2008-02-25 01:56:53 +00:00
|
|
|
/**
|
2008-02-25 02:01:48 +00:00
|
|
|
* \file WordList.cpp
|
2008-02-25 01:56:53 +00:00
|
|
|
* This file is part of LyX, the document processor.
|
|
|
|
* Licence details can be found in the file COPYING.
|
|
|
|
*
|
|
|
|
* \author Stefan Schimanski
|
|
|
|
*
|
|
|
|
* Full author contact details are available in file CREDITS.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <config.h>
|
|
|
|
|
2008-02-25 09:02:35 +00:00
|
|
|
#include "WordList.h"
|
2008-02-25 01:56:53 +00:00
|
|
|
|
2008-02-25 09:02:35 +00:00
|
|
|
#include "support/convert.h"
|
|
|
|
#include "support/debug.h"
|
|
|
|
#include "support/docstring.h"
|
2008-04-30 08:26:40 +00:00
|
|
|
#include "support/lassert.h"
|
2016-12-31 14:16:15 +00:00
|
|
|
#include "support/unique_ptr.h"
|
2014-03-21 11:24:47 +00:00
|
|
|
#include "support/weighted_btree.h"
|
2008-02-25 01:56:53 +00:00
|
|
|
|
2014-07-04 20:19:43 +00:00
|
|
|
#include <QThreadStorage>
|
|
|
|
|
2009-12-18 14:48:56 +00:00
|
|
|
#include <map>
|
|
|
|
|
|
|
|
using namespace std;
|
|
|
|
|
2008-02-25 01:56:53 +00:00
|
|
|
namespace lyx {
|
|
|
|
|
2008-02-26 13:07:59 +00:00
|
|
|
///
|
2016-12-31 14:16:15 +00:00
|
|
|
typedef map<string, unique_ptr<WordList>> GlobalWordList;
|
2014-07-04 20:19:43 +00:00
|
|
|
// Each thread uses its own word list, but only the one of the GUI thread is
|
|
|
|
// used to do real work. The others are only neded to prevent simultanous
|
|
|
|
// write access e.g. from a cloned buffer and a true document buffer.
|
|
|
|
QThreadStorage<GlobalWordList *> theGlobalWordList;
|
2008-02-26 13:07:59 +00:00
|
|
|
|
2011-10-15 13:38:52 +00:00
|
|
|
|
2016-12-31 14:22:07 +00:00
|
|
|
WordList & theWordList(string const & lang)
|
2008-02-26 13:07:59 +00:00
|
|
|
{
|
2014-07-04 20:19:43 +00:00
|
|
|
if (!theGlobalWordList.hasLocalData())
|
|
|
|
theGlobalWordList.setLocalData(new GlobalWordList);
|
2016-12-31 14:16:15 +00:00
|
|
|
GlobalWordList & globalWordList = *theGlobalWordList.localData();
|
|
|
|
GlobalWordList::iterator it = globalWordList.find(lang);
|
|
|
|
if (it != globalWordList.end())
|
2016-12-31 14:22:07 +00:00
|
|
|
return *it->second;
|
2016-12-31 14:16:15 +00:00
|
|
|
else
|
|
|
|
return *(globalWordList[lang] = make_unique<WordList>());
|
2011-10-15 13:38:52 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
2008-02-25 01:56:53 +00:00
|
|
|
///
|
|
|
|
struct WordList::Impl {
|
|
|
|
///
|
|
|
|
size_t c_;
|
|
|
|
///
|
2008-02-26 13:07:59 +00:00
|
|
|
typedef stx::weighted_btree<docstring, size_t, int> Words;
|
2008-02-25 01:56:53 +00:00
|
|
|
///
|
|
|
|
Words words_;
|
|
|
|
};
|
|
|
|
|
|
|
|
|
2016-12-31 14:16:15 +00:00
|
|
|
WordList::WordList() : d(make_unique<Impl>())
|
2008-02-25 01:56:53 +00:00
|
|
|
{
|
|
|
|
d->c_ = 0;
|
|
|
|
|
|
|
|
#if 0
|
|
|
|
for (size_t i = 1000000; i > 0; --i) {
|
|
|
|
d->words_.insert("a" + convert<docstring>(i), size_t(1), stx::Void());
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
docstring const & WordList::word(size_t idx) const
|
|
|
|
{
|
2008-02-25 09:02:35 +00:00
|
|
|
Impl::Words::const_iterator it = d->words_.find_summed_weight(idx);
|
2013-04-25 21:27:10 +00:00
|
|
|
LASSERT(it != d->words_.end(), { static docstring dummy; return dummy; });
|
2017-07-03 17:53:14 +00:00
|
|
|
|
2008-02-28 14:46:47 +00:00
|
|
|
// We use the key() method here, and not something like it->first
|
|
|
|
// because the btree only returns (iterator-) temporary value pairs.
|
|
|
|
// If we returned the first component of those here, we get an
|
|
|
|
// invalid reference and therefore strange crashes.
|
|
|
|
return it.key();
|
2008-02-25 01:56:53 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
size_t WordList::size() const
|
|
|
|
{
|
2008-02-26 13:07:59 +00:00
|
|
|
return d->words_.summed_weight();
|
2008-02-25 01:56:53 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void WordList::insert(docstring const & w)
|
|
|
|
{
|
2008-02-26 13:07:59 +00:00
|
|
|
Impl::Words::iterator it = d->words_.find(w);
|
|
|
|
if (it == d->words_.end())
|
|
|
|
d->words_.insert(w, size_t(1), 1);
|
|
|
|
else {
|
|
|
|
it.data()++;
|
|
|
|
d->words_.change_weight(it, 1);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void WordList::remove(docstring const & w)
|
|
|
|
{
|
|
|
|
Impl::Words::iterator it = d->words_.find(w);
|
|
|
|
if (it != d->words_.end()) {
|
|
|
|
it.data()--;
|
|
|
|
d->words_.change_weight(it, 0);
|
|
|
|
// We will not erase here, but instead we just leave it
|
|
|
|
// in the btree with weight 0. This avoid too much
|
|
|
|
// reorganisation of the tree all the time.
|
|
|
|
//if (it.data() == 0)
|
|
|
|
// d->words_.erase(w);
|
|
|
|
}
|
2008-02-25 01:56:53 +00:00
|
|
|
}
|
|
|
|
|
2008-02-25 08:54:51 +00:00
|
|
|
} // namespace lyx
|