lyx_mirror/src/buffer_funcs.C
Abdelrazak Younes bf61ecc2de Support code for advanced Toc management and bug correction.
* buffer.h:
  - structureChanged(): new boost signal. I intend to use that also when a text is changed in a section or caption item.

* buffer_funcs.C:
  - updateLabels(): emit Buffer::structureChanged() after TocBackend is updated.
  - checkBufferStructure(): new method for updating the TocBackend if needed.

* LyXView:
  - updateToc(): new slot for Buffer::structureChanged() signal.

* text.C:
  - call checkBufferStructure() whenever text is added or erased.

* TocBackend.[Ch]:
  - updateItem(): new method to update a specific item (called from checkBufferStructure()).



git-svn-id: svn://svn.lyx.org/lyx/lyx-devel/trunk@17413 a592a061-630c-0410-9148-cb99ea01b6c8
2007-03-12 11:23:41 +00:00

713 lines
18 KiB
C

/**
* \file buffer_funcs.C
* This file is part of LyX, the document processor.
* Licence details can be found in the file COPYING.
*
* \author Lars Gullik Bjønnes
* \author Alfredo Braunstein
*
* Full author contact details are available in file CREDITS.
*
*/
#include <config.h>
#include "buffer_funcs.h"
#include "buffer.h"
#include "bufferlist.h"
#include "bufferparams.h"
#include "dociterator.h"
#include "counters.h"
#include "errorlist.h"
#include "Floating.h"
#include "FloatList.h"
#include "gettext.h"
#include "language.h"
#include "LaTeX.h"
#include "lyxtextclass.h"
#include "paragraph.h"
#include "paragraph_funcs.h"
#include "ParagraphList.h"
#include "ParagraphParameters.h"
#include "pariterator.h"
#include "lyxvc.h"
#include "texrow.h"
#include "TocBackend.h"
#include "vc-backend.h"
#include "frontends/Alert.h"
#include "insets/insetbibitem.h"
#include "insets/insetcaption.h"
#include "insets/insetinclude.h"
#include "insets/insettabular.h"
#include "support/filetools.h"
#include "support/fs_extras.h"
#include "support/lyxlib.h"
#include <boost/bind.hpp>
#include <boost/filesystem/operations.hpp>
namespace lyx {
using namespace std;
using support::bformat;
using support::FileName;
using support::libFileSearch;
using support::makeAbsPath;
using support::makeDisplayPath;
using support::onlyFilename;
using support::onlyPath;
using support::unlink;
using std::min;
using std::string;
namespace Alert = frontend::Alert;
namespace fs = boost::filesystem;
namespace {
bool readFile(Buffer * const b, FileName const & s)
{
BOOST_ASSERT(b);
// File information about normal file
if (!fs::exists(s.toFilesystemEncoding())) {
docstring const file = makeDisplayPath(s.absFilename(), 50);
docstring text = bformat(_("The specified document\n%1$s"
"\ncould not be read."), file);
Alert::error(_("Could not read document"), text);
return false;
}
// Check if emergency save file exists and is newer.
FileName const e(s.absFilename() + ".emergency");
if (fs::exists(e.toFilesystemEncoding()) &&
fs::exists(s.toFilesystemEncoding()) &&
fs::last_write_time(e.toFilesystemEncoding()) > fs::last_write_time(s.toFilesystemEncoding()))
{
docstring const file = makeDisplayPath(s.absFilename(), 20);
docstring const text =
bformat(_("An emergency save of the document "
"%1$s exists.\n\n"
"Recover emergency save?"), file);
switch (Alert::prompt(_("Load emergency save?"), text, 0, 2,
_("&Recover"), _("&Load Original"),
_("&Cancel")))
{
case 0:
// the file is not saved if we load the emergency file.
b->markDirty();
return b->readFile(e);
case 1:
break;
default:
return false;
}
}
// Now check if autosave file is newer.
FileName const a(onlyPath(s.absFilename()) + '#' + onlyFilename(s.absFilename()) + '#');
if (fs::exists(a.toFilesystemEncoding()) &&
fs::exists(s.toFilesystemEncoding()) &&
fs::last_write_time(a.toFilesystemEncoding()) > fs::last_write_time(s.toFilesystemEncoding()))
{
docstring const file = makeDisplayPath(s.absFilename(), 20);
docstring const text =
bformat(_("The backup of the document "
"%1$s is newer.\n\nLoad the "
"backup instead?"), file);
switch (Alert::prompt(_("Load backup?"), text, 0, 2,
_("&Load backup"), _("Load &original"),
_("&Cancel") ))
{
case 0:
// the file is not saved if we load the autosave file.
b->markDirty();
return b->readFile(a);
case 1:
// Here we delete the autosave
unlink(a);
break;
default:
return false;
}
}
return b->readFile(s);
}
} // namespace anon
bool loadLyXFile(Buffer * b, FileName const & s)
{
BOOST_ASSERT(b);
if (fs::is_readable(s.toFilesystemEncoding())) {
if (readFile(b, s)) {
b->lyxvc().file_found_hook(s);
if (!fs::is_writable(s.toFilesystemEncoding()))
b->setReadonly(true);
return true;
}
} else {
docstring const file = makeDisplayPath(s.absFilename(), 20);
// Here we probably should run
if (LyXVC::file_not_found_hook(s)) {
docstring const text =
bformat(_("Do you want to retrieve the document"
" %1$s from version control?"), file);
int const ret = Alert::prompt(_("Retrieve from version control?"),
text, 0, 1, _("&Retrieve"), _("&Cancel"));
if (ret == 0) {
// How can we know _how_ to do the checkout?
// With the current VC support it has to be,
// a RCS file since CVS do not have special ,v files.
RCS::retrieve(s);
return loadLyXFile(b, s);
}
}
}
return false;
}
// FIXME newFile() should probably be a member method of Application...
Buffer * newFile(string const & filename, string const & templatename,
bool const isNamed)
{
// get a free buffer
Buffer * b = theBufferList().newBuffer(filename);
BOOST_ASSERT(b);
FileName tname;
// use defaults.lyx as a default template if it exists.
if (templatename.empty())
tname = libFileSearch("templates", "defaults.lyx");
else
tname = makeAbsPath(templatename);
if (!tname.empty()) {
if (!b->readFile(tname)) {
docstring const file = makeDisplayPath(tname.absFilename(), 50);
docstring const text = bformat(
_("The specified document template\n%1$s\ncould not be read."),
file);
Alert::error(_("Could not read template"), text);
theBufferList().release(b);
return 0;
}
}
if (!isNamed) {
b->setUnnamed();
b->setFileName(filename);
}
b->setReadonly(false);
b->fully_loaded(true);
b->updateDocLang(b->params().language);
return b;
}
void bufferErrors(Buffer const & buf, TeXErrors const & terr,
ErrorList & errorList)
{
TeXErrors::Errors::const_iterator cit = terr.begin();
TeXErrors::Errors::const_iterator end = terr.end();
for (; cit != end; ++cit) {
int id_start = -1;
int pos_start = -1;
int errorrow = cit->error_in_line;
bool found = buf.texrow().getIdFromRow(errorrow, id_start,
pos_start);
int id_end = -1;
int pos_end = -1;
do {
++errorrow;
found = buf.texrow().getIdFromRow(errorrow, id_end,
pos_end);
} while (found && id_start == id_end && pos_start == pos_end);
errorList.push_back(ErrorItem(cit->error_desc,
cit->error_text, id_start, pos_start, pos_end));
}
}
string const bufferFormat(Buffer const & buffer)
{
if (buffer.isDocBook())
return "docbook";
else if (buffer.isLiterate())
return "literate";
else
return "latex";
}
int countWords(DocIterator const & from, DocIterator const & to)
{
int count = 0;
bool inword = false;
for (DocIterator dit = from ; dit != to ; dit.forwardPos()) {
// Copied and adapted from isLetter() in ControlSpellChecker
if (dit.inTexted()
&& dit.pos() != dit.lastpos()
&& dit.paragraph().isLetter(dit.pos())
&& !dit.paragraph().isDeleted(dit.pos())) {
if (!inword) {
++count;
inword = true;
}
} else if (inword)
inword = false;
}
return count;
}
namespace {
depth_type getDepth(DocIterator const & it)
{
depth_type depth = 0;
for (size_t i = 0 ; i < it.depth() ; ++i)
if (!it[i].inset().inMathed())
depth += it[i].paragraph().getDepth() + 1;
// remove 1 since the outer inset does not count
return depth - 1;
}
depth_type getItemDepth(ParIterator const & it)
{
Paragraph const & par = *it;
LYX_LABEL_TYPES const labeltype = par.layout()->labeltype;
if (labeltype != LABEL_ENUMERATE && labeltype != LABEL_ITEMIZE)
return 0;
// this will hold the lowest depth encountered up to now.
depth_type min_depth = getDepth(it);
ParIterator prev_it = it;
while (true) {
if (prev_it.pit())
--prev_it.top().pit();
else {
// start of nested inset: go to outer par
prev_it.pop_back();
if (prev_it.empty()) {
// start of document: nothing to do
return 0;
}
}
// We search for the first paragraph with same label
// that is not more deeply nested.
Paragraph & prev_par = *prev_it;
depth_type const prev_depth = getDepth(prev_it);
if (labeltype == prev_par.layout()->labeltype) {
if (prev_depth < min_depth) {
return prev_par.itemdepth + 1;
}
else if (prev_depth == min_depth) {
return prev_par.itemdepth;
}
}
min_depth = std::min(min_depth, prev_depth);
// small optimization: if we are at depth 0, we won't
// find anything else
if (prev_depth == 0) {
return 0;
}
}
}
bool needEnumCounterReset(ParIterator const & it)
{
Paragraph const & par = *it;
BOOST_ASSERT(par.layout()->labeltype == LABEL_ENUMERATE);
depth_type const cur_depth = par.getDepth();
ParIterator prev_it = it;
while (prev_it.pit()) {
--prev_it.top().pit();
Paragraph const & prev_par = *prev_it;
if (prev_par.getDepth() <= cur_depth)
return prev_par.layout()->labeltype != LABEL_ENUMERATE;
}
// start of nested inset: reset
return true;
}
void setCaptionLabels(InsetBase & inset, string const & type,
docstring const label, Counters & counters)
{
LyXText * text = inset.getText(0);
if (!text)
return;
ParagraphList & pars = text->paragraphs();
if (pars.empty())
return;
docstring const counter = from_ascii(type);
ParagraphList::iterator p = pars.begin();
for (; p != pars.end(); ++p) {
InsetList::iterator it2 = p->insetlist.begin();
InsetList::iterator end2 = p->insetlist.end();
// Any caption within this float should have the same
// label prefix but different numbers.
for (; it2 != end2; ++it2) {
InsetBase & icap = *it2->inset;
// Look deeper just in case.
setCaptionLabels(icap, type, label, counters);
if (icap.lyxCode() == InsetBase::CAPTION_CODE) {
// We found a caption!
counters.step(counter);
int number = counters.value(counter);
InsetCaption & ic = static_cast<InsetCaption &>(icap);
ic.setType(type);
ic.setCount(number);
ic.setCustomLabel(label);
}
}
}
}
void setCaptions(Paragraph & par, LyXTextClass const & textclass)
{
if (par.insetlist.empty())
return;
Counters & counters = textclass.counters();
InsetList::iterator it = par.insetlist.begin();
InsetList::iterator end = par.insetlist.end();
for (; it != end; ++it) {
InsetBase & inset = *it->inset;
if (inset.lyxCode() == InsetBase::FLOAT_CODE
|| inset.lyxCode() == InsetBase::WRAP_CODE) {
docstring const & name = inset.getInsetName();
if (name.empty())
continue;
Floating const & fl = textclass.floats().getType(to_ascii(name));
// FIXME UNICODE
string const & type = fl.type();
docstring const label = from_utf8(fl.name());
setCaptionLabels(inset, type, label, counters);
}
else if (inset.lyxCode() == InsetBase::TABULAR_CODE
&& static_cast<InsetTabular &>(inset).tabular.isLongTabular()) {
// FIXME: are "table" and "Table" the correct type and label?
setCaptionLabels(inset, "table", from_ascii("Table"), counters);
}
}
}
// set the label of a paragraph. This includes the counters.
void setLabel(Buffer const & buf, ParIterator & it, LyXTextClass const & textclass)
{
Paragraph & par = *it;
LyXLayout_ptr const & layout = par.layout();
Counters & counters = textclass.counters();
if (it.pit() == 0) {
par.params().appendix(par.params().startOfAppendix());
} else {
par.params().appendix(it.plist()[it.pit() - 1].params().appendix());
if (!par.params().appendix() &&
par.params().startOfAppendix()) {
par.params().appendix(true);
textclass.counters().reset();
}
}
// Compute the item depth of the paragraph
par.itemdepth = getItemDepth(it);
if (layout->margintype == MARGIN_MANUAL) {
if (par.params().labelWidthString().empty())
par.params().labelWidthString(par.translateIfPossible(layout->labelstring(), buf.params()));
} else {
par.params().labelWidthString(docstring());
}
// Optimisation: setLabel() can be called for each for each
// paragraph of the document. So we make the string static to
// avoid the repeated instanciation.
static docstring itemlabel;
// is it a layout that has an automatic label?
if (layout->labeltype == LABEL_COUNTER) {
if (layout->toclevel <= buf.params().secnumdepth
&& (layout->latextype != LATEX_ENVIRONMENT
|| isFirstInSequence(it.pit(), it.plist()))) {
counters.step(layout->counter);
par.params().labelString(
par.expandLabel(layout, buf.params()));
} else
par.params().labelString(docstring());
} else if (layout->labeltype == LABEL_ITEMIZE) {
// At some point of time we should do something more
// clever here, like:
// par.params().labelString(
// buf.params().user_defined_bullet(par.itemdepth).getText());
// for now, use a simple hardcoded label
switch (par.itemdepth) {
case 0:
itemlabel = char_type(0x2022);
break;
case 1:
itemlabel = char_type(0x2013);
break;
case 2:
itemlabel = char_type(0x2217);
break;
case 3:
itemlabel = char_type(0x2219); // or 0x00b7
break;
}
par.params().labelString(itemlabel);
} else if (layout->labeltype == LABEL_ENUMERATE) {
// FIXME
// Yes I know this is a really, really! bad solution
// (Lgb)
docstring enumcounter = from_ascii("enum");
switch (par.itemdepth) {
case 2:
enumcounter += 'i';
case 1:
enumcounter += 'i';
case 0:
enumcounter += 'i';
break;
case 3:
enumcounter += "iv";
break;
default:
// not a valid enumdepth...
break;
}
// Maybe we have to reset the enumeration counter.
if (needEnumCounterReset(it))
counters.reset(enumcounter);
counters.step(enumcounter);
string format;
switch (par.itemdepth) {
case 0:
format = N_("\\arabic{enumi}.");
break;
case 1:
format = N_("(\\alph{enumii})");
break;
case 2:
format = N_("\\roman{enumiii}.");
break;
case 3:
format = N_("\\Alph{enumiv}.");
break;
default:
// not a valid enumdepth...
break;
}
par.params().labelString(counters.counterLabel(
par.translateIfPossible(from_ascii(format), buf.params())));
} else if (layout->labeltype == LABEL_BIBLIO) {// ale970302
counters.step(from_ascii("bibitem"));
int number = counters.value(from_ascii("bibitem"));
if (par.bibitem())
par.bibitem()->setCounter(number);
par.params().labelString(
par.translateIfPossible(layout->labelstring(), buf.params()));
// In biblio shouldn't be following counters but...
} else if (layout->labeltype == LABEL_SENSITIVE) {
// Search for the first float or wrap inset in the iterator
size_t i = it.depth();
InsetBase * in = 0;
while (i > 0) {
--i;
InsetBase::Code const code = it[i].inset().lyxCode();
if (code == InsetBase::FLOAT_CODE ||
code == InsetBase::WRAP_CODE) {
in = &it[i].inset();
break;
}
}
// FIXME Can getInsetName() return an empty name for wide or
// float insets? If not we can put the definition of type
// inside the if (in) clause and use that instead of
// if (!type.empty()).
docstring type;
if (in)
type = in->getInsetName();
if (!type.empty()) {
Floating const & fl = textclass.floats().getType(to_ascii(type));
// FIXME UNICODE
counters.step(from_ascii(fl.type()));
// Doesn't work... yet.
par.params().labelString(par.translateIfPossible(
bformat(from_ascii("%1$s #:"), from_utf8(fl.name())),
buf.params()));
} else {
// par->SetLayout(0);
par.params().labelString(par.translateIfPossible(
layout->labelstring(), buf.params()));
}
} else if (layout->labeltype == LABEL_NO_LABEL)
par.params().labelString(docstring());
else
par.params().labelString(
par.translateIfPossible(layout->labelstring(), buf.params()));
}
} // anon namespace
bool updateCurrentLabel(Buffer const & buf,
ParIterator & it)
{
if (it == par_iterator_end(buf.inset()))
return false;
// if (it.lastpit == 0 && LyXText::isMainText(buf))
// return false;
switch (it->layout()->labeltype) {
case LABEL_NO_LABEL:
case LABEL_MANUAL:
case LABEL_BIBLIO:
case LABEL_TOP_ENVIRONMENT:
case LABEL_CENTERED_TOP_ENVIRONMENT:
case LABEL_STATIC:
case LABEL_ITEMIZE:
setLabel(buf, it, buf.params().getLyXTextClass());
return true;
case LABEL_SENSITIVE:
case LABEL_COUNTER:
// do more things with enumerate later
case LABEL_ENUMERATE:
return false;
}
// This is dead code which get rid of a warning:
return true;
}
void updateLabels(Buffer const & buf,
ParIterator & from, ParIterator & to, bool childonly)
{
for (ParIterator it = from; it != to; ++it) {
if (it.pit() > it.lastpit())
return;
if (!updateCurrentLabel (buf, it)) {
updateLabels(buf, childonly);
return;
}
}
}
void updateLabels(Buffer const & buf, ParIterator & iter, bool childonly)
{
if (updateCurrentLabel(buf, iter))
return;
updateLabels(buf, childonly);
}
void updateLabels(Buffer const & buf, bool childonly)
{
// Use the master text class also for child documents
LyXTextClass const & textclass = buf.params().getLyXTextClass();
if (!childonly) {
// If this is a child document start with the master
Buffer const * const master = buf.getMasterBuffer();
if (master != &buf) {
updateLabels(*master);
return;
}
// start over the counters
textclass.counters().reset();
}
ParIterator const end = par_iterator_end(buf.inset());
for (ParIterator it = par_iterator_begin(buf.inset()); it != end; ++it) {
// reduce depth if necessary
if (it.pit()) {
Paragraph const & prevpar = it.plist()[it.pit() - 1];
it->params().depth(min(it->params().depth(),
prevpar.getMaxDepthAfter()));
} else
it->params().depth(0);
// set the counter for this paragraph
setLabel(buf, it, textclass);
// It is better to set the captions after setLabel because
// the caption number might need the section number in the
// future.
setCaptions(*it, textclass);
// Now included docs
InsetList::const_iterator iit = it->insetlist.begin();
InsetList::const_iterator end = it->insetlist.end();
for (; iit != end; ++iit) {
if (iit->inset->lyxCode() == InsetBase::INCLUDE_CODE)
static_cast<InsetInclude const *>(iit->inset)
->updateLabels(buf);
}
}
Buffer & cbuf = const_cast<Buffer &>(buf);
cbuf.tocBackend().update();
cbuf.structureChanged();
}
void checkBufferStructure(Buffer & buffer, ParIterator const & par_it)
{
if (par_it->layout()->labeltype == LABEL_COUNTER) {
buffer.tocBackend().updateItem(par_it);
buffer.structureChanged();
}
}
} // namespace lyx