X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;ds=sidebyside;f=src%2FEncoding.cpp;h=96e35ecbb707e04e75fee87cb30211fbbdc8173e;hb=e65e84878f93008006470db3372b8ee7bf994872;hp=a6f6a60ea3162144c5711fed6c7853658502c9d6;hpb=3013ce3716603e751152e162a5b155cd455c745e;p=lyx.git diff --git a/src/Encoding.cpp b/src/Encoding.cpp index a6f6a60ea3..96e35ecbb7 100644 --- a/src/Encoding.cpp +++ b/src/Encoding.cpp @@ -19,13 +19,15 @@ #include "support/debug.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 using namespace std; using namespace lyx::support; @@ -42,190 +44,6 @@ 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; - - typedef map CharInfoMap; CharInfoMap unicodesymbols; @@ -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) @@ -257,15 +75,16 @@ 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) + 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), @@ -288,10 +107,28 @@ 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 +142,10 @@ 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 +158,22 @@ 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; } @@ -380,7 +217,7 @@ pair Encoding::latexChar(char_type c) const } -pair Encoding::latexString(docstring const input, bool dryrun) const +pair Encoding::latexString(docstring const & input, bool dryrun) const { docstring result; docstring uncodable; @@ -423,16 +260,19 @@ 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 (pair 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; } @@ -481,6 +321,8 @@ 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) { + if (it->second.deprecated()) + continue; docstring const math = it->second.mathcommand(); docstring const text = it->second.textcommand(); if ((cmdtype & MATH_CMD) && math == cmd) { @@ -566,6 +408,8 @@ docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype, size_t unicmd_size = 0; char_type c = 0; for (; it != uniend; ++it) { + if (it->second.deprecated()) + continue; docstring const math = mathmode ? it->second.mathcommand() : docstring(); docstring const text = textmode ? it->second.textcommand() @@ -635,7 +479,7 @@ 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] == '}' ) { @@ -711,37 +555,6 @@ 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 +563,6 @@ 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); @@ -798,13 +605,25 @@ 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 0; if (!allowUnsafe && it->second.unsafe()) return 0; - return it != encodinglist.end() ? &it->second : 0; + return &it->second; } @@ -881,7 +700,7 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) if (!symbolslex.next(true)) break; string sflags = symbolslex.getString(); - + string tipashortcut; int flags = 0; @@ -921,6 +740,8 @@ void Encodings::read(FileName const & encfile, FileName const & symbolsfile) flags &= ~CharInfoMathNoTermination; } else if (contains(flag, "tipashortcut=")) { tipashortcut = split(flag, '='); + } else if (flag == "deprecated") { + flags |= CharInfoDeprecated; } else { lyxerr << "Ignoring unknown flag `" << flag << "' for symbol `0x"