#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/FileName.h"
+#include "support/gettext.h"
#include "support/lstrings.h"
#include "support/textutils.h"
#include "support/unicode.h"
namespace lyx {
+int const Encoding::any = -1;
+
Encodings encodings;
Encodings::MathCommandSet Encodings::mathcmd;
{0, 0, 0, 0}, // 0x067b
{0, 0, 0, 0}, // 0x067c
{0, 0, 0, 0}, // 0x067d
- {0xfb56, 0xfb57, 0xfb58, 0xfb59}, // 0x067e = peh
+ {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}, // 0x0683
{0, 0, 0, 0}, // 0x0684
{0, 0, 0, 0}, // 0x0685
- {0xfb7a, 0xfb7b, 0xfb7c, 0xfb7d}, // 0x0686 = tcheh
+ {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}, // 0x06a6
{0, 0, 0, 0}, // 0x06a7
{0, 0, 0, 0}, // 0x06a8
- {0xfb8e, 0xfb8f, 0xfb90, 0xfb91}, // 0x06a9 = farsi kaf
+ {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
+ {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}, // 0x06c9
{0, 0, 0, 0}, // 0x06ca
{0, 0, 0, 0}, // 0x06cb
- {0xfbfc, 0xfbfd, 0xfbfe, 0xfbff} // 0x06cc = farsi yeh
+ {0xfbfc, 0xfbfd, 0xfbfe, 0xfbff} // 0x06cc = farsi yeh
};
char_type const arabic_end = 0x06cc;
-/// 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;
- /// Is \c textpreamble a feature known by LaTeXFeatures, or a raw LaTeX
- /// command?
- bool textfeature;
- /// Is \c mathpreamble a feature known by LaTeXFeatures, or a raw LaTeX
- /// command?
- bool mathfeature;
- /// Always force the LaTeX command, even if the encoding contains
- /// this character?
- bool force;
-};
-
-
typedef map<char_type, CharInfo> CharInfoMap;
CharInfoMap unicodesymbols;
-typedef std::set<char_type> CharSet;
+typedef set<char_type> CharSet;
+typedef map<string, CharSet> CharSetMap;
CharSet forced;
+CharSetMap forcedselected;
-typedef std::set<char_type> MathAlphaSet;
+typedef set<char_type> MathAlphaSet;
MathAlphaSet mathalpha;
}
+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, Encoding::Package p)
- : name_(n), latexName_(l), guiName_(g), iconvName_(i), fixedwidth_(f), package_(p)
+ 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)
{
if (n == "ascii") {
// ASCII can encode 128 code points and nothing else
start_encodable_ = max_ucs4;
complete_ = true;
} else {
+ start_encodable_ = 0;
complete_ = false;
}
}
continue;
char_type const uc = ucs4[0];
CharInfoMap::const_iterator const it = unicodesymbols.find(uc);
- if (it == unicodesymbols.end() || !it->second.force)
+ if (it == unicodesymbols.end())
encodable_.insert(uc);
+ else if (!it->second.force()) {
+ if (forced_->empty() || forced_->find(uc) == forced_->end())
+ encodable_.insert(uc);
+ }
}
} else {
// We do not know how many code points this encoding has, and
vector<char> const eightbit = ucs4_to_eightbit(&c, 1, iconvName_);
if (!eightbit.empty()) {
CharInfoMap::const_iterator const it = unicodesymbols.find(c);
- if (it == unicodesymbols.end() || !it->second.force)
+ if (it == unicodesymbols.end())
encodable_.insert(c);
+ else if (!it->second.force()) {
+ if (forced_->empty() || forced_->find(c) == forced_->end())
+ encodable_.insert(c);
+ }
}
}
}
}
-docstring Encoding::latexChar(char_type c, bool no_commands) const
+bool Encoding::isForced(char_type c) const
+{
+ if (!forced.empty() && forced.find(c) != forced.end())
+ return true;
+ return !forced_->empty() && forced_->find(c) != forced_->end();
+}
+
+
+bool Encoding::encodable(char_type c) const
{
// assure the used encoding is properly initialized
init();
if (iconvName_ == "UTF-8" && package_ == none)
- return docstring(1, c);
- if (c < start_encodable_ && !encodings.isForced(c))
- return docstring(1, c);
+ return true;
+ if (c < start_encodable_ && !isForced(c))
+ return true;
if (encodable_.find(c) != encodable_.end())
- return docstring(1, c);
- if (no_commands)
- return docstring();
+ return true;
+ return false;
+}
+
+
+pair<docstring, bool> Encoding::latexChar(char_type c) const
+{
+ if (encodable(c))
+ return make_pair(docstring(1, c), false);
// c cannot (or should not) be encoded in this encoding
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())
- return "\\ensuremath{" + it->second.mathcommand + '}';
- return it->second.textcommand;
+ if (it->second.textcommand().empty())
+ return make_pair(
+ "\\ensuremath{" + it->second.mathcommand() + '}', false);
+ return make_pair(it->second.textcommand(), !it->second.textnotermination());
+}
+
+
+pair<docstring, docstring> Encoding::latexString(docstring const input, bool dryrun) const
+{
+ docstring result;
+ docstring uncodable;
+ bool terminate = false;
+ for (size_t n = 0; n < input.size(); ++n) {
+ try {
+ char_type const c = input[n];
+ pair<docstring, bool> latex_char = latexChar(c);
+ docstring const latex = latex_char.first;
+ if (terminate && !prefixIs(latex, '\\')
+ && !prefixIs(latex, '{')
+ && !prefixIs(latex, '}')) {
+ // Prevent eating of a following
+ // space or command corruption by
+ // following characters
+ if (latex == " ")
+ result += "{}";
+ else
+ result += " ";
+ }
+ result += latex;
+ terminate = latex_char.second;
+ } catch (EncodingException & /* e */) {
+ LYXERR0("Uncodable character in latexString!");
+ if (dryrun) {
+ result += "<" + _("LyX Warning: ")
+ + _("uncodable character") + " '";
+ result += docstring(1, input[n]);
+ result += "'>";
+ } else
+ uncodable += input[n];
+ }
+ }
+ return make_pair(result, uncodable);
}
bool Encodings::latexMathChar(char_type c, bool mathmode,
- Encoding const * encoding, docstring & command)
+ Encoding const * encoding, docstring & command,
+ bool & needsTermination)
{
+ command = empty_docstring();
if (encoding)
- command = encoding->latexChar(c, true);
+ if (encoding->encodable(c))
+ command = docstring(1, c);
+ needsTermination = false;
CharInfoMap::const_iterator const it = unicodesymbols.find(c);
if (it == unicodesymbols.end()) {
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);
}
if (mathmode)
}
-char_type Encodings::fromLaTeXCommand(docstring const & cmd, bool & combining)
+char_type Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype,
+ bool & combining, bool & needsTermination, set<string> * req)
{
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 (math == cmd || text == cmd) {
- combining = it->second.combining;
+ 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 ((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;
}
}
+ needsTermination = false;
return 0;
}
-docstring Encodings::fromLaTeXCommand(docstring const & cmd, docstring & rem,
- int cmdtype)
+docstring Encodings::fromLaTeXCommand(docstring const & cmd, int cmdtype,
+ bool & needsTermination, docstring & rem, set<string> * req)
{
+ needsTermination = false;
+ rem = empty_docstring();
bool const mathmode = cmdtype & MATH_CMD;
bool const textmode = cmdtype & TEXT_CMD;
docstring symbols;
- size_t i = 0;
size_t const cmdend = cmd.size();
+ size_t prefix = 0;
CharInfoMap::const_iterator const uniend = unicodesymbols.end();
- for (size_t j = 0; j < cmdend; ++j) {
+ for (size_t i = 0, j = 0; j < cmdend; ++j) {
// Also get the char after a backslash
- if (j + 1 < cmdend && cmd[j] == '\\')
+ if (j + 1 < cmdend && cmd[j] == '\\') {
++j;
+ prefix = 1;
+ // Detect things like \=*{e} as well
+ if (j + 3 < cmdend && cmd[j+1] == '*' &&
+ cmd[j+2] == '{') {
+ ++j;
+ prefix = 2;
+ }
+ }
+ // position of the last character before a possible macro
+ // argument
+ size_t m = j;
// If a macro argument follows, get it, too
+ // Do it here only for single character commands. Other
+ // combining commands need this too, but they are handled in
+ // the loop below for performance reasons.
if (j + 1 < cmdend && cmd[j + 1] == '{') {
size_t k = j + 1;
int count = 1;
}
if (k != docstring::npos)
j = k;
+ } else if (m + 1 < cmdend && isAlphaASCII(cmd[m])) {
+ while (m + 2 < cmdend && isAlphaASCII(cmd[m+1]))
+ m++;
}
// Start with this substring and try augmenting it when it is
// the prefix of some command in the unicodesymbols file
- docstring const subcmd = cmd.substr(i, j - i + 1);
+ docstring subcmd = cmd.substr(i, j - i + 1);
CharInfoMap::const_iterator it = unicodesymbols.begin();
+ // First part of subcmd which might be a combining character
+ docstring combcmd = (m == j) ? docstring() : cmd.substr(i, m - i + 1);
+ // The combining character of combcmd if it exists
+ CharInfoMap::const_iterator combining = uniend;
size_t unicmd_size = 0;
char_type c = 0;
for (; it != uniend; ++it) {
- docstring const math = mathmode ? it->second.mathcommand
+ 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))
+ combining = it;
size_t cur_size = max(math.size(), text.size());
// The current math or text unicode command cannot
// match, or we already matched a longer one
// If this is an exact match, we found a (longer)
// matching entry in the unicodesymbols file.
+ if (math != tmp && text != tmp)
+ continue;
+ // If we found a combining command, we need to append
+ // the macro argument if this has not been done above.
+ if (tmp == combcmd && combining != uniend &&
+ k < cmdend && cmd[k] == '{') {
+ size_t l = k;
+ int count = 1;
+ while (l < cmdend && count && l != docstring::npos) {
+ l = cmd.find_first_of(from_ascii("{}"), l + 1);
+ if (cmd[l] == '{')
+ ++count;
+ else
+ --count;
+ }
+ if (l != docstring::npos) {
+ j = l;
+ subcmd = cmd.substr(i, j - i + 1);
+ }
+ }
// If the entry doesn't start with '\', we take note
// of the match and continue (this is not a ultimate
// acceptance, as some other entry may match a longer
// (nonletter) char macro, or nothing else follows,
// or what follows is a nonletter char, or the last
// character is a }.
- if ((math == tmp || text == tmp)
- && (tmp[0] != '\\'
- || (tmp.size() == 2 && !isAlphaASCII(tmp[1]))
+ else if (tmp[0] != '\\'
+ || (tmp.size() == prefix + 1 &&
+ !isAlphaASCII(tmp[1]) &&
+ (prefix == 1 || !isAlphaASCII(tmp[2])))
|| k == cmdend
|| !isAlphaASCII(cmd[k])
- || tmp[tmp.size() - 1] == '}')
+ || tmp[tmp.size() - 1] == '}'
) {
c = it->first;
j = k - 1;
i = j + 1;
unicmd_size = cur_size;
+ if (math == tmp)
+ needsTermination = !it->second.mathnotermination();
+ else
+ 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 (unicmd_size)
symbols += c;
- else if (j + 1 == cmdend)
- // No luck. Return what remains
- rem = cmd.substr(i);
- }
- return symbols;
-}
-
-
-void Encodings::initUnicodeMath(Buffer const & buffer, bool clear_sets)
-{
-#ifdef TEX2LYX
- // The code below is not needed in tex2lyx and requires additional stuff
- (void)buffer;
- (void)clear_sets;
-#else
- if (clear_sets) {
- mathcmd.clear();
- textcmd.clear();
- mathsym.clear();
- }
-
- // Check master
- Inset & inset = buffer.inset();
- InsetIterator it = inset_iterator_begin(inset);
- InsetIterator const end = inset_iterator_end(inset);
- for (; it != end; ++it)
- it->initUnicodeMath();
-
- // Check children
- BufferList::iterator bit = theBufferList().begin();
- BufferList::iterator const bend = theBufferList().end();
- for (; bit != bend; ++bit)
- if (buffer.isChild(*bit))
- initUnicodeMath(**bit, false);
-#endif
-}
-
-
-void Encodings::validate(char_type c, LaTeXFeatures & features, bool for_mathed)
-{
-#ifdef TEX2LYX
- // The code below is not needed in tex2lyx and requires additional stuff
- (void)c;
- (void)features;
- (void)for_mathed;
-#else
- 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 use_math = (for_mathed && isMathCmd(c)) ||
- (!for_mathed && it->second.textcommand.empty());
- bool const use_text = (for_mathed && isTextCmd(c)) ||
- (!for_mathed && !it->second.textcommand.empty());
- if (use_math) {
- 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);
+ else if (combining != uniend &&
+ prefixIs(subcmd, combcmd + '{')) {
+ // We know that subcmd starts with combcmd and
+ // contains an argument in braces.
+ docstring const arg = subcmd.substr(
+ combcmd.length() + 1,
+ subcmd.length() - combcmd.length() - 2);
+ // If arg is a single character we can construct a
+ // combining sequence.
+ char_type a;
+ bool argcomb = false;
+ if (arg.size() == 1 && isAlnumASCII(arg[0]))
+ a = arg[0];
+ else {
+ // Use the version of fromLaTeXCommand() that
+ // parses only one command, since we cannot
+ // use more than one character.
+ bool dummy = false;
+ set<string> r;
+ a = fromLaTeXCommand(arg, cmdtype, argcomb,
+ dummy, &r);
+ if (a && req && !argcomb)
+ req->insert(r.begin(), r.end());
+ }
+ if (a && !argcomb) {
+ // In unicode the combining character comes
+ // after its base
+ symbols += a;
+ symbols += combining->first;
+ i = j + 1;
+ unicmd_size = 2;
}
}
- if (use_text) {
- 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 (j + 1 == cmdend && !unicmd_size) {
+ // No luck. Return what remains
+ rem = cmd.substr(i);
+ if (needsTermination && !rem.empty()) {
+ if (rem.substr(0, 2) == "{}") {
+ rem = rem.substr(2);
+ needsTermination = false;
+ } else if (rem[0] == ' ') {
+ needsTermination = false;
+ // LaTeX would swallow all spaces
+ rem = ltrim(rem);
+ }
}
}
}
- if (for_mathed && isMathSym(c)) {
- features.require("amstext");
- features.require("lyxmathsym");
- }
-#endif
+ return symbols;
}
}
+CharInfo const & Encodings::unicodeCharInfo(char_type c)
+{
+ static CharInfo empty;
+ CharInfoMap::const_iterator const it = unicodesymbols.find(c);
+ return it != unicodesymbols.end() ? it->second : empty;
+}
+
+
char_type Encodings::transformChar(char_type c, Encodings::LetterForm form)
{
return isArabicChar(c) ? arabic_table[c-arabic_start][form] : c;
{
CharInfoMap::const_iterator const it = unicodesymbols.find(c);
if (it != unicodesymbols.end())
- return it->second.combining;
+ return it->second.combining();
return false;
}
+string const Encodings::TIPAShortcut(char_type c)
+{
+ CharInfoMap::const_iterator const it = unicodesymbols.find(c);
+ if (it != unicodesymbols.end())
+ return it->second.tipashortcut();
+ return string();
+}
+
+
bool Encodings::isKnownScriptChar(char_type const c, string & preamble)
{
CharInfoMap::const_iterator const it = unicodesymbols.find(c);
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;
-}
-
-
-bool Encodings::isForced(char_type c)
-{
- return (!forced.empty() && forced.find(c) != forced.end());
+ return it->second.textpreamble() == preamble;
}
}
-Encoding const * Encodings::fromLyXName(string const & name) const
+Encoding const *
+Encodings::fromLyXName(string const & name, bool allowUnsafe) const
{
EncodingList::const_iterator const it = encodinglist.find(name);
+ if (!allowUnsafe && it->second.unsafe())
+ return 0;
return it != encodinglist.end() ? &it->second : 0;
}
-Encoding const * Encodings::fromLaTeXName(string const & n) const
+Encoding const *
+Encodings::fromLaTeXName(string const & n, int const & p, bool allowUnsafe) const
{
string name = n;
// FIXME: if we have to test for too many of these synonyms,
// most at the top of lib/encodings.
EncodingList::const_iterator const end = encodinglist.end();
for (EncodingList::const_iterator it = encodinglist.begin(); it != end; ++it)
- if (it->second.latexName() == name)
+ if ((it->second.latexName() == name) && (it->second.package() & p)
+ && (!it->second.unsafe() || allowUnsafe))
return &it->second;
return 0;
}
-Encodings::Encodings()
+Encoding const *
+Encodings::fromIconvName(string const & n, int const & 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;
}
+Encodings::Encodings()
+{}
+
+
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);
bool getNextToken = true;
while (symbolslex.isOK()) {
char_type symbol;
- CharInfo info;
- string flags;
if (getNextToken) {
if (!symbolslex.next(true))
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.combining = false;
- info.textfeature = false;
- info.force = false;
- 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.combining = true;
+ flags |= CharInfoCombining;
} else if (flag == "force") {
- info.force = true;
+ flags |= CharInfoForce;
forced.insert(symbol);
+ } else if (prefixIs(flag, "force=")) {
+ vector<string> encodings =
+ getVectorFromString(flag.substr(6), ";");
+ for (size_t i = 0; i < encodings.size(); ++i)
+ forcedselected[encodings[i]].insert(symbol);
+ flags |= CharInfoForceSelected;
+ } else if (prefixIs(flag, "force!=")) {
+ vector<string> encodings =
+ getVectorFromString(flag.substr(7), ";");
+ for (size_t i = 0; i < encodings.size(); ++i)
+ forcednotselected[encodings[i]].insert(symbol);
+ flags |= CharInfoForceSelected;
} else if (flag == "mathalpha") {
mathalpha.insert(symbol);
+ } else if (flag == "notermination=text") {
+ flags |= CharInfoTextNoTermination;
+ } else if (flag == "notermination=math") {
+ flags |= CharInfoMathNoTermination;
+ } else if (flag == "notermination=both") {
+ flags |= CharInfoTextNoTermination;
+ flags |= CharInfoMathNoTermination;
+ } else if (flag == "notermination=none") {
+ flags &= ~CharInfoTextNoTermination;
+ flags &= ~CharInfoMathNoTermination;
+ } else if (contains(flag, "tipashortcut=")) {
+ tipashortcut = split(flag, '=');
} else {
lyxerr << "Ignoring unknown flag `" << flag
<< "' for symbol `0x"
// 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;
breakout = true;
}
- if (!info.textpreamble.empty())
- info.textfeature = info.textpreamble[0] != '\\';
- if (!info.mathpreamble.empty())
- info.mathfeature = info.mathpreamble[0] != '\\';
-
+ // backward compatibility
+ 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.combining << ' ' << info.textfeature
- << " '" << to_utf8(info.mathcommand) << "' '"
- << info.mathpreamble << "' " << info.mathfeature);
+ << 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;
lex.next();
string const width = lex.getString();
bool fixedwidth = false;
+ bool unsafe = false;
if (width == "fixed")
fixedwidth = true;
else if (width == "variable")
fixedwidth = false;
+ else if (width == "variableunsafe") {
+ fixedwidth = false;
+ unsafe = true;
+ }
else
lex.printError("Unknown width");
LYXERR(Debug::INFO, "Reading encoding " << name);
encodinglist[name] = Encoding(name, latexname,
- guiname, iconvname, fixedwidth, package);
+ guiname, iconvname, fixedwidth, unsafe,
+ package);
if (lex.lex() != et_end)
lex.printError("Missing end");
break;
}
}
+
+ // 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());
+ }
+ }
+
}