X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FParagraph.cpp;h=b5f31281f4da9ee0dff6d11b8825bdb8a68a63f1;hb=3256dbc78a76f947080dc8e3414abc60e4b5c5af;hp=8b3d7df38f039a435d5356c85d083b6f0dcab784;hpb=75febee877693cdf229eadc6326579779bfddfbe;p=lyx.git diff --git a/src/Paragraph.cpp b/src/Paragraph.cpp index 8b3d7df38f..b5f31281f4 100644 --- a/src/Paragraph.cpp +++ b/src/Paragraph.cpp @@ -24,23 +24,20 @@ #include "Changes.h" #include "Counters.h" #include "Encoding.h" -#include "debug.h" -#include "gettext.h" #include "InsetList.h" #include "Language.h" #include "LaTeXFeatures.h" -#include "Color.h" #include "Layout.h" #include "Length.h" #include "Font.h" #include "FontList.h" #include "LyXRC.h" -#include "Messages.h" #include "OutputParams.h" #include "output_latex.h" #include "paragraph_funcs.h" #include "ParagraphParameters.h" #include "sgml.h" +#include "TextClass.h" #include "TexRow.h" #include "VSpace.h" @@ -49,30 +46,22 @@ #include "insets/InsetBibitem.h" #include "insets/InsetLabel.h" -#include "insets/InsetOptArg.h" +#include "support/convert.h" +#include "support/debug.h" +#include "support/gettext.h" #include "support/lstrings.h" +#include "support/Messages.h" #include "support/textutils.h" -#include "support/convert.h" -#include "support/unicode.h" #include #include -using std::endl; -using std::string; -using std::ostream; +using namespace std; +using namespace lyx::support; namespace lyx { -using support::contains; -using support::lowercase; -using support::prefixIs; -using support::suffixIs; -using support::rsplit; -using support::rtrim; -using support::uppercase; - namespace { /// Inset identifier (above 0x10ffff, for ucs-4) char_type const META_INSET = 0x200001; @@ -103,18 +92,18 @@ public: /// Output a space in appropriate formatting (or a surrogate pair /// if the next character is a combining character). /// \return whether a surrogate pair was output. - bool simpleTeXBlanks(Encoding const &, + bool simpleTeXBlanks(OutputParams const &, odocstream &, TexRow & texrow, pos_type i, unsigned int & column, Font const & font, Layout const & style); - /// Output consecutive known unicode chars, belonging to the same - /// language as specified by \p preamble, to \p os starting from \p c. + /// Output consecutive unicode chars, belonging to the same script as + /// specified by the latex macro \p ltx, to \p os starting from \p i. /// \return the number of characters written. - int knownLangChars(odocstream & os, char_type c, string & preamble, - Change &, Encoding const &, pos_type &); + int writeScriptChars(odocstream & os, docstring const & ltx, + Change &, Encoding const &, pos_type & i); /// This could go to ParagraphParameters if we want to. int startTeXParParams(BufferParams const &, odocstream &, TexRow &, @@ -163,17 +152,19 @@ public: bool latexSpecialPhrase( odocstream & os, pos_type & i, - unsigned int & column); + unsigned int & column, + OutputParams & runparams); /// void validate(LaTeXFeatures & features, Layout const & layout) const; - /// - pos_type size() const { return owner_->size(); } + /// Checks if the paragraph contains only text and no inset or font change. + bool onlyText(Buffer const & buf, Font const & outerfont, + pos_type initial) const; /// match a string against a particular point in the paragraph - bool isTextAt(std::string const & str, pos_type pos) const; + bool isTextAt(string const & str, pos_type pos) const; /// Which Paragraph owns us? Paragraph * owner_; @@ -209,14 +200,6 @@ public: }; - - -using std::endl; -using std::upper_bound; -using std::lower_bound; -using std::string; - - // Initialization of the counter for the paragraph id's, unsigned int Paragraph::Private::paragraph_id = 0; @@ -229,10 +212,10 @@ struct special_phrase { }; special_phrase const special_phrases[] = { - { "LyX", from_ascii("\\protect\\LyX{}"), false }, - { "TeX", from_ascii("\\protect\\TeX{}"), true }, - { "LaTeX2e", from_ascii("\\protect\\LaTeXe{}"), true }, - { "LaTeX", from_ascii("\\protect\\LaTeX{}"), true }, + { "LyX", from_ascii("\\LyX{}"), false }, + { "TeX", from_ascii("\\TeX{}"), true }, + { "LaTeX2e", from_ascii("\\LaTeXe{}"), true }, + { "LaTeX", from_ascii("\\LaTeX{}"), true }, }; size_t const phrases_nr = sizeof(special_phrases)/sizeof(special_phrase); @@ -266,15 +249,13 @@ bool Paragraph::isChanged(pos_type start, pos_type end) const } -bool Paragraph::isMergedOnEndOfParDeletion(bool trackChanges) const { +bool Paragraph::isMergedOnEndOfParDeletion(bool trackChanges) const +{ // keep the logic here in sync with the logic of eraseChars() - - if (!trackChanges) { + if (!trackChanges) return true; - } - - Change change = d->changes_.lookup(size()); + Change const change = d->changes_.lookup(size()); return change.type == Change::INSERTED && change.author == 0; } @@ -298,8 +279,8 @@ void Paragraph::setChange(Change const & change) if (change.type != Change::DELETED) { for (pos_type pos = 0; pos < size(); ++pos) { - if (isInset(pos)) - getInset(pos)->setChange(change); + if (Inset * inset = getInset(pos)) + inset->setChange(change); } } } @@ -308,22 +289,18 @@ void Paragraph::setChange(Change const & change) void Paragraph::setChange(pos_type pos, Change const & change) { BOOST_ASSERT(pos >= 0 && pos <= size()); - d->changes_.set(change, pos); // see comment in setChange(Change const &) above - - if (change.type != Change::DELETED && - pos < size() && isInset(pos)) { - getInset(pos)->setChange(change); - } + if (change.type != Change::DELETED && pos < size()) + if (Inset * inset = getInset(pos)) + inset->setChange(change); } Change const & Paragraph::lookupChange(pos_type pos) const { BOOST_ASSERT(pos >= 0 && pos <= size()); - return d->changes_.lookup(pos); } @@ -338,17 +315,15 @@ void Paragraph::acceptChanges(BufferParams const & bparams, pos_type start, switch (lookupChange(pos).type) { case Change::UNCHANGED: // accept changes in nested inset - if (pos < size() && isInset(pos)) - getInset(pos)->acceptChanges(bparams); - + if (Inset * inset = getInset(pos)) + inset->acceptChanges(bparams); break; case Change::INSERTED: d->changes_.set(Change(Change::UNCHANGED), pos); // also accept changes in nested inset - if (pos < size() && isInset(pos)) { - getInset(pos)->acceptChanges(bparams); - } + if (Inset * inset = getInset(pos)) + inset->acceptChanges(bparams); break; case Change::DELETED: @@ -376,9 +351,8 @@ void Paragraph::rejectChanges(BufferParams const & bparams, switch (lookupChange(pos).type) { case Change::UNCHANGED: // reject changes in nested inset - if (pos < size() && isInset(pos)) { - getInset(pos)->rejectChanges(bparams); - } + if (Inset * inset = getInset(pos)) + inset->rejectChanges(bparams); break; case Change::INSERTED: @@ -406,14 +380,14 @@ void Paragraph::rejectChanges(BufferParams const & bparams, void Paragraph::Private::insertChar(pos_type pos, char_type c, Change const & change) { - BOOST_ASSERT(pos >= 0 && pos <= size()); + BOOST_ASSERT(pos >= 0 && pos <= int(text_.size())); // track change changes_.insert(change, pos); // This is actually very common when parsing buffers (and // maybe inserting ascii text) - if (pos == size()) { + if (pos == pos_type(text_.size())) { // when appending characters, no need to update tables text_.push_back(c); return; @@ -523,19 +497,20 @@ int Paragraph::Private::latexSurrogatePair(odocstream & os, char_type c, } -bool Paragraph::Private::simpleTeXBlanks(Encoding const & encoding, +bool Paragraph::Private::simpleTeXBlanks(OutputParams const & runparams, odocstream & os, TexRow & texrow, pos_type i, unsigned int & column, Font const & font, Layout const & style) { - if (style.pass_thru) + if (style.pass_thru || runparams.verbatim) return false; - if (i + 1 < size()) { + if (i + 1 < int(text_.size())) { char_type next = text_[i + 1]; if (Encodings::isCombiningChar(next)) { + Encoding const & encoding = *(runparams.encoding); // This space has an accent, so we must always output it. column += latexSurrogatePair(os, ' ', next, encoding) - 1; return true; @@ -546,12 +521,12 @@ bool Paragraph::Private::simpleTeXBlanks(Encoding const & encoding, && column > lyxrc.plaintext_linelen && i && text_[i - 1] != ' ' - && (i + 1 < size()) + && (i + 1 < int(text_.size())) // same in FreeSpacing mode && !owner_->isFreeSpacing() // In typewriter mode, we want to avoid // ! . ? : at the end of a line - && !(font.family() == Font::TYPEWRITER_FAMILY + && !(font.fontInfo().family() == TYPEWRITER_FAMILY && (text_[i - 1] == '.' || text_[i - 1] == '?' || text_[i - 1] == ':' @@ -569,27 +544,35 @@ bool Paragraph::Private::simpleTeXBlanks(Encoding const & encoding, } -int Paragraph::Private::knownLangChars(odocstream & os, - char_type c, - string & preamble, - Change & runningChange, - Encoding const & encoding, - pos_type & i) -{ - // When the character is marked by the proper language, we simply - // get its code point in some encoding, otherwise we get the - // translation specified in the unicodesymbols file, which is - // something like "\textLANG{}". So, we have to retain - // "\textLANG{" for the first char but only "" for - // all subsequent chars. - docstring const latex1 = rtrim(encoding.latexChar(c), "}"); - int length = latex1.length(); - os << latex1; - while (i + 1 < size()) { - char_type next = text_[i + 1]; - // Stop here if next character belongs to another - // language or there is a change tracking status. - if (!Encodings::isKnownLangChar(next, preamble) || +int Paragraph::Private::writeScriptChars(odocstream & os, + docstring const & ltx, + Change & runningChange, + Encoding const & encoding, + pos_type & i) +{ + // FIXME: modifying i here is not very nice... + + // We only arrive here when a proper language for character text_[i] has + // not been specified (i.e., it could not be translated in the current + // latex encoding) and it belongs to a known script. + // Parameter ltx contains the latex translation of text_[i] as specified in + // the unicodesymbols file and is something like "\textXXX{}". + // The latex macro name "textXXX" specifies the script to which text_[i] + // belongs and we use it in order to check whether characters from the + // same script immediately follow, such that we can collect them in a + // single "\textXXX" macro. So, we have to retain "\textXXX{" + // for the first char but only "" for all subsequent chars. + docstring::size_type const brace1 = ltx.find_first_of(from_ascii("{")); + docstring::size_type const brace2 = ltx.find_last_of(from_ascii("}")); + string script = to_ascii(ltx.substr(1, brace1 - 1)); + int length = ltx.substr(0, brace2).length(); + os << ltx.substr(0, brace2); + int size = text_.size(); + while (i + 1 < size) { + char_type const next = text_[i + 1]; + // Stop here if next character belongs to another script + // or there is a change in change tracking status. + if (!Encodings::isKnownScriptChar(next, script) || runningChange != owner_->lookupChange(i + 1)) break; Font prev_font; @@ -604,27 +587,21 @@ int Paragraph::Private::knownLangChars(odocstream & os, if (cit->pos() >= i + 1) break; } - // Stop here if there is a font attribute change. + // Stop here if there is a font attribute or encoding change. if (found && cit != end && prev_font != cit->font()) break; - docstring const latex = rtrim(encoding.latexChar(next), "}"); - docstring::size_type const j = + docstring const latex = encoding.latexChar(next); + docstring::size_type const b1 = latex.find_first_of(from_ascii("{")); - if (j == docstring::npos) { - os << latex; - length += latex.length(); - } else { - os << latex.substr(j + 1); - length += latex.substr(j + 1).length(); - } + docstring::size_type const b2 = + latex.find_last_of(from_ascii("}")); + int const len = b2 - b1 - 1; + os << latex.substr(b1 + 1, len); + length += len; ++i; } - // When the proper language is set, we are simply passed a code - // point, so we should not try to close the \textLANG command. - if (prefixIs(latex1, from_ascii("\\" + preamble))) { - os << '}'; - ++length; - } + os << '}'; + ++length; return length; } @@ -634,7 +611,7 @@ bool Paragraph::Private::isTextAt(string const & str, pos_type pos) const pos_type const len = str.length(); // is the paragraph large enough? - if (pos + len > size()) + if (pos + len > int(text_.size())) return false; // does the wanted text start at point? @@ -685,7 +662,7 @@ void Paragraph::Private::latexInset(Buffer const & buf, open_font = false; } - if (running_font.family() == Font::TYPEWRITER_FAMILY) + if (running_font.fontInfo().family() == TYPEWRITER_FAMILY) os << '~'; basefont = owner_->getLayoutFont(bparams, outerfont); @@ -694,7 +671,6 @@ void Paragraph::Private::latexInset(Buffer const & buf, if (runparams.moving_arg) os << "\\protect "; - os << "\\\\\n"; } texrow.newline(); texrow.start(owner_->id(), i + 1); @@ -715,10 +691,10 @@ void Paragraph::Private::latexInset(Buffer const & buf, bool close = false; odocstream::pos_type const len = os.tellp(); - if ((inset->lyxCode() == GRAPHICS_CODE - || inset->lyxCode() == MATH_CODE - || inset->lyxCode() == HYPERLINK_CODE) - && running_font.isRightToLeft()) { + if (inset->forceLTR() + && running_font.isRightToLeft() + // ERT is an exception, it should be output with no decorations at all + && inset->lyxCode() != ERT_CODE) { if (running_font.language()->lang() == "farsi") os << "\\beginL{}"; else @@ -812,7 +788,7 @@ void Paragraph::Private::latexSpecialChar( if (lyxrc.fontenc == "T1" && latexSpecialT1(c, os, i, column)) return; - if (running_font.family() == Font::TYPEWRITER_FAMILY + if (running_font.fontInfo().family() == TYPEWRITER_FAMILY && latexSpecialTypewriter(c, os, i, column)) return; @@ -879,14 +855,14 @@ void Paragraph::Private::latexSpecialChar( default: // LyX, LaTeX etc. - if (latexSpecialPhrase(os, i, column)) + if (latexSpecialPhrase(os, i, column, runparams)) return; if (c == '\0') return; Encoding const & encoding = *(runparams.encoding); - if (i + 1 < size()) { + if (i + 1 < int(text_.size())) { char_type next = text_[i + 1]; if (Encodings::isCombiningChar(next)) { column += latexSurrogatePair(os, c, next, encoding) - 1; @@ -894,14 +870,13 @@ void Paragraph::Private::latexSpecialChar( break; } } - string preamble; - if (Encodings::isKnownLangChar(c, preamble)) { - column += knownLangChars(os, c, preamble, running_change, - encoding, i) - 1; - break; - } + string script; docstring const latex = encoding.latexChar(c); - if (latex.length() > 1 && latex[latex.length() - 1] != '}') { + if (Encodings::isKnownScriptChar(c, script) + && prefixIs(latex, from_ascii("\\" + script))) + column += writeScriptChars(os, latex, + running_change, encoding, i) - 1; + else if (latex.length() > 1 && latex[latex.length() - 1] != '}') { // Prevent eating of a following // space or command corruption by // following characters @@ -925,7 +900,7 @@ bool Paragraph::Private::latexSpecialT1(char_type const c, odocstream & os, os.put(c); // In T1 encoding, these characters exist // but we should avoid ligatures - if (i + 1 > size() || text_[i + 1] != c) + if (i + 1 >= int(text_.size()) || text_[i + 1] != c) return true; os << "\\,{}"; column += 3; @@ -947,7 +922,7 @@ bool Paragraph::Private::latexSpecialTypewriter(char_type const c, odocstream & { switch (c) { case '-': - if (i + 1 < size() && text_[i + 1] == '-') { + if (i + 1 < int(text_.size()) && text_[i + 1] == '-') { // "--" in Typewriter mode -> "-{}-" os << "-{}"; column += 2; @@ -983,7 +958,7 @@ bool Paragraph::Private::latexSpecialTypewriter(char_type const c, odocstream & bool Paragraph::Private::latexSpecialPhrase(odocstream & os, pos_type & i, - unsigned int & column) + unsigned int & column, OutputParams & runparams) { // FIXME: if we have "LaTeX" with a font // change in the middle (before the 'T', then @@ -994,6 +969,8 @@ bool Paragraph::Private::latexSpecialPhrase(odocstream & os, pos_type & i, for (size_t pnr = 0; pnr < phrases_nr; ++pnr) { if (!isTextAt(special_phrases[pnr].phrase, i)) continue; + if (runparams.moving_arg) + os << "\\protect"; os << special_phrases[pnr].macro; i += special_phrases[pnr].phrase.length() - 1; column += special_phrases[pnr].macro.length() - 1; @@ -1033,7 +1010,7 @@ void Paragraph::Private::validate(LaTeXFeatures & features, } // then the contents - for (pos_type i = 0; i < size() ; ++i) { + for (pos_type i = 0; i < int(text_.size()) ; ++i) { for (size_t pnr = 0; pnr < phrases_nr; ++pnr) { if (!special_phrases[pnr].builtin && isTextAt(special_phrases[pnr].phrase, i)) { @@ -1109,7 +1086,7 @@ void Paragraph::write(Buffer const & buf, ostream & os, params().write(os); - Font font1(Font::ALL_INHERIT, bparams.language); + Font font1(inherit_font, bparams.language); Change running_change = Change(Change::UNCHANGED); @@ -1134,9 +1111,7 @@ void Paragraph::write(Buffer const & buf, ostream & os, char_type const c = d->text_[i]; switch (c) { case META_INSET: - { - Inset const * inset = getInset(i); - if (inset) + if (Inset const * inset = getInset(i)) { if (inset->directWrite()) { // international char, let it write // code directly so it's shorter in @@ -1150,8 +1125,8 @@ void Paragraph::write(Buffer const & buf, ostream & os, os << "\n\\end_inset\n\n"; column = 0; } - } - break; + } + break; case '\\': os << "\n\\backslash\n"; column = 0; @@ -1171,11 +1146,9 @@ void Paragraph::write(Buffer const & buf, ostream & os, } // this check is to amend a bug. LyX sometimes // inserts '\0' this could cause problems. - if (c != '\0') { - std::vector tmp = ucs4_to_utf8(c); - tmp.push_back('\0'); - os << &tmp[0]; - } else + if (c != '\0') + os << to_utf8(docstring(1, c)); + else lyxerr << "ERROR (Paragraph::writeFile):" " NULL char in structure." << endl; ++column; @@ -1215,12 +1188,12 @@ void Paragraph::appendChar(char_type c, Font const & font, void Paragraph::appendString(docstring const & s, Font const & font, Change const & change) { - size_t end = s.size(); + pos_type end = s.size(); size_t oldsize = d->text_.size(); size_t newsize = oldsize + end; size_t capacity = d->text_.capacity(); if (newsize >= capacity) - d->text_.reserve(std::max(capacity + 100, newsize)); + d->text_.reserve(max(capacity + 100, newsize)); // when appending characters, no need to update tables d->text_.append(s); @@ -1230,7 +1203,8 @@ void Paragraph::appendString(docstring const & s, Font const & font, // track change d->changes_.insert(change, i); } - d->fontlist_.setRange(oldsize, newsize, font); + d->fontlist_.set(oldsize, font); + d->fontlist_.set(newsize - 1, font); } @@ -1274,6 +1248,13 @@ bool Paragraph::insetAllowed(InsetCode code) } +void Paragraph::resetFonts(Font const & font) +{ + d->fontlist_.clear(); + d->fontlist_.set(0, font); + d->fontlist_.set(d->text_.size() - 1, font); +} + // Gets uninstantiated font setting at position. Font const Paragraph::getFontSettings(BufferParams const & bparams, pos_type pos) const @@ -1290,7 +1271,7 @@ Font const Paragraph::getFontSettings(BufferParams const & bparams, if (pos == size() && !empty()) return getFontSettings(bparams, pos - 1); - return Font(Font::ALL_INHERIT, getParLanguage(bparams)); + return Font(inherit_font, getParLanguage(bparams)); } @@ -1304,11 +1285,11 @@ FontSpan Paragraph::fontSpan(pos_type pos) const for (; cit != end; ++cit) { if (cit->pos() >= pos) { if (pos >= beginOfBody()) - return FontSpan(std::max(start, beginOfBody()), + return FontSpan(max(start, beginOfBody()), cit->pos()); else return FontSpan(start, - std::min(beginOfBody() - 1, + min(beginOfBody() - 1, cit->pos())); } start = cit->pos() + 1; @@ -1327,7 +1308,7 @@ Font const Paragraph::getFirstFontSettings(BufferParams const & bparams) const if (!empty() && !d->fontlist_.empty()) return d->fontlist_.begin()->font(); - return Font(Font::ALL_INHERIT, bparams.language); + return Font(inherit_font, bparams.language); } @@ -1345,12 +1326,12 @@ Font const Paragraph::getFont(BufferParams const & bparams, pos_type pos, pos_type const body_pos = beginOfBody(); if (pos < body_pos) - font.realize(d->layout_->labelfont); + font.fontInfo().realize(d->layout_->labelfont); else - font.realize(d->layout_->font); + font.fontInfo().realize(d->layout_->font); - font.realize(outerfont); - font.realize(bparams.getFont()); + font.fontInfo().realize(outerfont.fontInfo()); + font.fontInfo().realize(bparams.getFont().fontInfo()); return font; } @@ -1359,28 +1340,26 @@ Font const Paragraph::getFont(BufferParams const & bparams, pos_type pos, Font const Paragraph::getLabelFont (BufferParams const & bparams, Font const & outerfont) const { - Font tmpfont = layout()->labelfont; - tmpfont.setLanguage(getParLanguage(bparams)); - tmpfont.realize(outerfont); - tmpfont.realize(bparams.getFont()); - return tmpfont; + FontInfo tmpfont = layout()->labelfont; + tmpfont.realize(outerfont.fontInfo()); + tmpfont.realize(bparams.getFont().fontInfo()); + return Font(tmpfont, getParLanguage(bparams)); } Font const Paragraph::getLayoutFont (BufferParams const & bparams, Font const & outerfont) const { - Font tmpfont = layout()->font; - tmpfont.setLanguage(getParLanguage(bparams)); - tmpfont.realize(outerfont); - tmpfont.realize(bparams.getFont()); - return tmpfont; + FontInfo tmpfont = layout()->font; + tmpfont.realize(outerfont.fontInfo()); + tmpfont.realize(bparams.getFont().fontInfo()); + return Font(tmpfont, getParLanguage(bparams)); } /// Returns the height of the highest font in range -Font_size Paragraph::highestFontInRange - (pos_type startpos, pos_type endpos, Font_size def_size) const +FontSize Paragraph::highestFontInRange + (pos_type startpos, pos_type endpos, FontSize def_size) const { return d->fontlist_.highestInRange(startpos, endpos, def_size); } @@ -1521,7 +1500,7 @@ void Paragraph::setLabelWidthString(docstring const & s) docstring const Paragraph::translateIfPossible(docstring const & s, BufferParams const & bparams) const { - if (!support::isAscii(s) || s.empty()) { + if (!isAscii(s) || s.empty()) { // This must be a user defined layout. We cannot translate // this, since gettext accepts only ascii keys. return s; @@ -1622,17 +1601,6 @@ void Paragraph::setBeginOfBody() } -InsetBibitem * Paragraph::bibitem() const -{ - if (!d->insetlist_.empty()) { - Inset * inset = d->insetlist_.begin()->inset; - if (inset->lyxCode() == BIBITEM_CODE) - return static_cast(inset); - } - return 0; -} - - bool Paragraph::forceDefaultParagraphs() const { return inInset() && inInset()->forceDefaultParagraphs(0); @@ -1826,7 +1794,7 @@ bool Paragraph::latex(Buffer const & buf, odocstream & os, TexRow & texrow, OutputParams const & runparams) const { - LYXERR(Debug::LATEX) << "SimpleTeXOnePar... " << this << endl; + LYXERR(Debug::LATEX, "SimpleTeXOnePar... " << this); bool return_value = false; @@ -1959,13 +1927,22 @@ bool Paragraph::latex(Buffer const & buf, open_font = false; } + // close babel's font environment before opening CJK. + if (!running_font.language()->babel().empty() && + font.language()->encoding()->package() == Encoding::CJK) { + string end_tag = subst(lyxrc.language_command_end, + "$$lang", + running_font.language()->babel()); + os << from_ascii(end_tag); + column += end_tag.length(); + } + // Switch file encoding if necessary (and allowed) if (!runparams.verbatim && - runparams.encoding->package() == Encoding::inputenc && - font.language()->encoding()->package() == Encoding::inputenc) { - std::pair const enc_switch = switchEncoding(os, bparams, - runparams.moving_arg, *(runparams.encoding), - *(font.language()->encoding())); + runparams.encoding->package() == Encoding::none && + font.language()->encoding()->package() == Encoding::none) { + pair const enc_switch = switchEncoding(os, bparams, + runparams, *(font.language()->encoding())); if (enc_switch.first) { column += enc_switch.second; runparams.encoding = font.language()->encoding(); @@ -2003,7 +1980,7 @@ bool Paragraph::latex(Buffer const & buf, // style->pass_thru is false. if (i != body_pos - 1) { if (d->simpleTeXBlanks( - *(runparams.encoding), os, texrow, + runparams, os, texrow, i, column, font, *style)) { // A surrogate pair was output. We // must not call latexSpecialChar @@ -2028,9 +2005,24 @@ bool Paragraph::latex(Buffer const & buf, texrow, rp, running_font, basefont, outerfont, open_font, runningChange, *style, i, column); - else - d->latexSpecialChar(os, rp, running_font, runningChange, - *style, i, column); + else { + try { + d->latexSpecialChar(os, rp, running_font, runningChange, + *style, i, column); + } catch (EncodingException & e) { + if (runparams.dryrun) { + os << "<" << _("LyX Warning: ") + << _("uncodable character") << " '"; + os.put(c); + os << "'>"; + } else { + // add location information and throw again. + e.par_id = id(); + e.pos = i; + throw(e); + } + } + } // Set the encoding to that returned from simpleTeXSpecialChars (see // comment for encoding member in OutputParams.h) @@ -2071,59 +2063,15 @@ bool Paragraph::latex(Buffer const & buf, runparams.moving_arg); } - LYXERR(Debug::LATEX) << "SimpleTeXOnePar...done " << this << endl; + LYXERR(Debug::LATEX, "SimpleTeXOnePar...done " << this); return return_value; } -namespace { - -enum PAR_TAG { - PAR_NONE=0, - TT = 1, - SF = 2, - BF = 4, - IT = 8, - SL = 16, - EM = 32 -}; - - -string tag_name(PAR_TAG const & pt) { - switch (pt) { - case PAR_NONE: return "!-- --"; - case TT: return "tt"; - case SF: return "sf"; - case BF: return "bf"; - case IT: return "it"; - case SL: return "sl"; - case EM: return "em"; - } - return ""; -} - - -inline -void operator|=(PAR_TAG & p1, PAR_TAG const & p2) -{ - p1 = static_cast(p1 | p2); -} - - -inline -void reset(PAR_TAG & p1, PAR_TAG const & p2) -{ - p1 = static_cast(p1 & ~p2); -} - -} // anon - - bool Paragraph::emptyTag() const { for (pos_type i = 0; i < size(); ++i) { - if (isInset(i)) { - Inset const * inset = getInset(i); + if (Inset const * inset = getInset(i)) { InsetCode lyx_code = inset->lyxCode(); if (lyx_code != TOC_CODE && lyx_code != INCLUDE_CODE && @@ -2147,8 +2095,7 @@ bool Paragraph::emptyTag() const string Paragraph::getID(Buffer const & buf, OutputParams const & runparams) const { for (pos_type i = 0; i < size(); ++i) { - if (isInset(i)) { - Inset const * inset = getInset(i); + if (Inset const * inset = getInset(i)) { InsetCode lyx_code = inset->lyxCode(); if (lyx_code == LABEL_CODE) { InsetLabel const * const il = static_cast(inset); @@ -2156,7 +2103,6 @@ string Paragraph::getID(Buffer const & buf, OutputParams const & runparams) cons return "id='" + to_utf8(sgml::cleanID(buf, runparams, id)) + "'"; } } - } return string(); } @@ -2166,8 +2112,7 @@ pos_type Paragraph::getFirstWord(Buffer const & buf, odocstream & os, OutputPara { pos_type i; for (i = 0; i < size(); ++i) { - if (isInset(i)) { - Inset const * inset = getInset(i); + if (Inset const * inset = getInset(i)) { inset->docbook(buf, os, runparams); } else { char_type c = d->text_[i]; @@ -2180,13 +2125,13 @@ pos_type Paragraph::getFirstWord(Buffer const & buf, odocstream & os, OutputPara } -bool Paragraph::onlyText(Buffer const & buf, Font const & outerfont, pos_type initial) const +bool Paragraph::Private::onlyText(Buffer const & buf, Font const & outerfont, pos_type initial) const { Font font_old; - - for (pos_type i = initial; i < size(); ++i) { - Font font = getFont(buf.params(), i, outerfont); - if (isInset(i)) + pos_type size = text_.size(); + for (pos_type i = initial; i < size; ++i) { + Font font = owner_->getFont(buf.params(), i, outerfont); + if (text_[i] == META_INSET) return false; if (i != initial && font != font_old) return false; @@ -2206,10 +2151,10 @@ void Paragraph::simpleDocBookOnePar(Buffer const & buf, bool emph_flag = false; LayoutPtr const & style = layout(); - Font font_old = + FontInfo font_old = style->labeltype == LABEL_MANUAL ? style->labelfont : style->font; - if (style->pass_thru && !onlyText(buf, outerfont, initial)) + if (style->pass_thru && !d->onlyText(buf, outerfont, initial)) os << "]]>"; // parsing main loop @@ -2217,8 +2162,8 @@ void Paragraph::simpleDocBookOnePar(Buffer const & buf, Font font = getFont(buf.params(), i, outerfont); // handle tag - if (font_old.emph() != font.emph()) { - if (font.emph() == Font::ON) { + if (font_old.emph() != font.fontInfo().emph()) { + if (font.fontInfo().emph() == FONT_ON) { os << ""; emph_flag = true; } else if (i != initial) { @@ -2227,8 +2172,7 @@ void Paragraph::simpleDocBookOnePar(Buffer const & buf, } } - if (isInset(i)) { - Inset const * inset = getInset(i); + if (Inset const * inset = getInset(i)) { inset->docbook(buf, os, runparams); } else { char_type c = d->text_[i]; @@ -2238,7 +2182,7 @@ void Paragraph::simpleDocBookOnePar(Buffer const & buf, else os << sgml::escapeChar(c); } - font_old = font; + font_old = font.fontInfo(); } if (emph_flag) { @@ -2247,43 +2191,42 @@ void Paragraph::simpleDocBookOnePar(Buffer const & buf, if (style->free_spacing) os << '\n'; - if (style->pass_thru && !onlyText(buf, outerfont, initial)) + if (style->pass_thru && !d->onlyText(buf, outerfont, initial)) os << "lyxCode() == HFILL_CODE; + Inset const * inset = getInset(pos); + return inset && inset->lyxCode() == HFILL_CODE; } bool Paragraph::isNewline(pos_type pos) const { - return isInset(pos) - && getInset(pos)->lyxCode() == NEWLINE_CODE; + Inset const * inset = getInset(pos); + return inset && inset->lyxCode() == NEWLINE_CODE; } bool Paragraph::isLineSeparator(pos_type pos) const { char_type const c = d->text_[pos]; - return isLineSeparatorChar(c) - || (c == META_INSET && getInset(pos) && - getInset(pos)->isLineSeparator()); + if (isLineSeparatorChar(c)) + return true; + Inset const * inset = getInset(pos); + return inset && inset->isLineSeparator(); } /// Used by the spellchecker bool Paragraph::isLetter(pos_type pos) const { - if (isInset(pos)) - return getInset(pos)->isLetter(); - else { - char_type const c = d->text_[pos]; - return isLetterChar(c) || isDigit(c); - } + if (Inset const * inset = getInset(pos)) + return inset->isLetter(); + char_type const c = d->text_[pos]; + return isLetterChar(c) || isDigit(c); } @@ -2396,8 +2339,7 @@ Inset * Paragraph::inInset() const InsetCode Paragraph::ownerCode() const { - return d->inset_owner_ ? - d->inset_owner_->lyxCode() : NO_CODE; + return d->inset_owner_ ? d->inset_owner_->lyxCode() : NO_CODE; } @@ -2417,10 +2359,7 @@ bool Paragraph::isFreeSpacing() const { if (layout()->free_spacing) return true; - - // for now we just need this, later should we need this in some - // other way we can always add a function to Inset too. - return ownerCode() == ERT_CODE || ownerCode() == LISTINGS_CODE; + return d->inset_owner_ && d->inset_owner_->isFreeSpacing(); } @@ -2428,7 +2367,7 @@ bool Paragraph::allowEmpty() const { if (layout()->keepempty) return true; - return ownerCode() == ERT_CODE || ownerCode() == LISTINGS_CODE; + return d->inset_owner_ && d->inset_owner_->allowEmpty(); } @@ -2580,31 +2519,20 @@ Inset * Paragraph::releaseInset(pos_type pos) Inset * Paragraph::getInset(pos_type pos) { - return d->insetlist_.get(pos); + return (pos < pos_type(d->text_.size()) && d->text_[pos] == META_INSET) + ? d->insetlist_.get(pos) : 0; } Inset const * Paragraph::getInset(pos_type pos) const { - return d->insetlist_.get(pos); -} - - -int Paragraph::numberOfOptArgs() const -{ - int num = 0; - InsetList::const_iterator it = insetList().begin(); - InsetList::const_iterator end = insetList().end(); - for (; it != end ; ++it) { - if (it->inset->lyxCode() == OPTARG_CODE) - ++num; - } - return num; + return (pos < pos_type(d->text_.size()) && d->text_[pos] == META_INSET) + ? d->insetlist_.get(pos) : 0; } void Paragraph::changeCase(BufferParams const & bparams, pos_type pos, - pos_type right, TextCase action) + pos_type & right, TextCase action) { // process sequences of modified characters; in change // tracking mode, this approach results in much better @@ -2620,7 +2548,7 @@ void Paragraph::changeCase(BufferParams const & bparams, pos_type pos, char_type newChar = oldChar; // ignore insets and don't play with deleted text! - if (isInset(pos) && !isDeleted(pos)) { + if (oldChar != META_INSET && !isDeleted(pos)) { switch (action) { case text_lowercase: newChar = lowercase(oldChar);