* Licence details can be found in the file COPYING.
*
* \author Asger Alstrup
- * \author Lars Gullik Bjønnes
+ * \author Lars Gullik Bjønnes
* \author Jean-Marc Lasgouttes
* \author Angus Leeming
* \author John Levon
- * \author André Pönitz
+ * \author André Pönitz
* \author Dekel Tsur
- * \author Jürgen Vigna
+ * \author Jürgen Vigna
*
* Full author contact details are available in file CREDITS.
*/
#include "LyXRC.h"
#include "OutputParams.h"
#include "output_latex.h"
+#include "output_xhtml.h"
#include "paragraph_funcs.h"
#include "ParagraphParameters.h"
+#include "SpellChecker.h"
#include "sgml.h"
#include "TextClass.h"
#include "TexRow.h"
#include "Text.h"
#include "VSpace.h"
+#include "WordLangTuple.h"
#include "WordList.h"
#include "frontends/alert.h"
#include "insets/InsetBibitem.h"
#include "insets/InsetLabel.h"
-#include "support/lassert.h"
-#include "support/convert.h"
#include "support/debug.h"
+#include "support/docstring_list.h"
#include "support/ExceptionMessage.h"
#include "support/gettext.h"
+#include "support/lassert.h"
#include "support/lstrings.h"
#include "support/Messages.h"
#include "support/textutils.h"
}
+bool Paragraph::isFullyDeleted(pos_type start, pos_type end) const
+{
+ LASSERT(start >= 0 && start <= size(), /**/);
+ LASSERT(end > start && end <= size() + 1, /**/);
+
+ return d->changes_.isFullyDeleted(start, end);
+}
+
+
bool Paragraph::isChanged(pos_type start, pos_type end) const
{
LASSERT(start >= 0 && start <= size(), /**/);
}
bool close = false;
- odocstringstream ods;
+ odocstream::pos_type const len = os.tellp();
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")
- ods << "\\beginL{}";
+ os << "\\beginL{}";
else
- ods << "\\L{";
+ os << "\\L{";
close = true;
}
if (open_font && inset->noFontChange()) {
bool closeLanguage = arabtex
|| basefont.isRightToLeft() == running_font.isRightToLeft();
- unsigned int count = running_font.latexWriteEndChanges(ods,
+ unsigned int count = running_font.latexWriteEndChanges(os,
bparams, runparams, basefont, basefont, closeLanguage);
column += count;
// if any font properties were closed, update the running_font,
int tmp;
try {
- tmp = inset->latex(ods, runparams);
+ tmp = inset->latex(os, runparams);
} catch (EncodingException & e) {
// add location information and throw again.
e.par_id = id_;
if (close) {
if (running_font.language()->lang() == "farsi")
- ods << "\\endL{}";
+ os << "\\endL{}";
else
- ods << '}';
+ os << '}';
}
- os << ods.str();
-
if (tmp) {
for (int j = 0; j < tmp; ++j)
texrow.newline();
texrow.start(owner_->id(), i + 1);
column = 0;
} else {
- column += ods.str().size();
+ column += os.tellp() - len;
}
if (owner_->lookupChange(i).type == Change::DELETED)
return;
}
- if (lyxrc.fontenc == "T1" && latexSpecialT1(c, os, i, column))
+ // If T1 font encoding is used, use the special
+ // characters it provides.
+ // NOTE: some languages reset the font encoding
+ // internally
+ if (!running_font.language()->internalFontEncoding()
+ && lyxrc.fontenc == "T1" && latexSpecialT1(c, os, i, column))
return;
+ // \tt font needs special treatment
if (running_font.fontInfo().family() == TYPEWRITER_FAMILY
&& latexSpecialTypewriter(c, os, i, column))
return;
column += 17;
break;
- case '*': case '[':
+ case '*':
+ case '[':
+ case ']':
// avoid being mistaken for optional arguments
os << '{';
os.put(c);
int column = 0;
for (pos_type i = 0; i <= size(); ++i) {
- Change change = lookupChange(i);
+ Change const change = lookupChange(i);
Changes::lyxMarkChange(os, column, running_change, change);
running_change = change;
if (i == size())
break;
- // Write font changes
- Font const & font2 = getFontSettings(bparams, i);
+ // Write font changes (ignore spelling markers)
+ Font font2 = getFontSettings(bparams, i);
+ font2.setMisspelled(false);
if (font2 != font1) {
font2.lyxWriteChanges(font1, os);
column = 0;
d->text_.append(s);
// FIXME: Optimize this!
- for (pos_type i = oldsize; i != newsize; ++i) {
+ for (size_t i = oldsize; i != newsize; ++i) {
// track change
d->changes_.insert(change, i);
}
// the next two functions are for the manual labels
docstring const Paragraph::getLabelWidthString() const
{
- if (d->layout_->margintype == MARGIN_MANUAL)
+ if (d->layout_->margintype == MARGIN_MANUAL
+ || d->layout_->latextype == LATEX_BIB_ENVIRONMENT)
return d->params_.labelWidthString();
else
return _("Senseless with this layout!");
if (fmt.empty() && layout.labeltype == LABEL_COUNTER
&& !layout.counter.empty())
- fmt = "\\the" + layout.counter;
+ return tclass.counters().theCounter(layout.counter);
// handle 'inherited level parts' in 'fmt',
// i.e. the stuff between '@' in '@Section@.\arabic{subsection}'
{
int column = 0;
- switch (params_.align()) {
+ LyXAlignment const curAlign = params_.align();
+
+ if (curAlign == layout_->align)
+ return column;
+
+ switch (curAlign) {
case LYX_ALIGN_NONE:
case LYX_ALIGN_BLOCK:
case LYX_ALIGN_LAYOUT:
InsetCode code = owner_->ownerCode();
bool const lastpar = runparams.isLastPar;
- switch (params_.align()) {
+ switch (curAlign) {
case LYX_ALIGN_NONE:
case LYX_ALIGN_BLOCK:
case LYX_ALIGN_LAYOUT:
// This one spits out the text of the paragraph
bool Paragraph::latex(BufferParams const & bparams,
- Font const & outerfont,
- odocstream & os, TexRow & texrow,
- OutputParams const & runparams) const
+ Font const & outerfont,
+ odocstream & os, TexRow & texrow,
+ OutputParams const & runparams,
+ int start_pos, int end_pos) const
{
LYXERR(Debug::LATEX, "Paragraph::latex... " << this);
unsigned int column = 0;
if (body_pos > 0) {
- // the optional argument is kept in curly brackets in
- // case it contains a ']'
- os << "[{";
- column += 2;
+ os << '[';
+ column += 1;
basefont = getLabelFont(bparams, outerfont);
} else {
basefont = getLayoutFont(bparams, outerfont);
runningChange, Change(Change::UNCHANGED));
runningChange = Change(Change::UNCHANGED);
- os << "}] ";
- column +=3;
+ os << "] ";
+ column +=2;
}
if (style.isCommand()) {
os << '{';
if (!runparams.verbatim &&
runparams.encoding->package() != Encoding::none &&
font.language()->encoding()->package() != Encoding::none) {
- odocstringstream ods;
- pair<bool, int> const enc_switch = switchEncoding(ods, bparams,
+ pair<bool, int> const enc_switch = switchEncoding(os, bparams,
runparams, *(font.language()->encoding()));
- os << ods.str();
if (enc_switch.first) {
column += enc_switch.second;
runparams.encoding = font.language()->encoding();
os << fontchange;
}
- if (c == ' ') {
+ // FIXME: think about end_pos implementation...
+ if (c == ' ' && i >= start_pos && (end_pos == -1 || i < end_pos)) {
// FIXME: integrate this case in latexSpecialChar
// Do not print the separation of the optional argument
// if style.pass_thru is false. This works because
// Two major modes: LaTeX or plain
// Handle here those cases common to both modes
// and then split to handle the two modes separately.
- if (c == META_INSET)
- d->latexInset(bparams, os,
- texrow, rp, running_font,
- basefont, outerfont, open_font,
- runningChange, style, i, column);
- else {
- try {
- d->latexSpecialChar(os, rp, running_font, runningChange,
- style, i, column);
- } catch (EncodingException & e) {
+ if (c == META_INSET) {
+ if (i >= start_pos && (end_pos == -1 || i < end_pos)) {
+ d->latexInset(bparams, os,
+ texrow, rp, running_font,
+ basefont, outerfont, open_font,
+ runningChange, style, i, column);
+ }
+ } else {
+ if (i >= start_pos && (end_pos == -1 || i < end_pos)) {
+ try {
+ d->latexSpecialChar(os, rp, running_font, runningChange,
+ style, i, 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)
// Needed if there is an optional argument but no contents.
if (body_pos > 0 && body_pos == size()) {
- os << "}]~";
+ os << "]~";
return_value = false;
}
}
-pos_type Paragraph::firstWord(odocstream & os, OutputParams const & runparams)
+pos_type Paragraph::firstWordDocBook(odocstream & os, OutputParams const & runparams)
const
{
pos_type i;
}
+pos_type Paragraph::firstWordLyXHTML(odocstream & os, OutputParams const & runparams)
+ const
+{
+ pos_type i;
+ for (i = 0; i < size(); ++i) {
+ if (Inset const * inset = getInset(i)) {
+ inset->xhtml(os, runparams);
+ } else {
+ char_type c = d->text_[i];
+ if (c == ' ')
+ break;
+ os << html::escapeChar(c);
+ }
+ }
+ return i;
+}
+
+
bool Paragraph::Private::onlyText(Buffer const & buf, Font const & outerfont, pos_type initial) const
{
Font font_old;
}
+docstring Paragraph::simpleLyXHTMLOnePar(Buffer const & buf,
+ odocstream & os,
+ OutputParams const & runparams,
+ Font const & outerfont,
+ pos_type initial) const
+{
+ docstring retval;
+
+ // FIXME We really need to manage the tag nesting here.
+ // Probably in the same sort of way as in output_xhtml.
+ bool emph_flag = false;
+ bool bold_flag = false;
+ std::string closing_tag;
+
+ Layout const & style = *d->layout_;
+ FontInfo font_old =
+ style.labeltype == LABEL_MANUAL ? style.labelfont : style.font;
+
+ //if (style.pass_thru && !d->onlyText(buf, outerfont, initial))
+ // os << "]]>";
+
+ // parsing main loop
+ for (pos_type i = initial; i < size(); ++i) {
+ Font font = getFont(buf.params(), i, outerfont);
+
+ // emphasis
+ if (font_old.emph() != font.fontInfo().emph()) {
+ if (font.fontInfo().emph() == FONT_ON) {
+ os << "<em>";
+ emph_flag = true;
+ } else if (emph_flag && i != initial) {
+ os << "</em>";
+ emph_flag = false;
+ }
+ }
+ // bold
+ if (font_old.series() != font.fontInfo().series()) {
+ if (font.fontInfo().series() == BOLD_SERIES) {
+ os << "<strong>";
+ bold_flag = true;
+ } else if (bold_flag && i != initial) {
+ os << "</strong>";
+ bold_flag = false;
+ }
+ }
+ // FIXME Other such tags?
+
+ if (Inset const * inset = getInset(i)) {
+ retval += inset->xhtml(os, runparams);
+ } else {
+ char_type c = d->text_[i];
+
+ if (style.pass_thru)
+ os.put(c);
+ else if (c == '-') {
+ docstring str;
+ int j = i + 1;
+ if (j < size() && d->text_[j] == '-') {
+ j += 1;
+ if (j < size() && d->text_[j] == '-') {
+ str += from_ascii("—");
+ i += 2;
+ } else {
+ str += from_ascii("–");
+ i += 1;
+ }
+ }
+ else
+ str += c;
+ os << str;
+ } else
+ os << html::escapeChar(c);
+ }
+ font_old = font.fontInfo();
+ }
+
+ // FIXME This could be out of order. See above.
+ if (emph_flag)
+ os << "</em>";
+ if (bold_flag)
+ os << "</strong>";
+
+ return retval;
+}
+
+
bool Paragraph::isHfill(pos_type pos) const
{
Inset const * inset = getInset(pos);
}
-/// Used by the spellchecker
-bool Paragraph::isLetter(pos_type pos) const
+bool Paragraph::isWordSeparator(pos_type pos) const
{
if (Inset const * inset = getInset(pos))
- return inset->isLetter();
+ return !inset->isLetter();
char_type const c = d->text_[pos];
- return isLetterChar(c) || isDigit(c);
+ // We want to pass the ' and escape chars to the spellchecker
+ static docstring const quote = from_utf8(lyxrc.spellchecker_esc_chars + '\'');
+ return (!isLetterChar(c) && !isDigit(c) && !contains(quote, c))
+ || pos == size();
}
}
+docstring Paragraph::stringify(pos_type beg, pos_type end, int options, OutputParams & runparams) const
+{
+ odocstringstream os;
+
+ if (beg == 0
+ && options & AS_STR_LABEL
+ && !d->params_.labelString().empty())
+ os << d->params_.labelString() << ' ';
+
+ for (pos_type i = beg; i < end; ++i) {
+ char_type const c = d->text_[i];
+ if (isPrintable(c) || c == '\t'
+ || (c == '\n' && options & AS_STR_NEWLINES))
+ os.put(c);
+ else if (c == META_INSET && options & AS_STR_INSETS) {
+ getInset(i)->plaintext(os, runparams);
+ }
+ }
+
+ return os.str();
+}
+
+
void Paragraph::setInsetOwner(Inset const * inset)
{
d->inset_owner_ = inset;
Inset const & Paragraph::inInset() const
{
LASSERT(d->inset_owner_, throw ExceptionMessage(BufferException,
- _("Memory problem"), _("Paragraph not properly initiliazed")));
+ _("Memory problem"), _("Paragraph not properly initialized")));
return *d->inset_owner_;
}
}
}
- if (!isLetter(pos) || isDeleted(pos)) {
+ if (isWordSeparator(pos) || isDeleted(pos)) {
// permit capitalization again
capitalize = true;
}
- if (oldChar != newChar)
+ if (oldChar != newChar) {
changes += newChar;
+ if (pos != right - 1)
+ continue;
+ // step behind the changing area
+ pos++;
+ }
- if (oldChar == newChar || pos == right - 1) {
- if (oldChar != newChar) {
- // step behind the changing area
- pos++;
- }
- int erasePos = pos - changes.size();
- for (size_t i = 0; i < changes.size(); i++) {
- insertChar(pos, changes[i],
- getFontSettings(bparams,
- erasePos),
- trackChanges);
- if (!eraseChar(erasePos, trackChanges)) {
- ++erasePos;
- ++pos; // advance
- ++right; // expand selection
- }
+ int erasePos = pos - changes.size();
+ for (size_t i = 0; i < changes.size(); i++) {
+ insertChar(pos, changes[i],
+ getFontSettings(bparams,
+ erasePos),
+ trackChanges);
+ if (!eraseChar(erasePos, trackChanges)) {
+ ++erasePos;
+ ++pos; // advance
+ ++right; // expand selection
}
- changes.clear();
}
+ changes.clear();
}
}
// if necessary, check whether string matches word
if (mw) {
- if (pos > 0 && isLetter(pos - 1))
+ if (pos > 0 && !isWordSeparator(pos - 1))
return false;
if (pos + strsize < parsize
- && isLetter(pos + strsize))
+ && !isWordSeparator(pos + strsize))
return false;
}
}
-void Paragraph::collectWords(CursorSlice const & sl)
+void Paragraph::locateWord(pos_type & from, pos_type & to,
+ word_location const loc) const
{
- // find new words
- bool inword = false;
+ switch (loc) {
+ case WHOLE_WORD_STRICT:
+ if (from == 0 || from == size()
+ || isWordSeparator(from)
+ || isWordSeparator(from - 1)) {
+ to = from;
+ return;
+ }
+ // no break here, we go to the next
- //lyxerr << "Words: ";
- pos_type n = size();
- for (pos_type pos = 0; pos != n; ++pos) {
- if (isDeleted(pos))
- continue;
+ case WHOLE_WORD:
+ // If we are already at the beginning of a word, do nothing
+ if (!from || isWordSeparator(from - 1))
+ break;
+ // no break here, we go to the next
- if (!isLetter(pos)) {
- inword = false;
- continue;
- }
+ case PREVIOUS_WORD:
+ // always move the cursor to the beginning of previous word
+ while (from && !isWordSeparator(from - 1))
+ --from;
+ break;
+ case NEXT_WORD:
+ LYXERR0("Paragraph::locateWord: NEXT_WORD not implemented yet");
+ break;
+ case PARTIAL_WORD:
+ // no need to move the 'from' cursor
+ break;
+ }
+ to = from;
+ while (to < size() && !isWordSeparator(to))
+ ++to;
+}
- if (inword)
- continue;
- inword = true;
- CursorSlice from = sl;
- CursorSlice to = sl;
- from.pos() = pos;
- to.pos() = pos;
- from.text()->getWord(from, to, WHOLE_WORD);
- if (to.pos() - from.pos() < 6)
+void Paragraph::collectWords()
+{
+ pos_type n = size();
+ WordLangTuple wl;
+ docstring_list suggestions;
+ for (pos_type pos = 0; pos < n; ++pos) {
+ if (isWordSeparator(pos))
continue;
- docstring word = asString(from.pos(), to.pos(), false);
- d->words_.insert(word);
- //lyxerr << word << " ";
+ pos_type from = pos;
+ locateWord(from, pos, WHOLE_WORD);
+ if (pos - from >= 6) {
+ docstring word = asString(from, pos, AS_STR_NONE);
+ d->words_.insert(word);
+ }
+ if (lyxrc.spellcheck_continuously
+ && spellCheck(from, pos, wl, suggestions)) {
+ for (size_t i = 0; i != suggestions.size(); ++i)
+ d->words_.insert(suggestions[i]);
+ }
}
- //lyxerr << std::endl;
}
}
-void Paragraph::updateWords(CursorSlice const & sl)
+void Paragraph::updateWords()
{
- LASSERT(&sl.paragraph() == this, /**/);
deregisterWords();
- collectWords(sl);
+ collectWords();
registerWords();
}
+
+bool Paragraph::spellCheck(pos_type & from, pos_type & to, WordLangTuple & wl,
+ docstring_list & suggestions) const
+{
+ SpellChecker * speller = theSpellChecker();
+ if (!speller)
+ return false;
+
+ locateWord(from, to, WHOLE_WORD);
+ if (from == to || from >= pos_type(d->text_.size()))
+ return false;
+
+ docstring word = asString(from, to, AS_STR_INSETS);
+ string const lang_code = lyxrc.spellchecker_alt_lang.empty()
+ ? getFontSettings(d->inset_owner_->buffer().params(), from).language()->code()
+ : lyxrc.spellchecker_alt_lang;
+ wl = WordLangTuple(word, lang_code);
+ SpellChecker::Result res = speller->check(wl);
+ // Just ignore any error that the spellchecker reports.
+ // FIXME: we should through out an exception and catch it in the GUI to
+ // display the error.
+ if (!speller->error().empty())
+ return false;
+
+ bool const misspelled = res != SpellChecker::OK
+ && res != SpellChecker::IGNORED_WORD;
+
+ if (lyxrc.spellcheck_continuously)
+ d->fontlist_.setMisspelled(from, to, misspelled);
+
+ if (misspelled) {
+ while (!(word = speller->nextMiss()).empty())
+ suggestions.push_back(word);
+ }
+ return misspelled;
+}
+
+
+bool Paragraph::isMisspelled(pos_type pos) const
+{
+ pos_type from = pos;
+ pos_type to = pos;
+ WordLangTuple wl;
+ docstring_list suggestions;
+ return spellCheck(from, to, wl, suggestions);
+}
+
+
} // namespace lyx