreamble().empty()) - req->insert(it->second.mathpreamble()); - if (text == tmp && it->second.textfeature() && - !it->second.textpreamble().empty()) - req->insert(it->second.textpreamble()); + if (math == tmp && it->second.mathFeature() && + !it->second.mathPreamble().empty()) + req->insert(it->second.mathPreamble()); + if (text == tmp && it->second.textFeature() && + !it->second.textPreamble().empty()) + req->insert(it->second.textPreamble()); } } } @@ -711,37 +614,6 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, } -bool Encodings::isHebrewComposeChar(char_type c) -{ - return c <= 0x05c2 && c >= 0x05b0 && c != 0x05be && c != 0x05c0; -} - - -// Special Arabic letters are ones that do not get connected from left -// they are hamza, alef_madda, alef_hamza, waw_hamza, alef_hamza_under, -// alef, tah_marbota, dal, thal, rah, zai, wow, alef_maksoura - -bool Encodings::isArabicSpecialChar(char_type c) -{ - return (c >= 0x0621 && c <= 0x0625) || (c >= 0x0630 && c <= 0x0632) - || c == 0x0627 || c == 0x0629 || c == 0x062f || c == 0x0648 - || c == 0x0649 || c == 0x0698; -} - - -bool Encodings::isArabicComposeChar(char_type c) -{ - return c >= 0x064b && c <= 0x0652; -} - - -bool Encodings::isArabicChar(char_type c) -{ - return c >= arabic_start && c <= arabic_end - && arabic_table[c-arabic_start][0]; -} - - CharInfo const & Encodings::unicodeCharInfo(char_type c) { static CharInfo empty; @@ -750,12 +622,6 @@ CharInfo const & Encodings::unicodeCharInfo(char_type c) } -char_type Encodings::transformChar(char_type c, Encodings::LetterForm form) -{ - return isArabicChar(c) ? arabic_table[c-arabic_start][form] : c; -} - - bool Encodings::isCombiningChar(char_type c) { CharInfoMap::const_iterator const it = unicodesymbols.find(c); @@ -769,26 +635,34 @@ string const Encodings::TIPAShortcut(char_type c) { CharInfoMap::const_iterator const it = unicodesymbols.find(c); if (it != unicodesymbols.end()) - return it->second.tipashortcut(); + return it->second.tipaShortcut(); return string(); } -bool Encodings::isKnownScriptChar(char_type const c, string & preamble) +string const Encodings::isKnownScriptChar(char_type const c) { CharInfoMap::const_iterator const it = unicodesymbols.find(c); if (it == unicodesymbols.end()) - return false; + return string(); + // FIXME: parse complex textPreamble (may be list or alternatives, + // e.g., "subscript,textgreek" or "textcomp|textgreek") + if (it->second.textPreamble() == "textgreek" + || it->second.textPreamble() == "textcyrillic") + return it->second.textPreamble(); + return string(); +} - if (it->second.textpreamble() != "textgreek" && it->second.textpreamble() != "textcyr") - return false; - if (preamble.empty()) { - preamble = it->second.textpreamble(); - return true; - } - return it->second.textpreamble() == preamble; +bool Encodings::fontencSupportsScript(string const & fontenc, string const & script) +{ + if (script == "textgreek") + return (fontenc == "LGR" || fontenc == "TU"); + if (script == "textcyrillic") + return (fontenc == "T2A" || fontenc == "T2B" || fontenc == "T2C" + || fontenc == "X2" || fontenc == "TU"); + return false; } @@ -798,18 +672,30 @@ bool Encodings::isMathAlpha(char_type c) } +bool Encodings::isUnicodeTextOnly(char_type c) +{ + if (isASCII(c) || isMathAlpha(c)) + return false; + + CharInfoMap::const_iterator const it = unicodesymbols.find(c); + return it == unicodesymbols.end() || it->second.mathCommand().empty(); +} + + Encoding const * Encodings::fromLyXName(string const & name, bool allowUnsafe) const { EncodingList::const_iterator const it = encodinglist.find(name); + if (it == encodinglist.end()) + return nullptr; if (!allowUnsafe && it->second.unsafe()) - return 0; - return it != encodinglist.end() ? &it->second : 0; + return nullptr; + return &it->second; } Encoding const * -Encodings::fromLaTeXName(string const & n, int const & p, bool allowUnsafe) const +Encodings::fromLaTeXName(string const & n, int p, bool allowUnsafe) const { string name = n; // FIXME: if we have to test for too many of these synonyms, @@ -827,19 +713,19 @@ Encodings::fromLaTeXName(string const & n, int const & p, bool allowUnsafe) cons if ((it->second.latexName() == name) && (it->second.package() & p) && (!it->second.unsafe() || allowUnsafe)) return &it->second; - return 0; + return nullptr; } Encoding const * -Encodings::fromIconvName(string const & n, int const & p, bool allowUnsafe) const +Encodings::fromIconvName(string const & n, int p, bool allowUnsafe) const { EncodingList::const_iterator const end = encodinglist.end(); for (EncodingList::const_iterator it = encodinglist.begin(); it != end; ++it) if ((it->second.iconvName() == n) && (it->second.package() & p) && (!it->second.unsafe() || allowUnsafe)) return &it->second; - return 0; + return nullptr; } @@ -851,41 +737,59 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) { // We must read the symbolsfile first, because the Encoding // constructor depends on it. - CharSetMap forcednotselected; - Lexer symbolslex; - symbolslex.setFile(symbolsfile); + CharSetMap forcedNotSelected; + Lexer symbolsLex; + symbolsLex.setFile(symbolsfile); bool getNextToken = true; - while (symbolslex.isOK()) { + while (symbolsLex.isOK()) { char_type symbol; if (getNextToken) { - if (!symbolslex.next(true)) + if (!symbolsLex.next(true)) break; } else getNextToken = true; - istringstream is(symbolslex.getString()); + istringstream is(symbolsLex.getString()); // reading symbol directly does not work if // char_type == wchar_t. - boost::uint32_t tmp; + uint32_t tmp; if(!(is >> hex >> tmp)) break; symbol = tmp; - if (!symbolslex.next(true)) + // Special case: more than one entry for one character (to add other LaTeX commands). + if (unicodesymbols.find(symbol) != unicodesymbols.end()) { + if (!symbolsLex.next(true)) + break; + docstring textCommand = symbolsLex.getDocString(); + if (!symbolsLex.next(true)) + break; + string mathCommand = symbolsLex.getString(); + + if (!textCommand.empty()) + unicodesymbols.at(symbol).addTextCommand(textCommand); + if (!mathCommand.empty()) + unicodesymbols.at(symbol).addMathCommand(textCommand); + + continue; + } + + // If the symbol is not the same as the previous entry, consider it is a totally new symbol. + if (!symbolsLex.next(true)) break; - docstring textcommand = symbolslex.getDocString(); - if (!symbolslex.next(true)) + docstring textCommand = symbolsLex.getDocString(); + if (!symbolsLex.next(true)) break; - string textpreamble = symbolslex.getString(); - if (!symbolslex.next(true)) + string textPreamble = symbolsLex.getString(); + if (!symbolsLex.next(true)) break; - string sflags = symbolslex.getString(); - - string tipashortcut; + string sflags = symbolsLex.getString(); + + string tipaShortcut; int flags = 0; - if (suffixIs(textcommand, '}')) + if (suffixIs(textCommand, '}')) flags |= CharInfoTextNoTermination; while (!sflags.empty()) { string flag; @@ -896,16 +800,16 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) flags |= CharInfoForce; forced.insert(symbol); } else if (prefixIs(flag, "force=")) { - vector encodings = + vector encs = getVectorFromString(flag.substr(6), ";"); - for (size_t i = 0; i < encodings.size(); ++i) - forcedselected[encodings[i]].insert(symbol); + for (auto const & enc : encs) + forcedSelected[enc].insert(symbol); flags |= CharInfoForceSelected; } else if (prefixIs(flag, "force!=")) { - vector encodings = + vector encs = getVectorFromString(flag.substr(7), ";"); - for (size_t i = 0; i < encodings.size(); ++i) - forcednotselected[encodings[i]].insert(symbol); + for (auto const & enc : encs) + forcedNotSelected[enc].insert(symbol); flags |= CharInfoForceSelected; } else if (flag == "mathalpha") { mathalpha.insert(symbol); @@ -920,7 +824,9 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) flags &= ~CharInfoTextNoTermination; flags &= ~CharInfoMathNoTermination; } else if (contains(flag, "tipashortcut=")) { - tipashortcut = split(flag, '='); + tipaShortcut = split(flag, '='); + } else if (flag == "deprecated") { + flags |= CharInfoDeprecated; } else { lyxerr << "Ignoring unknown flag `" << flag << "' for symbol `0x" @@ -928,27 +834,27 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) << "'." << endl; } } - // mathcommand and mathpreamble have been added for 1.6.0. + // mathCommand and mathPreamble have been added for 1.6.0. // make them optional so that old files still work. - int const lineno = symbolslex.lineNumber(); + int const lineNo = symbolsLex.lineNumber(); bool breakout = false; - docstring mathcommand; - string mathpreamble; - if (symbolslex.next(true)) { - if (symbolslex.lineNumber() != lineno) { - // line in old format without mathcommand and mathpreamble + docstring mathCommand; + string mathPreamble; + if (symbolsLex.next(true)) { + if (symbolsLex.lineNumber() != lineNo) { + // line in old format without mathCommand and mathPreamble getNextToken = false; } else { - mathcommand = symbolslex.getDocString(); - if (suffixIs(mathcommand, '}')) + mathCommand = symbolsLex.getDocString(); + if (suffixIs(mathCommand, '}')) flags |= CharInfoMathNoTermination; - if (symbolslex.next(true)) { - if (symbolslex.lineNumber() != lineno) { - // line in new format with mathcommand only + if (symbolsLex.next(true)) { + if (symbolsLex.lineNumber() != lineNo) { + // line in new format with mathCommand only getNextToken = false; } else { - // line in new format with mathcommand and mathpreamble - mathpreamble = symbolslex.getString(); + // line in new format with mathCommand and mathPreamble + mathPreamble = symbolsLex.getString(); } } else breakout = true; @@ -958,27 +864,27 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) } // backward compatibility - if (mathpreamble == "esintoramsmath") - mathpreamble = "esint|amsmath"; + if (mathPreamble == "esintoramsmath") + mathPreamble = "esint|amsmath"; - if (!textpreamble.empty()) - if (textpreamble[0] != '\\') + if (!textPreamble.empty()) + if (textPreamble[0] != '\\') flags |= CharInfoTextFeature; - if (!mathpreamble.empty()) - if (mathpreamble[0] != '\\') + if (!mathPreamble.empty()) + if (mathPreamble[0] != '\\') flags |= CharInfoMathFeature; CharInfo info = CharInfo( - textcommand, mathcommand, - textpreamble, mathpreamble, - tipashortcut, flags); + textCommand, mathCommand, + textPreamble, mathPreamble, + tipaShortcut, flags); LYXERR(Debug::INFO, "Read unicode symbol " << symbol << " '" - << to_utf8(info.textcommand()) << "' '" << info.textpreamble() - << " '" << info.textfeature() << ' ' << info.textnotermination() - << ' ' << to_utf8(info.mathcommand()) << "' '" << info.mathpreamble() - << "' " << info.mathfeature() << ' ' << info.mathnotermination() + << to_utf8(info.textCommand()) << "' '" << info.textPreamble() + << " '" << info.textFeature() << ' ' << info.textNoTermination() + << ' ' << to_utf8(info.mathCommand()) << "' '" << info.mathPreamble() + << "' " << info.mathFeature() << ' ' << info.mathNoTermination() << ' ' << info.combining() << ' ' << info.force() - << ' ' << info.forceselected()); + << ' ' << info.forceSelected()); // we assume that at least one command is nonempty when using unicodesymbols if (info.isUnicodeSymbol()) { @@ -995,12 +901,12 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) et_end }; - LexerKeyword encodingtags[] = { + LexerKeyword encodingTags[] = { { "encoding", et_encoding }, { "end", et_end } }; - Lexer lex(encodingtags); + Lexer lex(encodingTags); lex.setFile(encfile); lex.setContext("Encodings::read"); while (lex.isOK()) { @@ -1010,21 +916,21 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) lex.next(); string const name = lex.getString(); lex.next(); - string const latexname = lex.getString(); + string const latexName = lex.getString(); lex.next(); - string const guiname = lex.getString(); + string const guiName = lex.getString(); lex.next(); - string const iconvname = lex.getString(); + string const iconvName = lex.getString(); lex.next(); string const width = lex.getString(); - bool fixedwidth = false; + bool fixedWidth = false; bool unsafe = false; if (width == "fixed") - fixedwidth = true; + fixedWidth = true; else if (width == "variable") - fixedwidth = false; + fixedWidth = false; else if (width == "variableunsafe") { - fixedwidth = false; + fixedWidth = false; unsafe = true; } else @@ -1045,9 +951,9 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) lex.printError("Unknown package"); LYXERR(Debug::INFO, "Reading encoding " << name); - encodinglist[name] = Encoding(name, latexname, - guiname, iconvname, fixedwidth, unsafe, - package); + encodinglist[name] = Encoding(name, latexName, + guiName, iconvName, fixedWidth, unsafe, + package); if (lex.lex() != et_end) lex.printError("Missing end"); @@ -1064,9 +970,9 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) } } - // Move all information from forcednotselected to forcedselected - for (CharSetMap::const_iterator it1 = forcednotselected.begin(); it1 != forcednotselected.end(); ++it1) { - for (CharSetMap::iterator it2 = forcedselected.begin(); it2 != forcedselected.end(); ++it2) { + // Move all information from forcedNotSelected to forcedSelected + for (CharSetMap::const_iterator it1 = forcedNotSelected.begin(); it1 != forcedNotSelected.end(); ++it1) { + for (CharSetMap::iterator it2 = forcedSelected.begin(); it2 != forcedSelected.end(); ++it2) { if (it2->first != it1->first) it2->second.insert(it1->second.begin(), it1->second.end()); }