#include "ParagraphParameters.h"
#include "SpellChecker.h"
#include "sgml.h"
+#include "texstream.h"
#include "TextClass.h"
#include "TexRow.h"
#include "Text.h"
#include "insets/InsetLabel.h"
#include "insets/InsetSpecialChar.h"
+#include "mathed/InsetMathHull.h"
+
#include "support/debug.h"
#include "support/docstring_list.h"
#include "support/ExceptionMessage.h"
#include "support/lstrings.h"
#include "support/textutils.h"
+#include <atomic>
#include <sstream>
#include <vector>
class Paragraph::Private
{
- // Enforce our own "copy" constructor by declaring the standard one and
- // the assignment operator private without implementing them.
- Private(Private const &);
- Private & operator=(Private const &);
+ // Enforce our own "copy" constructor
+ Private(Private const &) = delete;
+ Private & operator=(Private const &) = delete;
+ // Unique ID generator
+ static int make_id();
public:
///
Private(Paragraph * owner, Layout const & layout);
pos_type i,
unsigned int & column);
///
+ bool latexSpecialTU(
+ char_type const c,
+ otexstream & os,
+ pos_type i,
+ unsigned int & column);
+ ///
bool latexSpecialT3(
char_type const c,
otexstream & os,
}
-// Initialization of the counter for the paragraph id's,
-//
-// FIXME: There should be a more intelligent way to generate and use the
-// paragraph ids per buffer instead a global static counter for all InsetText
-// in the running program.
-static int paragraph_id = -1;
+//static
+int Paragraph::Private::make_id()
+{
+ // The id is unique per session across buffers because it is used in
+ // LFUN_PARAGRAPH_GOTO to switch to a different buffer, for instance in the
+ // outliner.
+ // (thread-safe)
+ static atomic_uint next_id(0);
+ return next_id++;
+}
+
Paragraph::Private::Private(Private const & p, Paragraph * owner)
: owner_(owner), inset_owner_(p.inset_owner_), fontlist_(p.fontlist_),
+ id_(make_id()),
params_(p.params_), changes_(p.changes_), insetlist_(p.insetlist_),
begin_of_body_(p.begin_of_body_), text_(p.text_), words_(p.words_),
layout_(p.layout_)
{
- id_ = ++paragraph_id;
requestSpellCheck(p.text_.size());
}
Paragraph::Private::Private(Private const & p, Paragraph * owner,
pos_type beg, pos_type end)
- : owner_(owner), inset_owner_(p.inset_owner_),
+ : owner_(owner), inset_owner_(p.inset_owner_), id_(make_id()),
params_(p.params_), changes_(p.changes_),
insetlist_(p.insetlist_, beg, end),
begin_of_body_(p.begin_of_body_), words_(p.words_),
layout_(p.layout_)
{
- id_ = ++paragraph_id;
if (beg >= pos_type(p.text_.size()))
return;
text_ = p.text_.substr(beg, end - beg);
}
-void Paragraph::addChangesToToc(DocIterator const & cdit,
- Buffer const & buf, bool output_active) const
+void Paragraph::addChangesToToc(DocIterator const & cdit, Buffer const & buf,
+ bool output_active, TocBackend & backend) const
+{
+ d->changes_.addToToc(cdit, buf, output_active, backend);
+}
+
+
+void Paragraph::addChangesToBuffer(Buffer const & buf) const
{
- d->changes_.addToToc(cdit, buf, output_active);
+ d->changes_.updateBuffer(buf);
+}
+
+
+bool Paragraph::isChangeUpdateRequired() const
+{
+ return d->changes_.isUpdateRequired();
}
{
// 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) or its latex translation has been forced, and it
- // belongs to a known script.
+ // We only arrive here when character text_[i] could not be translated
+ // into the current latex encoding (or its latex translation has been forced,)
+ // and it belongs to a known script.
+ // TODO: We need \textcyr and \textgreek wrappers also for characters
+ // that can be encoded in the "LaTeX encoding" but not in the
+ // current *font encoding*.
+ // (See #9681 for details and test)
// Parameter ltx contains the latex translation of text_[i] as specified
// in the unicodesymbols file and is something like "\textXXX{<spec>}".
// The latex macro name "textXXX" specifies the script to which text_[i]
bool closing_brace = true;
if (script == "textgreek" && encoding.latexName() == "iso-8859-7") {
// Correct encoding is being used, so we can avoid \textgreek.
+ // TODO: wrong test: we need to check the *font encoding*
+ // (i.e. the active language and its FontEncoding tag)
+ // instead of the LaTeX *input encoding*!
+ // See #9637 for details and test-cases.
pos = brace1 + 1;
length -= pos;
closing_brace = false;
// decorations at all
&& inset->lyxCode() != ERT_CODE) {
if (running_font.language()->lang() == "farsi")
- os << "\\beginL{}";
+ os << "\\beginL" << termcmd;
else
os << "\\L{";
close = true;
}
}
- int prev_rows = os.texrow().rows();
+ size_t const previous_row_count = os.texrow().rows();
try {
runparams.lastid = id_;
if (close) {
if (running_font.language()->lang() == "farsi")
- os << "\\endL{}";
+ os << "\\endL" << termcmd;
else
os << '}';
}
- if (os.texrow().rows() > prev_rows) {
+ if (os.texrow().rows() > previous_row_count) {
os.texrow().start(owner_->id(), i + 1);
column = 0;
} else {
char_type const c = (runparams.use_polyglossia) ?
owner_->getUChar(bparams, i) : text_[i];
- if (style.pass_thru || runparams.pass_thru) {
+ if (style.pass_thru || runparams.pass_thru
+ || contains(style.pass_thru_chars, c)
+ || contains(runparams.pass_thru_chars, c)) {
if (c != '\0') {
Encoding const * const enc = runparams.encoding;
if (enc && !enc->encodable(c))
return;
// If T1 font encoding is used, use the special
// characters it provides.
- // NOTE: Some languages reset the font encoding internally.
- // If we are using such a language, we do not output
- // special T1 chars.
+ // NOTE: Some languages reset the font encoding internally to a
+ // non-standard font encoding. If we are using such a language,
+ // we do not output special T1 chars.
if (!runparams.inIPA && !running_font.language()->internalFontEncoding()
- && bparams.font_encoding() == "T1" && latexSpecialT1(c, os, i, column))
+ && !runparams.isFullUnicode() && bparams.main_font_encoding() == "T1"
+ && latexSpecialT1(c, os, i, column))
return;
+ // NOTE: XeTeX and LuaTeX use EU1/2 (pre 2017) or TU (as of 2017) encoding
+ else if (!runparams.inIPA && !running_font.language()->internalFontEncoding()
+ && runparams.isFullUnicode() && latexSpecialTU(c, os, i, column))
+ return;
// Otherwise, we use what LaTeX provides us.
switch (c) {
case '\\':
- os << "\\textbackslash{}";
+ os << "\\textbackslash" << termcmd;
column += 15;
break;
case '<':
- os << "\\textless{}";
+ os << "\\textless" << termcmd;
column += 10;
break;
case '>':
- os << "\\textgreater{}";
+ os << "\\textgreater" << termcmd;
column += 13;
break;
case '|':
- os << "\\textbar{}";
+ os << "\\textbar" << termcmd;
column += 9;
break;
case '-':
os << '-';
- if (i + 1 < end_pos && text_[i+1] == '-') {
+ if (i + 1 < static_cast<pos_type>(text_.size()) &&
+ (end_pos == -1 || i + 1 < end_pos) &&
+ text_[i+1] == '-') {
// Prevent "--" becoming an endash and "---" becoming
// an emdash.
// Within \ttfamily, "--" is merged to "-" (no endash)
}
break;
case '\"':
- os << "\\char`\\\"{}";
+ os << "\\char34" << termcmd;
column += 9;
break;
break;
case '~':
- os << "\\textasciitilde{}";
+ os << "\\textasciitilde" << termcmd;
column += 16;
break;
case '^':
- os << "\\textasciicircum{}";
+ os << "\\textasciicircum" << termcmd;
column += 17;
break;
// written. (Asger)
break;
+ case 0x2013:
+ case 0x2014:
+ if (bparams.use_dash_ligatures && !bparams.useNonTeXFonts) {
+ if (c == 0x2013) {
+ // en-dash
+ os << "--";
+ column +=2;
+ } else {
+ // em-dash
+ os << "---";
+ column +=3;
+ }
+ break;
+ }
+ // fall through
default:
if (c == '\0')
return;
// but we should avoid ligatures
if (i + 1 >= int(text_.size()) || text_[i + 1] != c)
return true;
- os << "\\textcompwordmark{}";
+ os << "\\textcompwordmark" << termcmd;
column += 19;
return true;
case '|':
return true;
case '\"':
// soul.sty breaks with \char`\"
- os << "\\textquotedbl{}";
+ os << "\\textquotedbl" << termcmd;
column += 14;
return true;
default:
}
+bool Paragraph::Private::latexSpecialTU(char_type const c, otexstream & os,
+ pos_type i, unsigned int & column)
+{
+ // TU encoding is currently on par with T1.
+ return latexSpecialT1(c, os, i, column);
+}
+
+
bool Paragraph::Private::latexSpecialT3(char_type const c, otexstream & os,
pos_type /*i*/, unsigned int & column)
{
os.put(c);
return true;
case '|':
- os << "\\textvertline{}";
+ os << "\\textvertline" << termcmd;
column += 14;
return true;
default:
BufferParams const & bp = features.runparams().is_child
? buf.masterParams() : buf.params();
Font f;
- TexRow texrow;
// Using a string stream here circumvents the encoding
// switching machinery of odocstream. Therefore the
// output is wrong if this paragraph contains content
// that needs to switch encoding.
- odocstringstream ods;
- otexstream os(ods, texrow);
+ otexstringstream os;
+ os << layout_->preamble();
if (is_command) {
os << '\\' << from_ascii(layout_->latexname());
// we have to provide all the optional arguments here, even though
}
os << from_ascii(layout_->latexparam());
}
- docstring::size_type const length = ods.str().length();
+ size_t const length = os.length();
// this will output "{" at the beginning, but not at the end
owner_->latex(bp, f, os, features.runparams(), 0, -1, true);
- if (ods.str().length() > length) {
+ if (os.length() > length) {
if (is_command) {
- ods << '}';
+ os << '}';
if (!layout_->postcommandargs().empty()) {
OutputParams rp = features.runparams();
rp.local_font = &owner_->getFirstFontSettings(bp);
latexArgInsets(*owner_, os, rp, layout_->postcommandargs(), "post:");
}
}
- string const snippet = to_utf8(ods.str());
- features.addPreambleSnippet(snippet);
+ features.addPreambleSnippet(os.release(), true);
}
}
InsetList::const_iterator iend = insetlist_.end();
for (; icit != iend; ++icit) {
if (icit->inset) {
+ features.inDeletedInset(owner_->isDeleted(icit->pos));
icit->inset->validate(features);
+ features.inDeletedInset(false);
if (layout_->needprotect &&
icit->inset->lyxCode() == FOOT_CODE)
- features.require("footmisc");
+ features.require("NeedLyXFootnoteCode");
}
}
FontSpan Paragraph::fontSpan(pos_type pos) const
{
- LBUFERR(pos < size());
+ LBUFERR(pos <= size());
+
+ if (pos == size())
+ return FontSpan(pos, pos);
pos_type start = 0;
FontList::const_iterator cit = d->fontlist_.begin();
}
-/// Returns the height of the highest font in range
-FontSize Paragraph::highestFontInRange
- (pos_type startpos, pos_type endpos, FontSize def_size) const
-{
- return d->fontlist_.highestInRange(startpos, endpos, def_size);
-}
-
-
char_type Paragraph::getUChar(BufferParams const & bparams, pos_type pos) const
{
char_type c = d->text_[pos];
}
-char Paragraph::getAlign() const
+LyXAlignment Paragraph::getAlign() const
{
if (d->params_.align() == LYX_ALIGN_LAYOUT)
return d->layout_->align;
docstring parent(fmt, i + 1, j - i - 1);
docstring label = from_ascii("??");
if (tclass.hasLayout(parent))
- docstring label = expandParagraphLabel(tclass[parent], bparams,
+ label = expandParagraphLabel(tclass[parent], bparams,
process_appendix);
fmt = docstring(fmt, 0, i) + label
+ docstring(fmt, j + 1, docstring::npos);
pos_type end = size();
if (i < end && !(isNewline(i) || isEnvSeparator(i))) {
++i;
- char_type previous_char = 0;
- char_type temp = 0;
if (i < end) {
- previous_char = d->text_[i];
+ char_type previous_char = d->text_[i];
if (!(isNewline(i) || isEnvSeparator(i))) {
++i;
while (i < end && previous_char != ' ') {
- temp = d->text_[i];
+ char_type temp = d->text_[i];
if (isNewline(i) || isEnvSeparator(i))
break;
++i;
{
int column = 0;
- if (params_.noindent() && !layout_->pass_thru
- && (layout_->toggle_indent != ITOGGLE_NEVER)) {
+ bool canindent =
+ (bparams.paragraph_separation == BufferParams::ParagraphIndentSeparation) ?
+ (layout_->toggle_indent != ITOGGLE_NEVER) :
+ (layout_->toggle_indent == ITOGGLE_ALWAYS);
+
+ if (canindent && params_.noindent() && !layout_->pass_thru) {
os << "\\noindent ";
column += 10;
}
corrected_env(os, begin_tag, "flushright", code, lastpar, column);
break;
} case LYX_ALIGN_RIGHT: {
- string output;
if (owner_->getParLanguage(bparams)->babel() != "hebrew")
corrected_env(os, begin_tag, "flushright", code, lastpar, column);
else
runparams);
}
+ runparams.wasDisplayMath = runparams.inDisplayMath;
+ runparams.inDisplayMath = false;
+ bool deleted_display_math = false;
+
+ // Check whether a display math inset follows
+ if (d->text_[i] == META_INSET
+ && i >= start_pos && (end_pos == -1 || i < end_pos)) {
+ InsetMath const * im = getInset(i)->asInsetMath();
+ if (im && im->asHullInset()
+ && im->asHullInset()->outerDisplay()) {
+ runparams.inDisplayMath = true;
+ // runparams.inDeletedInset will be set by
+ // latexInset later, but we need this info
+ // before it is called. On the other hand, we
+ // cannot set it here because it is a counter.
+ deleted_display_math = isDeleted(i);
+ }
+ }
+
Change const & change = runparams.inDeletedInset
? runparams.changeOfDeletedInset : lookupChange(i);
}
basefont = getLayoutFont(bparams, outerfont);
running_font = basefont;
-
column += Changes::latexMarkChange(os, bparams, runningChange,
change, runparams);
runningChange = change;
++column;
// Fully instantiated font
- Font const font = getFont(bparams, i, outerfont);
+ Font const current_font = getFont(bparams, i, outerfont);
Font const last_font = running_font;
// Do we need to close the previous font?
if (open_font &&
- (font != running_font ||
- font.language() != running_font.language()))
+ (current_font != running_font ||
+ current_font.language() != running_font.language()))
{
column += running_font.latexWriteEndChanges(
os, bparams, runparams, basefont,
- (i == body_pos-1) ? basefont : font);
+ (i == body_pos-1) ? basefont : current_font);
running_font = basefont;
open_font = false;
}
string const lang_end_command = runparams.use_polyglossia ?
"\\end{$$lang}" : lyxrc.language_command_end;
if (!running_lang.empty() &&
- font.language()->encoding()->package() == Encoding::CJK) {
+ current_font.language()->encoding()->package() == Encoding::CJK) {
string end_tag = subst(lang_end_command,
"$$lang",
running_lang);
os << from_ascii(end_tag);
column += end_tag.length();
+ if (runparams.use_polyglossia)
+ popPolyglossiaLang();
}
// Switch file encoding if necessary (and allowed)
if (!runparams.pass_thru && !style.pass_thru &&
runparams.encoding->package() != Encoding::none &&
- font.language()->encoding()->package() != Encoding::none) {
+ current_font.language()->encoding()->package() != Encoding::none) {
pair<bool, int> const enc_switch =
switchEncoding(os.os(), bparams, runparams,
- *(font.language()->encoding()));
+ *(current_font.language()->encoding()));
if (enc_switch.first) {
column += enc_switch.second;
- runparams.encoding = font.language()->encoding();
+ runparams.encoding = current_font.language()->encoding();
}
}
char_type const c = d->text_[i];
+ // A display math inset inside an ulem command will be output
+ // as a box of width \columnwidth, so we have to either disable
+ // indentation if the inset starts a paragraph, or start a new
+ // line to accommodate such box. This has to be done before
+ // writing any font changing commands.
+ if (runparams.inDisplayMath && !deleted_display_math
+ && runparams.inulemcmd) {
+ if (os.afterParbreak())
+ os << "\\noindent";
+ else
+ os << "\\\\\n";
+ }
+
// Do we need to change font?
- if ((font != running_font ||
- font.language() != running_font.language()) &&
+ if ((current_font != running_font ||
+ current_font.language() != running_font.language()) &&
i != body_pos - 1)
{
odocstringstream ods;
- column += font.latexWriteStartChanges(ods, bparams,
+ column += current_font.latexWriteStartChanges(ods, bparams,
runparams, basefont,
last_font);
- running_font = font;
+ // Check again for display math in ulem commands as a
+ // font change may also occur just before a math inset.
+ if (runparams.inDisplayMath && !deleted_display_math
+ && runparams.inulemcmd) {
+ if (os.afterParbreak())
+ os << "\\noindent";
+ else
+ os << "\\\\\n";
+ }
+ running_font = current_font;
open_font = true;
docstring fontchange = ods.str();
// check whether the fontchange ends with a \\textcolor
// style.pass_thru is false.
if (i != body_pos - 1) {
if (d->simpleTeXBlanks(runparams, os,
- i, column, font, style)) {
+ i, column, current_font, style)) {
// A surrogate pair was output. We
// must not call latexSpecialChar
// in this iteration, since it would output
OutputParams rp = runparams;
rp.free_spacing = style.free_spacing;
- rp.local_font = &font;
+ rp.local_font = ¤t_font;
rp.intitle = style.intitle;
// Two major modes: LaTeX or plain
basefont, outerfont, open_font,
runningChange, style, i, column);
}
- } else {
- if (i >= start_pos && (end_pos == -1 || i < end_pos)) {
- try {
- d->latexSpecialChar(os, bparams, rp, running_font, runningChange,
- style, i, end_pos, column);
- } catch (EncodingException & e) {
+ } else if (i >= start_pos && (end_pos == -1 || i < end_pos)) {
+ try {
+ d->latexSpecialChar(os, bparams, rp,
+ running_font, runningChange,
+ style, i, end_pos, column);
+ } catch (EncodingException & e) {
if (runparams.dryrun) {
os << "<" << _("LyX Warning: ")
<< _("uncodable character") << " '";
}
}
}
- }
// Set the encoding to that returned from latexSpecialChar (see
// comment for encoding member in OutputParams.h)
runparams.encoding = rp.encoding;
+
+ // Also carry on the info on a closed ulem command for insets
+ // such as Note that do not produce any output, so that no
+ // command is ever executed but its opening was recorded.
+ runparams.inulemcmd = rp.inulemcmd;
}
// If we have an open font definition, we have to close it
if (allowcust && d->endTeXParParams(bparams, os, runparams)
&& runparams.encoding != prev_encoding) {
runparams.encoding = prev_encoding;
- if (!runparams.isFullUnicode())
- os << setEncoding(prev_encoding->iconvName());
+ os << setEncoding(prev_encoding->iconvName());
}
LYXERR(Debug::LATEX, "Paragraph::latex... done " << this);
XHTMLStream & xs,
OutputParams const & runparams,
Font const & outerfont,
+ bool start_paragraph, bool close_paragraph,
pos_type initial) const
{
docstring retval;
Layout const & style = *d->layout_;
- xs.startParagraph(allowEmpty());
+ if (start_paragraph)
+ xs.startDivision(allowEmpty());
FontInfo font_old =
style.labeltype == LABEL_MANUAL ? style.labelfont : style.font;
if (!runparams.for_toc || inset->isInToc()) {
OutputParams np = runparams;
np.local_font = &font;
- if (!inset->getLayout().htmlisblock())
+ // If the paragraph has size 1, then we are in the "special
+ // case" where we do not output the containing paragraph info
+ if (!inset->getLayout().htmlisblock() && size() != 1)
np.html_in_par = true;
retval += inset->xhtml(xs, np);
}
font_old = font.fontInfo();
}
+ // FIXME XHTML
+ // I'm worried about what happens if a branch, say, is itself
+ // wrapped in some font stuff. I think that will not work.
xs.closeFontTags();
- xs.endParagraph();
+ if (close_paragraph)
+ xs.endDivision();
+
return retval;
}
bool Paragraph::isNewline(pos_type pos) const
{
+ // U+2028 LINE SEPARATOR
+ // U+2029 PARAGRAPH SEPARATOR
+ char_type const c = d->text_[pos];
+ if (c == 0x2028 || c == 0x2029)
+ return true;
Inset const * inset = getInset(pos);
return inset && inset->lyxCode() == NEWLINE_CODE;
}
}
-bool Paragraph::isSameSpellRange(pos_type pos1, pos_type pos2) const
+FontSpan const & Paragraph::getSpellRange(pos_type pos) const
{
- return pos1 == pos2
- || d->speller_state_.getRange(pos1) == d->speller_state_.getRange(pos2);
+ return d->speller_state_.getRange(pos);
}
}
-void Paragraph::forOutliner(docstring & os, size_t maxlen) const
+void Paragraph::forOutliner(docstring & os, size_t const maxlen,
+ bool const shorten, bool const label) const
{
- if (!d->params_.labelString().empty())
- os += d->params_.labelString() + ' ';
- for (pos_type i = 0; i < size() && os.length() < maxlen; ++i) {
+ size_t tmplen = shorten ? maxlen + 1 : maxlen;
+ if (label && !labelString().empty())
+ os += labelString() + ' ';
+ for (pos_type i = 0; i < size() && os.length() < tmplen; ++i) {
if (isDeleted(i))
continue;
char_type const c = d->text_[i];
if (isPrintable(c))
os += c;
- else if (c == '\t' || c == '\n')
- os += ' ';
else if (c == META_INSET)
- getInset(i)->forOutliner(os, maxlen);
+ getInset(i)->forOutliner(os, tmplen, false);
}
+ if (shorten)
+ Text::shortenForOutliner(os, maxlen);
}
}
+void Paragraph::resetBuffer()
+{
+ d->insetlist_.resetBuffer();
+}
+
+
Inset * Paragraph::releaseInset(pos_type pos)
{
Inset * inset = d->insetlist_.release(pos);
Private::LangWordsMap::const_iterator itl = d->words_.begin();
Private::LangWordsMap::const_iterator ite = d->words_.end();
for (; itl != ite; ++itl) {
- WordList * wl = theWordList(itl->first);
+ WordList & wl = theWordList(itl->first);
Private::Words::const_iterator it = (itl->second).begin();
Private::Words::const_iterator et = (itl->second).end();
for (; it != et; ++it)
- wl->remove(*it);
+ wl.remove(*it);
}
d->words_.clear();
}
Private::LangWordsMap::const_iterator itl = d->words_.begin();
Private::LangWordsMap::const_iterator ite = d->words_.end();
for (; itl != ite; ++itl) {
- WordList * wl = theWordList(itl->first);
+ WordList & wl = theWordList(itl->first);
Private::Words::const_iterator it = (itl->second).begin();
Private::Words::const_iterator et = (itl->second).end();
for (; it != et; ++it)
- wl->insert(*it);
+ wl.insert(*it);
}
}