X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FEncoding.cpp;h=73edc843ff361bfe977441284f67f10d498239e6;hb=57b69a5efddf9f3c148007322f00dad6c253a2ed;hp=d925292742fc30868602b0df0bc6b624dc58917d;hpb=b293bd3670e551c10590af266f4d605bc5ff4152;p=lyx.git diff --git a/src/Encoding.cpp b/src/Encoding.cpp index d925292742..73edc843ff 100644 --- a/src/Encoding.cpp +++ b/src/Encoding.cpp @@ -14,23 +14,19 @@ #include "Encoding.h" -#include "Buffer.h" -#include "BufferList.h" -#include "InsetIterator.h" -#include "LaTeXFeatures.h" #include "Lexer.h" -#include "LyXRC.h" #include "support/debug.h" #include "support/gettext.h" -#include "support/FileName.h" #include "support/lstrings.h" +#include "support/mutex.h" #include "support/textutils.h" #include "support/unicode.h" #include #include +#include using namespace std; using namespace lyx::support; @@ -47,241 +43,6 @@ 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; - - -enum CharInfoFlags { - /// - CharInfoCombining = 1, - /// - CharInfoTextFeature = 2, - /// - CharInfoMathFeature = 4, - /// - CharInfoForce = 8, - /// - CharInfoTextNoTermination = 16, - /// - CharInfoMathNoTermination = 32, - /// - CharInfoForceSelected = 64, -}; - -/// Information about a single UCS4 character -struct CharInfo { - /// LaTeX command (text mode) for this character - docstring textcommand; - /// LaTeX command (math mode) for this character - docstring mathcommand; - /// Needed LaTeX preamble (or feature) for text mode - string textpreamble; - /// Needed LaTeX preamble (or feature) for math mode - string mathpreamble; - /// Is this a combining character? - bool combining() const { return flags & CharInfoCombining ? true : false; } - /// Is \c textpreamble a feature known by LaTeXFeatures, or a raw LaTeX - /// command? - bool textfeature() const { return flags & CharInfoTextFeature ? true : false; } - /// Is \c mathpreamble a feature known by LaTeXFeatures, or a raw LaTeX - /// command? - bool mathfeature() const { return flags & CharInfoMathFeature ? true : false; } - /// Always force the LaTeX command, even if the encoding contains - /// this character? - bool force() const { return flags & CharInfoForce ? true : false; } - /// Force the LaTeX command for some encodings? - bool forceselected() const { return flags & CharInfoForceSelected ? true : false; } - /// TIPA shortcut - string tipashortcut; - /// \c textcommand needs no termination (such as {} or space). - bool textnotermination() const { return flags & CharInfoTextNoTermination ? true : false; } - /// \c mathcommand needs no termination (such as {} or space). - bool mathnotermination() const { return flags & CharInfoMathNoTermination ? true : false; } - /// - unsigned int flags; -}; - - typedef map CharInfoMap; CharInfoMap unicodesymbols; @@ -312,6 +73,17 @@ const char * EncodingException::what() const throw() } +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) +{ +} + + 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), @@ -334,10 +106,28 @@ 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_) { @@ -351,10 +141,10 @@ 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 { @@ -367,22 +157,22 @@ 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; } @@ -419,14 +209,14 @@ pair Encoding::latexChar(char_type c) const if (it == unicodesymbols.end()) throw EncodingException(c); // at least one of mathcommand and textcommand is nonempty - if (it->second.textcommand.empty()) + 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; @@ -502,15 +292,15 @@ bool Encodings::latexMathChar(char_type c, bool mathmode, return false; } // at least one of mathcommand and textcommand is nonempty - bool use_math = (mathmode && !it->second.mathcommand.empty()) || - (!mathmode && it->second.textcommand.empty()); + bool use_math = (mathmode && !it->second.mathcommand().empty()) || + (!mathmode && it->second.textcommand().empty()); if (use_math) { - command = it->second.mathcommand; + command = it->second.mathcommand(); needsTermination = !it->second.mathnotermination(); addMathCmd(c); } else { if (!encoding || command.empty()) { - command = it->second.textcommand; + command = it->second.textcommand(); needsTermination = !it->second.textnotermination(); addTextCmd(c); } @@ -527,22 +317,24 @@ 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 (it->second.deprecated()) + continue; + 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); + !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); + !it->second.textpreamble().empty()) + req->insert(it->second.textpreamble()); return it->first; } } @@ -584,8 +376,11 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, if (j + 1 < cmdend && cmd[j + 1] == '{') { size_t k = j + 1; int count = 1; - while (k < cmdend && count && k != docstring::npos) { + while (k < cmdend && count) { k = cmd.find_first_of(from_ascii("{}"), k + 1); + // braces may not be balanced + if (k == docstring::npos) + break; if (cmd[k] == '{') ++count; else @@ -609,9 +404,11 @@ 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)) @@ -650,8 +447,11 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, k < cmdend && cmd[k] == '{') { size_t l = k; int count = 1; - while (l < cmdend && count && l != docstring::npos) { + while (l < cmdend && count) { l = cmd.find_first_of(from_ascii("{}"), l + 1); + // braces may not be balanced + if (l == docstring::npos) + break; if (cmd[l] == '{') ++count; else @@ -675,7 +475,7 @@ 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] == '}' ) { @@ -689,11 +489,11 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, needsTermination = !it->second.textnotermination(); if (req) { if (math == tmp && it->second.mathfeature() && - !it->second.mathpreamble.empty()) - req->insert(it->second.mathpreamble); + !it->second.mathpreamble().empty()) + req->insert(it->second.mathpreamble()); if (text == tmp && it->second.textfeature() && - !it->second.textpreamble.empty()) - req->insert(it->second.textpreamble); + !it->second.textpreamble().empty()) + req->insert(it->second.textpreamble()); } } } @@ -751,131 +551,11 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, } -void Encodings::initUnicodeMath(Buffer const & buffer, bool for_master) -{ -#ifdef TEX2LYX - // The code below is not needed in tex2lyx and requires additional stuff - (void)buffer; - (void)for_master; -#else - if (for_master) { - mathcmd.clear(); - textcmd.clear(); - mathsym.clear(); - } - - // Check this buffer - Inset & inset = buffer.inset(); - InsetIterator it = inset_iterator_begin(inset); - InsetIterator const end = inset_iterator_end(inset); - for (; it != end; ++it) - it->initUnicodeMath(); - - if (!for_master) - return; - - // Check children - ListOfBuffers blist = buffer.getDescendents(); - ListOfBuffers::const_iterator bit = blist.begin(); - ListOfBuffers::const_iterator const bend = blist.end(); - for (; bit != bend; ++bit) - initUnicodeMath(**bit, false); -#endif -} - - -void Encodings::validate(char_type c, LaTeXFeatures & features, bool for_mathed) +CharInfo const & Encodings::unicodeCharInfo(char_type c) { -#ifdef TEX2LYX - // The code below is not needed in tex2lyx and requires additional stuff - (void)c; - (void)features; - (void)for_mathed; -#else + static CharInfo empty; CharInfoMap::const_iterator const it = unicodesymbols.find(c); - if (it != unicodesymbols.end()) { - // In mathed, c could be used both in textmode and mathmode - bool const math_mode = for_mathed && isMathCmd(c); - bool const use_math = math_mode || - (!for_mathed && it->second.textcommand.empty()); - bool const use_text = (for_mathed && isTextCmd(c)) || - (!for_mathed && !it->second.textcommand.empty()); - bool const plain_utf8 = (features.runparams().encoding->name() == "utf8-plain"); - bool const unicode_math = (features.isRequired("unicode-math") - && features.isAvailable("unicode-math")); - // with utf8-plain, we only load packages when in mathed (see #7766) - // and if we do not use unicode-math - if ((math_mode && !unicode_math) - || (use_math && !plain_utf8)) { - if (!it->second.mathpreamble.empty()) { - if (it->second.mathfeature()) { - string feats = it->second.mathpreamble; - while (!feats.empty()) { - string feat; - feats = split(feats, feat, ','); - features.require(feat); - } - } else - features.addPreambleSnippet(it->second.mathpreamble); - } - } - // with utf8-plain, we do not load packages (see #7766) - if (use_text && !plain_utf8) { - if (!it->second.textpreamble.empty()) { - if (it->second.textfeature()) { - string feats = it->second.textpreamble; - while (!feats.empty()) { - string feat; - feats = split(feats, feat, ','); - features.require(feat); - } - } else - features.addPreambleSnippet(it->second.textpreamble); - } - } - } - if (for_mathed && isMathSym(c)) { - features.require("amstext"); - features.require("lyxmathsym"); - } -#endif -} - - -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]; -} - - -char_type Encodings::transformChar(char_type c, Encodings::LetterForm form) -{ - return isArabicChar(c) ? arabic_table[c-arabic_start][form] : c; + return it != unicodesymbols.end() ? it->second : empty; } @@ -892,7 +572,7 @@ 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(); } @@ -904,14 +584,14 @@ bool Encodings::isKnownScriptChar(char_type const c, string & preamble) if (it == unicodesymbols.end()) return false; - if (it->second.textpreamble != "textgreek" && it->second.textpreamble != "textcyr") + if (it->second.textpreamble() != "textgreek" && it->second.textpreamble() != "textcyr") return false; if (preamble.empty()) { - preamble = it->second.textpreamble; + preamble = it->second.textpreamble(); return true; } - return it->second.textpreamble == preamble; + return it->second.textpreamble() == preamble; } @@ -925,9 +605,11 @@ Encoding const * Encodings::fromLyXName(string const & name, bool allowUnsafe) const { EncodingList::const_iterator const it = encodinglist.find(name); + if (it == encodinglist.end()) + return 0; if (!allowUnsafe && it->second.unsafe()) return 0; - return it != encodinglist.end() ? &it->second : 0; + return &it->second; } @@ -980,8 +662,6 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) bool getNextToken = true; while (symbolslex.isOK()) { char_type symbol; - CharInfo info; - string flags; if (getNextToken) { if (!symbolslex.next(true)) @@ -999,53 +679,55 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) if (!symbolslex.next(true)) break; - info.textcommand = symbolslex.getDocString(); + docstring textcommand = symbolslex.getDocString(); if (!symbolslex.next(true)) break; - info.textpreamble = symbolslex.getString(); + string textpreamble = symbolslex.getString(); if (!symbolslex.next(true)) break; - flags = symbolslex.getString(); - - info.flags = 0; - if (suffixIs(info.textcommand, '}')) - info.flags |= CharInfoTextNoTermination; - if (suffixIs(info.mathcommand, '}')) - info.flags |= CharInfoMathNoTermination; - while (!flags.empty()) { + string sflags = symbolslex.getString(); + + string tipashortcut; + int flags = 0; + + if (suffixIs(textcommand, '}')) + flags |= CharInfoTextNoTermination; + while (!sflags.empty()) { string flag; - flags = split(flags, flag, ','); + sflags = split(sflags, flag, ','); if (flag == "combining") { - info.flags |= CharInfoCombining; + flags |= CharInfoCombining; } else if (flag == "force") { - info.flags |= CharInfoForce; + flags |= CharInfoForce; forced.insert(symbol); } else if (prefixIs(flag, "force=")) { vector encodings = getVectorFromString(flag.substr(6), ";"); for (size_t i = 0; i < encodings.size(); ++i) forcedselected[encodings[i]].insert(symbol); - info.flags |= CharInfoForceSelected; + flags |= CharInfoForceSelected; } else if (prefixIs(flag, "force!=")) { vector encodings = getVectorFromString(flag.substr(7), ";"); for (size_t i = 0; i < encodings.size(); ++i) forcednotselected[encodings[i]].insert(symbol); - info.flags |= CharInfoForceSelected; + flags |= CharInfoForceSelected; } else if (flag == "mathalpha") { mathalpha.insert(symbol); } else if (flag == "notermination=text") { - info.flags |= CharInfoTextNoTermination; + flags |= CharInfoTextNoTermination; } else if (flag == "notermination=math") { - info.flags |= CharInfoMathNoTermination; + flags |= CharInfoMathNoTermination; } else if (flag == "notermination=both") { - info.flags |= CharInfoTextNoTermination; - info.flags |= CharInfoMathNoTermination; + flags |= CharInfoTextNoTermination; + flags |= CharInfoMathNoTermination; } else if (flag == "notermination=none") { - info.flags &= ~CharInfoTextNoTermination; - info.flags &= ~CharInfoMathNoTermination; + flags &= ~CharInfoTextNoTermination; + flags &= ~CharInfoMathNoTermination; } else if (contains(flag, "tipashortcut=")) { - info.tipashortcut = split(flag, '='); + tipashortcut = split(flag, '='); + } else if (flag == "deprecated") { + flags |= CharInfoDeprecated; } else { lyxerr << "Ignoring unknown flag `" << flag << "' for symbol `0x" @@ -1057,19 +739,23 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) // make them optional so that old files still work. 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 getNextToken = false; } else { - info.mathcommand = symbolslex.getDocString(); + mathcommand = symbolslex.getDocString(); + if (suffixIs(mathcommand, '}')) + flags |= CharInfoMathNoTermination; 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 - info.mathpreamble = symbolslex.getString(); + mathpreamble = symbolslex.getString(); } } else breakout = true; @@ -1079,27 +765,32 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) } // backward compatibility - if (info.mathpreamble == "esintoramsmath") - info.mathpreamble = "esint|amsmath"; - - if (!info.textpreamble.empty()) - if (info.textpreamble[0] != '\\') - info.flags |= CharInfoTextFeature; - if (!info.mathpreamble.empty()) - if (info.mathpreamble[0] != '\\') - info.flags |= CharInfoMathFeature; - + if (mathpreamble == "esintoramsmath") + mathpreamble = "esint|amsmath"; + + if (!textpreamble.empty()) + if (textpreamble[0] != '\\') + flags |= CharInfoTextFeature; + if (!mathpreamble.empty()) + if (mathpreamble[0] != '\\') + flags |= CharInfoMathFeature; + + CharInfo info = CharInfo( + 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() - << ' ' << info.combining() << ' ' << info.force() - << ' ' << info.forceselected()); + << 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()); // we assume that at least one command is nonempty when using unicodesymbols - if (!info.textcommand.empty() || !info.mathcommand.empty()) + if (info.isUnicodeSymbol()) { unicodesymbols[symbol] = info; + } if (breakout) break;