std::pair<int, int> NGramsDBMS::FindRawNGramsID(
const lem::FString &suffix,
const lem::FString &sgm,
const lem::UCString &w1,
const lem::UCString &w2,
const lem::UCString &w3,
const lem::UCString &w4,
const lem::UCString &w5
)
{
FString sql = lem::format_str("SELECT w, N.id"
" FROM NGRAM_WORDS%s%s wrd1, NGRAM_WORDS%s%s wrd2, NGRAM_WORDS%s%s wrd3, NGRAM_WORDS%s%s wrd4, NGRAM_WORDS%s%s wrd5, NGRAM5%s%s N"
" WHERE wrd1.word='%s' AND wrd2.word='%s' AND wrd3.word='%s' AND wrd4.word='%s' AND wrd5.word='%s' AND iword1=wrd1.id AND iword2=wrd2.id AND iword3=wrd3.id AND iword4=wrd4.id AND iword5=wrd5.id"
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, to_utf8(w1.c_str()).c_str(), to_utf8(w2.c_str()).c_str()
, to_utf8(w3.c_str()).c_str()
, to_utf8(w4.c_str()).c_str(), to_utf8(w5.c_str()).c_str());
std::pair<int, int> r = SelectIntPair(sql);
return r;
}
SG_EntryGroup::KEY SG_EntryGroup::BuildKey( const lem::UCString &str )
{
switch( str.length() )
{
case 1: return KEY( str.front(), 0, 0 );
case 2: return KEY( str.front(), str[1], 0 );
default: return KEY( str.front(), str[1], str[2] );
}
}
bool LA_PreprocessorRules::Crop(
const lem::UCString &word,
lem::MCollect<lem::UCString> &results,
lem::MCollect<lem::Real1> &rels,
LA_RecognitionTrace *trace
) const
{
bool applied = false;
if (!crop_rules.empty())
{
// сначала применяем префиксные правила
typedef CROP_RULES::const_iterator IT;
LA_CropRule::HashType prefix_hash = LA_CropRule::CalcHash(word.c_str(), true, false);
std::pair<IT, IT> pp = prefix_crop_rules.equal_range(prefix_hash);
lem::UCString result;
for (auto it = pp.first; it != pp.second; ++it)
{
const LA_CropRule *r = it->second;
if (r->Apply(word, result))
{
applied = true;
results.push_back(result);
rels.push_back(r->GetRel());
if (trace != nullptr)
{
trace->CropRuleApplied(word, result, r);
}
}
}
// теперь отсекаем аффикс
LA_CropRule::HashType affix_hash = LA_CropRule::CalcHash(word.c_str(), false, true);
pp = affix_crop_rules.equal_range(affix_hash);
for (auto it = pp.first; it != pp.second; ++it)
{
const LA_CropRule *r = it->second;
if (r->Apply(word, result))
{
applied = true;
results.push_back(result);
rels.push_back(r->GetRel());
if (trace != nullptr)
{
trace->CropRuleApplied(word, result, r);
}
}
}
}
return applied;
}
void Lemmatizator::DecodeWord1( const lem::CString & a, lem::UCString & u ) const
{
for( int i=0; i<a.length(); ++i )
u.ptr()[i] = i2wchar[ unsigned(a[i]) ];
u.ptr()[ a.length() ] = 0;
u.calc_hash();
return;
}
const lem::CString lem::to_ascii( const lem::UCString &str, const CodeConverter *cp )
{
if( str.empty() )
return CString();
lem::CString ascii;
lem_unicode_to_ascii( ascii.ptr(), str.c_str(), cp ? cp : &lem::UI::get_UI().GetSessionCp() );
ascii.ptr()[str.length()] = 0;
ascii.calc_hash();
return ascii;
}
void Lemmatizator::EncodeWord1( const lem::UCString & u, lem::CString & a ) const
{
for( int i=0; i<u.length(); ++i )
{
std::map<wchar_t,int>::const_iterator it = wchar2i.find(u[i]);
if( it==wchar2i.end() )
a.ptr()[i] = (char)0xff;
else
a.ptr()[i] = it->second;
}
a.ptr()[ u.length() ] = 0;
a.calc_hash();
return;
}
lem::UCString Lemmatizator::GetSuffix( const lem::UCString & word ) const
{
LEM_CHECKIT_Z( !word.empty() );
lem::UCString uword(word);
uword.to_lower();
if( word.length()==7 && uword==L"~~end~~" )
return word;
else if( word.length()==9 && uword==L"~~begin~~" )
return word;
else if( word.length()<=model_suffix_len )
return uword;
else
return lem::UCString(L'~')+lem::right( uword, model_suffix_len );
}
void TreeScorerBoundVariables::AddVariable( const lem::UCString & upper_name, const Solarix::Word_Form * wordform )
{
LEM_CHECKIT_Z( !upper_name.empty() );
LEM_CHECKIT_Z( wordform!=NULL );
bound_variables[ upper_name ] = wordform;
return;
}
std::pair<int, int> NGramsDBMS::FindRawNGramsID(
const lem::FString &suffix,
const lem::FString &segment,
const lem::UCString &left,
const lem::UCString &right
)
{
FString sql = lem::format_str("SELECT w, N.id"
" FROM NGRAM_WORDS%s%s wrd1, NGRAM_WORDS%s%s wrd2, NGRAM2%s%s N"
" WHERE wrd1.word='%s' AND wrd2.word='%s' AND iword1=wrd1.id AND iword2=wrd2.id"
, suffix.c_str(), segment.c_str()
, suffix.c_str(), segment.c_str()
, suffix.c_str(), segment.c_str()
, lem::to_utf8(left.c_str()).c_str(), to_utf8(right.c_str()).c_str());
std::pair<int, int> r = SelectIntPair(sql);
return r;
}
void NGramsDBMS::InsertWord(
const lem::FString &suffix,
int id,
const lem::UCString &word
)
{
FString an(to_utf8(word.c_str()));
FString sql = lem::format_str("INSERT INTO NGRAM_WORDS%s( id, word ) VALUES ( %d, '%s' )", suffix.c_str(), id, an.c_str());
ExecuteSql(sql);
return;
}
int NGramsDBMS::FindWord(
const lem::FString &suffix,
const lem::FString &sgm,
const lem::UCString &word
)
{
FString sql = lem::format_str("SELECT id FROM NGRAM_WORDS%s%s WHERE word='%s'"
, suffix.c_str(), sgm.c_str(), to_utf8(word.c_str()).c_str()
);
const int id = SelectInt(sql, -1);
return id;
}
int Lemmatizator::MatchSuffix( const lem::UCString & suffix ) const
{
LEM_CHECKIT_Z( !suffix.empty() );
int id = -1;
std::map<lem::UCString,int>::const_iterator it = suffix2id.find(suffix);
if( it!=suffix2id.end() )
id = it->second;
return id;
}
SyllabContext::SyllabContext(GraphGram &alphabet, const lem::UCString &word, int id_language)
{
LEM_CHECKIT_Z(id_language != UNKNOWN);
const SG_Language &lang = alphabet.GetDict().GetSynGram().languages()[id_language];
const lem::MCollect<int> & id_alphabets = lang.GetAlphabets();
points.push_back(new SyllabContextLeftBoundary());
// Разбираем исходное слово по символам, каждый символ ищем в алфавите и создаем по результатам одну точку контекста
lem::WideStringUcs4 ucs4(word.c_str());
lem::uint32_t c;
while ((c = ucs4.Fetch()) != 0)
{
const Word_Coord wc = alphabet.entries().FindSymbol(c, id_alphabets);
if (wc.IsUnknown())
{
SyllabContextUnknownSymbol *point = new SyllabContextUnknownSymbol(c);
points.push_back(point);
}
else
{
const GG_Entry &e = alphabet.entries()[wc.GetEntry()];
const int id_class = e.GetClass();
const GG_EntryForm & f = e.forms()[wc.GetForm()];
lem::MCollect<GramCoordPair> coords = f.dims();
for (lem::Container::size_type k = 0; k < e.attrs().size(); ++k)
coords.push_back(e.attrs()[k]);
SyllabContextSymbol *point = new SyllabContextSymbol(c, e.GetName(), wc.GetEntry(), id_class, coords);
points.push_back(point);
}
}
points.push_back(new SyllabContextRightBoundary());
return;
}
void NGramsDBMS::FindRaw3GramsWithCenter(
const lem::FString &suffix,
const lem::FString &sgm,
float max_w3,
const lem::UCString ¢er,
int min_freq,
std::map< std::pair<lem::UCString, lem::UCString>, float > & reslist
)
{
FString sql = lem::format_str(
"SELECT wrd1.word, wrd3.word, w"
" FROM NGRAM3%s%s, NGRAM_WORDS%s%s wrd1, NGRAM_WORDS%s%s wrd2, NGRAM_WORDS%s%s wrd3"
" WHERE wrd2.word='%s' AND iword2=wrd2.id AND wrd1.id=iword1 AND wrd3.id=iword3"
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, to_utf8(center.c_str()).c_str());
std::unique_ptr<LS_ResultSet> rs(Select(sql));
while (rs->Fetch())
{
const int f = rs->GetInt(2);
if (f >= min_freq)
{
UCString cw1(rs->GetUCString(0));
UCString cw3(rs->GetUCString(1));
const float ff = float(f) / max_w3;
auto it = reslist.find(std::make_pair(cw1, cw3));
if (it == reslist.end())
reslist.insert(std::make_pair(std::make_pair(cw1, cw3), ff));
else
it->second += ff;
}
}
return;
}
void NGramsDBMS::Find2GramsWithRight(
const lem::FString &suffix,
const lem::FString &sgm,
float max_w2,
const lem::UCString &right,
int min_freq,
std::map< lem::UCString, float > & reslist
)
{
FString sql = lem::format_str(
"SELECT wrd2.word, w"
" FROM NGRAM_WORDS%s%s wrd1, NGRAM2%s%s, NGRAM_WORDS%s%s wrd2"
" WHERE wrd1.word='%s' AND iword2=wrd1.id AND wrd2.id=iword1"
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, suffix.c_str(), sgm.c_str()
, to_utf8(right.c_str()).c_str());
std::unique_ptr<LS_ResultSet> rs(Select(sql));
while (rs->Fetch())
{
const int f = rs->GetInt(1);
if (f >= min_freq)
{
UCString cs(rs->GetUCString(0));
const float ff = float(f) / max_w2;
auto it = reslist.find(cs);
if (it == reslist.end())
reslist.insert(std::make_pair(cs, ff));
else
it->second += ff;
}
}
return;
}
SG_calibrator::SG_calibrator(const lem::UCString & keyword, const SynGram &sg, const Sol_IO &io, Macro_Parser &txtfile)
{
if (keyword.eqi(L"wordentry_freq"))
freq_type = WordEntryFreq;
else if (keyword.eqi(L"wordform_score"))
freq_type = WordFormScore;
else if (keyword.eqi(L"wordforms_score"))
freq_type = WordFormsScore;
id_class = UNKNOWN;
freq = 0;
word = txtfile.read().string();
word.strip(L'"');
// если далее идет открывающая фигурная скобка, то значит конкретизируется словоформа (или несколько
// словоформ).
if (txtfile.probe(B_OFIGPAREN))
{
while (!txtfile.eof())
{
if (txtfile.pick().GetToken() == B_CFIGPAREN)
{
txtfile.read();
break;
}
// для обычных: координата:состояние
// для бистабильных: координата
lem::Iridium::BethToken coord_name = txtfile.read();
if (id_class == UNKNOWN)
{
const int id_class0 = sg.FindClass(coord_name);
if (id_class0 != UNKNOWN)
{
id_class = id_class0;
continue;
}
}
bool AFFIRM = true;
if (coord_name.GetToken() == B_NEGATIVE)
{
// Оператор отрицания перед определением координаты!
AFFIRM = false;
coord_name = txtfile.read();
}
const GramCoordAdr iglob_coord = sg.FindCoord(coord_name.string());
if (!iglob_coord.IsDefined())
{
sg.GetIO().merr().printf("Unknown coordinate %us\n", coord_name.c_str());
lem::Iridium::Print_Error(coord_name, txtfile);
throw lem::E_BaseException();
}
if (sg.coords()[iglob_coord.GetIndex()].IsBistable())
{
// Имя состояния не может быть указано.
coords.push_back(GramCoordPair(iglob_coord, AFFIRM));
}
else
{
// После двоеточия должно идти имя состояния для координаты.
txtfile.read_it(B_COLON);
// Имя состояния.
BethToken state_name = txtfile.read();
// Получим индекс состояния для определенной координаты.
const int istate = sg.coords()[iglob_coord.GetIndex()]
.FindState(state_name.string());
if (istate == UNKNOWN)
{
// Нет такого состояния для этого измерения.
lem::Iridium::Print_Error(state_name, txtfile);
sg.GetIO().merr().printf(
"State [%vfE%us%vn] is not declared for coordinate [%vfE%us%vn]\n"
, state_name.c_str(), coord_name.c_str()
);
throw E_ParserError();
}
coords.push_back(GramCoordEx(iglob_coord, istate, AFFIRM));
}
}
}
txtfile.read_it(B_EQUAL);
if (txtfile.probe(B_SUB))
freq = -txtfile.read_int();
else
freq = txtfile.read_int();
return;
}
void LexicalAutomat::SaveRules_SQL( OFormatter &out, OFormatter &alters, const SQL_Production &sql_version )
{
if( sql_version.norules )
return;
PM_Automat::SaveRules_SQL( "la", out, alters, sql_version );
lem::Ptr<LS_ResultSet> rs_predicates( GetStorage().ListPredicateTemplate() );
while( rs_predicates->Fetch() )
{
int id = rs_predicates->GetInt(0);
lem::Ptr<PredicateTemplate> t( GetStorage().LoadPredicateTemplate(id) );
lem::UFString src( sql_version.SqlStr(t->GetSrc()) );
lem::UFString params;
for( lem::Container::size_type i=0; i<t->GetParams().size(); ++i )
if( i>0 )
{
params.Add_Dirty( L"," );
params.Add_Dirty( t->GetParams()[i].c_str() );
}
else
{
params.Add_Dirty( t->GetParams()[i].c_str() );
}
out.printf( "INSERT INTO predicate_template( src, params ) VALUES ( '%us', '%us' );\n", src.c_str(), params.c_str() );
}
rs_predicates.Delete();
out.eol();
out.flush();
lem::Ptr<LS_ResultSet> rs_assocs( GetStorage().ListAssociatedEntries() );
while( rs_assocs->Fetch() )
{
const int id = rs_assocs->GetInt(0);
lem::Ptr<WordAssociation> assoc( GetStorage().LoadAssocitation(id) );
assoc->SaveSQL( out, sql_version );
}
rs_assocs.Delete();
out.flush();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> misspelling( GetStorage().ListMisspelling() );
while( misspelling->Fetch() )
{
int id_language = misspelling->GetInt(1);
UFString old_word = sql_version.SqlStr(misspelling->GetUFString(2));
UFString new_word = sql_version.SqlStr(misspelling->GetUFString(3));
old_word = sql_version.SqlStr(old_word);
new_word = sql_version.SqlStr(new_word);
out.printf( "INSERT INTO misspelling( id_language, original_word, substitution ) VALUES ( %d, '%us', '%us' );\n",
id_language, old_word.c_str(), new_word.c_str() );
}
misspelling.Delete();
out.eol();
lem::Ptr<WordEntrySetEnumerator> wordentry_sets( wordentry_set->ListWordEntrySets() );
while( wordentry_sets->Fetch() )
{
const WordEntrySetItem &wes = wordentry_sets->GetItem();
lem::UFString s;
for( std::set<int>::const_iterator it=wes.ies.begin(); it!=wes.ies.end(); ++it )
{
if( !s.empty() )
s.Add_Dirty(L' ');
s.Add_Dirty( lem::to_ustr(*it).c_str() );
}
out.printf( "INSERT INTO word_entry_set( id, name, ies ) VALUES ( %d, '%us', '%us' );\n",
wes.GetId(), lem::to_upper(wes.GetName()).c_str(), s.c_str() );
}
wordentry_sets.Delete();
out.eol();
lem::Ptr<WordSetEnumerator> word_sets( wordentry_set->ListWordSets() );
while( word_sets->Fetch() )
{
const WordSetItem &wes = word_sets->GetItem();
lem::UFString s;
for( std::set<lem::UCString>::const_iterator it=wes.words.begin(); it!=wes.words.end(); ++it )
{
if( !s.empty() )
s.Add_Dirty(LexiconStorage::WORD_SET_DELIMITER);
s.Add_Dirty( sql_version.SqlStr(*it) );
}
out.printf( "INSERT INTO word_set( id, name, words, case_sensitive ) VALUES ( %d, '%us', '%us', %d );\n",
wes.GetId(), lem::to_upper(wes.GetName()).c_str(), s.c_str(), wes.case_sensitive ? 1 : 0 );
}
word_sets.Delete();
out.eol();
// именованные наборы словосочетаний
lem::Ptr<LS_ResultSet> rs_collocation_sets( storage->ListCollocationSets() );
while( rs_collocation_sets->Fetch() )
{
int id = rs_collocation_sets->GetInt(0);
lem::UFString name = sql_version.SqlStr(rs_collocation_sets->GetUFString(1));
int case_sensitive = rs_collocation_sets->GetInt(2);
out.printf( "INSERT INTO collocation_set( id, name, case_sensitive ) VALUES ( %d, '%us', %d );\n",
id, name.c_str(), case_sensitive );
}
rs_collocation_sets.Delete();
lem::Ptr<LS_ResultSet> rs_collocation_items( storage->ListCollocationSetItems() );
while( rs_collocation_items->Fetch() )
{
int id = rs_collocation_items->GetInt(0);
int id_set = rs_collocation_items->GetInt(1);
lem::UFString words = sql_version.SqlStr(rs_collocation_items->GetUFString(2));
int n_word = rs_collocation_items->GetInt(3);
lem::UFString headword = sql_version.SqlStr(rs_collocation_items->GetUFString(4));
out.printf( "INSERT INTO collocation_set_item( id, id_set, words, n_word, headword ) VALUES ( %d, %d, '%us', %d, '%us' );\n",
id, id_set, words.c_str(), n_word, headword.c_str() );
}
rs_collocation_items.Delete();
lem::Ptr<LS_ResultSet> rs_collocation_headwords( storage->ListCollocationSetHeadwords() );
while( rs_collocation_headwords->Fetch() )
{
int id = rs_collocation_headwords->GetInt(0);
lem::UFString headword = sql_version.SqlStr(rs_collocation_headwords->GetUFString(1));
int min_len = rs_collocation_headwords->GetInt(2);
int max_len = rs_collocation_headwords->GetInt(3);
out.printf( "INSERT INTO collocation_headword( id, headword, min_len, max_len ) VALUES ( %d, '%us', %d, %d );\n",
id, headword.c_str(), min_len, max_len );
}
rs_collocation_headwords.Delete();
// --------------------------------------------
lem::Ptr<LS_ResultSet> rs_wordform_sets( storage->ListWordformSets() );
while( rs_wordform_sets->Fetch() )
{
int id = rs_wordform_sets->GetInt(0);
lem::UFString name = sql_version.SqlStr(rs_wordform_sets->GetUFString(1));
out.printf( "INSERT INTO wordform_set( id, name ) VALUES ( %d, '%us' );\n",
id, name.c_str() );
}
rs_wordform_sets.Delete();
lem::Ptr<LS_ResultSet> rs_wordform_set_coords( storage->ListWordformSetCoords() );
while( rs_wordform_set_coords->Fetch() )
{
int id = rs_wordform_set_coords->GetInt(0);
lem::UFString coords = rs_wordform_set_coords->GetUFString(1);
out.printf( "INSERT INTO wordform_set_coords( id, coords ) VALUES ( %d, '%us' );\n",
id, coords.c_str() );
}
rs_wordform_set_coords.Delete();
lem::Ptr<LS_ResultSet> rs_wordform_set_items( storage->ListWordformSetItems() );
while( rs_wordform_set_items->Fetch() )
{
const int id = rs_wordform_set_items->GetInt(0);
const int id_set = rs_wordform_set_items->GetInt(1);
lem::UFString word = sql_version.SqlStr( rs_wordform_set_items->GetUFString(2) );
const int id_entry = rs_wordform_set_items->GetInt(3);
const int id_coords = rs_wordform_set_items->GetInt(4);
out.printf( "INSERT INTO wordform_set_item( id, id_set, word, id_entry, id_coords ) VALUES ( %d, %d, '%us', %d, %d );\n",
id, id_set, word.c_str(), id_entry, id_coords );
}
rs_wordform_set_items.Delete();
// --------------------------------------------
lem::Ptr<LS_ResultSet> rs_unb( storage->ListUnbreakableRules(ANY_STATE) );
while( rs_unb->Fetch() )
{
const int id_rule = rs_unb->GetInt(0);
lem::Ptr<LA_UnbreakableRule> rule( storage->GetUnbreakableRule(id_rule) );
lem::UFString pattern(rule->GetPattern());
pattern = sql_version.SqlStr(pattern);
out.printf( "INSERT INTO unbreakable( id, name, id_language, is_regex,"
" pattern, id_src ) VALUES ( %d, '%us', %d, %d, '%us', %d );\n",
rule->GetId(), lem::to_upper(rule->GetName()).c_str(), rule->GetLanguage(),
rule->IsRegex() ? 1 : 0,
pattern.c_str(), rule->GetSourceLocation() );
}
out.eol();
lem::Ptr<LS_ResultSet> rs_recog = storage->ListRecognitionRules(ANY_STATE,true);
while( rs_recog->Fetch() )
{
const int id = rs_recog->GetInt(0);
lem::Ptr<LA_RecognitionRule> rule = storage->GetRecognitionRule(id);
lem::UFString condition(rule->GetCondition());
condition = sql_version.SqlStr(condition);
lem::UFString word;
if( rule->IsWordMatcher() )
{
word = sql_version.SqlStr(condition);
word.dress(L'\'');
}
else
word = L"NULL";
lem::UFString coords;
coords.reserve(64);
for( lem::Container::size_type i=0; i<rule->GetCoords().size(); ++i )
{
if( i>0 )
coords.Add_Dirty(L' ');
coords.Add_Dirty( to_ustr( rule->GetCoords()[i].GetCoord().GetIndex() ).c_str() );
coords.Add_Dirty( L':' );
coords.Add_Dirty( to_ustr( rule->GetCoords()[i].GetState() ).c_str() );
}
out.printf( "INSERT INTO recog_rule( id, name, id_language, is_syllab, is_regex, is_prefix, is_affix,"
" r_condition, id_entry, rel, coords, id_src, word ) VALUES ( %d, '%us', %d, %d, %d, %d, %d, '%us', %d, %d, '%us', %d, %us );\n",
rule->GetId(), lem::to_upper(rule->GetName()).c_str(), rule->GetLanguage(),
rule->IsSyllab() ? 1 : 0,
rule->IsRegex() ? 1 : 0,
rule->IsPrefix() ? 1 : 0,
rule->IsAffix() ? 1 : 0,
condition.c_str(), rule->GetEntryKey(), rule->GetRel().GetInt(), coords.c_str(), rule->GetSourceLocation(), word.c_str() );
}
out.eol();
lem::Ptr<LS_ResultSet> rs_ph( storage->ListPhoneticRules(ANY_STATE) );
while( rs_ph->Fetch() )
{
int id = rs_ph->GetInt(0);
lem::Ptr<LA_PhoneticRule> rule = storage->GetPhoneticRule(id);
lem::UFString condition(rule->GetCondition().GetContext().c_str());
condition = sql_version.SqlStr(condition);
lem::UFString result(rule->GetResult().GetContext().c_str());
result = sql_version.SqlStr(result);
out.printf( "INSERT INTO ph_rule( id, name, id_language, is_prefix, is_affix,"
" r_condition, r_result, rel, id_src ) VALUES ( %d, '%us', %d, %d, %d, '%us', '%us', %d, %d );\n",
rule->GetId(), lem::to_upper(rule->GetName()).c_str(), rule->GetLanguage(),
rule->IsPrefix() ? 1 : 0, rule->IsAffix() ? 1 : 0,
condition.c_str(), result.c_str(), rule->GetVal().GetInt(), rule->GetSourceLocation() );
}
out.eol();
lem::Ptr<LS_ResultSet> rs_syn = storage->ListSynPatternTrees();
while( rs_syn->Fetch() )
{
const int id_tree = rs_syn->GetInt(0);
const lem::UCString name = rs_syn->GetUCString(1);
const lem::UCString uname = rs_syn->GetUCString(2);
const int id_language = rs_syn->GetInt(3);
const int pattern_type = rs_syn->GetInt(4);
std::pair<SynPatternTreeNode*,lem::UCString> tree_name = storage->GetSynPatternTree(id_tree);
lem::Ptr<SynPatternTreeNode> tree = tree_name.first;
lem::MemStream mem(true);
tree->SaveBin(mem);
const int sz = mem.tellp();
const char* bytes = mem.get_Block();
lem::FString hex;
Storage_Data2Hex( bytes, sz, hex );
if( sql_version.type==SQL_Production::SQLite )
{
int id0=-1;
int segment_no=0;
out.printf( "INSERT INTO SynPatternTree( id_tree, id0, segment_no, name, uname, id_language, pattern_type, bin )"
" VALUES ( %d, %d, %d, '%us', '%us', %d, %d, X'%s' );\n", id_tree, id0, segment_no, name.c_str(), uname.c_str(), id_language, pattern_type, hex.c_str() );
}
else
{
out.printf( "INSERT INTO SynPatternTree( id_tree, name, uname, id_language, pattern_type, bin )"
" VALUES ( %d, '%us', '%us', %d, %d, 0x%s );\n", id_tree, name.c_str(), uname.c_str(), id_language, pattern_type, hex.c_str() );
}
}
out.eol();
lem::Ptr<LS_ResultSet> rs_c = storage->ListCropRules(ANY_STATE);
while( rs_c->Fetch() )
{
const int id = rs_c->GetInt(0);
lem::Ptr<LA_CropRule> rule = storage->GetPreprocessorCropRule(id);
lem::UFString condition(rule->GetCondition());
condition = sql_version.SqlStr(condition);
lem::UFString result(rule->GetResult());
result = sql_version.SqlStr(result);
out.printf( "INSERT INTO crop_rule( id, name, id_language, is_prefix, is_affix,"
" r_condition, r_result, is_regex, case_sensitive, id_src, rel )"
" VALUES ( %d, '%us', %d, %d, %d, '%us', '%us', %d, %d, %d, %d );\n",
id, lem::to_upper(rule->GetName()).c_str(), rule->GetLanguage(),
rule->IsPrefix() ? 1 : 0, rule->IsAffix() ? 1 : 0,
condition.c_str(), result.c_str(), rule->IsRegex() ? 1 : 0, rule->IsCaseSensitive() ? 1 : 0,
rule->GetSourceLocation(), rule->GetRel().GetInt() );
}
out.eol();
// ------------------
// база знаний
// ------------------
lem::Ptr<LS_ResultSet> rs_kb1 = storage->ListFactGroups();
while( rs_kb1->Fetch() )
{
const int id = rs_kb1->GetInt(0);
const int id_language = rs_kb1->GetInt(1);
lem::UCString name = rs_kb1->GetUCString(2);
const int n_arg = rs_kb1->GetInt(3);
const int n_ret = rs_kb1->GetInt(4);
const int query_mode = rs_kb1->GetInt(5);
const int ret_type = rs_kb1->GetInt(6);
const int violation_score = rs_kb1->GetInt(7);
const int violation_handler = rs_kb1->GetInt(8);
out.printf( "INSERT INTO kb_facts( id, id_language, name, n_arg, n_ret, query_mode, ret_type, violation_score, violation_handler ) VALUES ( %d, %d, '%us', %d, %d, %d, %d, %d, %d );\n"
, id, id_language, name.c_str(), n_arg, n_ret, query_mode, ret_type, violation_score, violation_handler );
}
rs_kb1.Delete();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> rs_kb2 = storage->ListFacts();
lem::UCString null_str(L"NULL");
while( rs_kb2->Fetch() )
{
const int id = rs_kb2->GetInt(0);
const int id_group = rs_kb2->GetInt(1);
const int boolean_return = rs_kb2->GetInt(2);
const int integer_return = rs_kb2->GetInt(3);
lem::CString bool_ret, int_ret;
if( boolean_return==-1 )
bool_ret = "NULL";
else
bool_ret = lem::to_str(boolean_return);
if( integer_return==-1 )
int_ret = "NULL";
else
int_ret = lem::to_str(integer_return);
const int false_score = rs_kb2->GetInt(14);
out.printf( "INSERT INTO kb_fact( id, id_group, boolean_return, integer_return, false_score )"
" VALUES ( %d, %d, %d, %d, %d );\n"
, id, id_group, bool_ret.c_str(), int_ret.c_str(), false_score );
}
rs_kb2.Delete();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> rs_kbi1 = storage->ListFactIndex1();
while( rs_kbi1->Fetch() )
{
const int id = rs_kbi1->GetInt(0);
const int id_group = rs_kbi1->GetInt(1);
const int id_fact = rs_kbi1->GetInt(2);
lem::UCString word1 = rs_kbi1->GetUCString(9);
lem::UCString word2 = rs_kbi1->GetUCString(10);
lem::UCString word3 = rs_kbi1->GetUCString(11);
lem::UCString word4 = rs_kbi1->GetUCString(12);
lem::UCString word5 = rs_kbi1->GetUCString(13);
if( word1.empty() )
word1 = null_str;
else
word1.dress_apostrophes();
if( word2.empty() )
word2 = null_str;
else
word2.dress_apostrophes();
if( word3.empty() )
word3 = null_str;
else
word3.dress_apostrophes();
if( word4.empty() )
word4 = null_str;
else
word4.dress_apostrophes();
if( word5.empty() )
word5 = null_str;
else
word5.dress_apostrophes();
const int false_score = rs_kbi1->GetInt(14);
out.printf( "INSERT INTO kb_fact_index1( id, id_group, id_fact,"
" word1, word2, word3, word4, word5"
" ) VALUES ( %d, %d, %d,"
" %s, %s, %s, %s, %s );\n"
, id, id_group, id_fact
, word1.c_str(), word2.c_str(), word3.c_str(), word4.c_str(), word5.c_str()
);
}
rs_kbi1.Delete();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> rs_kbi2 = storage->ListFactIndex2();
while( rs_kbi2->Fetch() )
{
const int id = rs_kbi2->GetInt(0);
const int id_group = rs_kbi2->GetInt(1);
const int id_fact = rs_kbi2->GetInt(2);
const int id_entry1 = rs_kbi2->GetInt(4);
const int id_entry2 = rs_kbi2->GetInt(5);
const int id_entry3 = rs_kbi2->GetInt(6);
const int id_entry4 = rs_kbi2->GetInt(7);
const int id_entry5 = rs_kbi2->GetInt(8);
CString entry1("NULL"), entry2("NULL"), entry3("NULL"), entry4("NULL"), entry5("NULL");
if( id_entry1!=-1 ) entry1 = lem::to_str(id_entry1);
if( id_entry2!=-1 ) entry2 = lem::to_str(id_entry2);
if( id_entry3!=-1 ) entry3 = lem::to_str(id_entry3);
if( id_entry4!=-1 ) entry4 = lem::to_str(id_entry4);
if( id_entry5!=-1 ) entry5 = lem::to_str(id_entry5);
out.printf( "INSERT INTO kb_fact_index2( id, id_group, id_fact,"
" id_entry1, id_entry2, id_entry3, id_entry4, id_entry5"
" ) VALUES ( %d, %d, %d,"
" %s, %s, %s, %s, %s );\n"
, id, id_group, id_fact
, entry1.c_str(), entry2.c_str(), entry3.c_str(), entry4.c_str(), entry5.c_str()
);
}
rs_kbi2.Delete();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> rs_kb3 = storage->ListFactArguments();
while( rs_kb3->Fetch() )
{
const int id = rs_kb3->GetInt(0);
const int id_fact = rs_kb3->GetInt(1);
const int argument_index = rs_kb3->GetInt(2);
lem::UFString word_text( sql_version.SqlStr(rs_kb3->GetUFString(3)) );
const int id_entry = rs_kb3->GetInt(4);
const int id_class = rs_kb3->GetInt(5);
lem::UFString wordentryset_name( sql_version.SqlStr(rs_kb3->GetUFString(6)) );
lem::UFString wordset_name( sql_version.SqlStr(rs_kb3->GetUFString(7)) );
const int n_coords = rs_kb3->GetInt(8);
const int id_coord1 = rs_kb3->GetInt(9);
const int id_state1 = rs_kb3->GetInt(10);
const int affirm1 = rs_kb3->GetInt(11);
const int id_coord2 = rs_kb3->GetInt(12);
const int id_state2 = rs_kb3->GetInt(13);
const int affirm2 = rs_kb3->GetInt(14);
const int id_coord3 = rs_kb3->GetInt(15);
const int id_state3 = rs_kb3->GetInt(16);
const int affirm3 = rs_kb3->GetInt(17);
const int id_coord4 = rs_kb3->GetInt(18);
const int id_state4 = rs_kb3->GetInt(19);
const int affirm4 = rs_kb3->GetInt(20);
const int thesauruscheck_link = rs_kb3->GetInt(21);
const int thesauruscheck_entry = rs_kb3->GetInt(22);
const int is_positive = rs_kb3->GetInt(23);
const int is_regex = rs_kb3->GetInt(24);
const int case_sensitive = rs_kb3->GetInt(25);
const int id_metaentry = rs_kb3->GetInt(26);
out.printf( "INSERT INTO kb_argument_point("
" id, id_fact, argument_index, word_text, id_entry,"
" id_class, wordentryset_name, wordset_name, n_coords, id_coord1,"
" id_state1, affirm1, id_coord2, id_state2, affirm2,"
" id_coord3, id_state3, affirm3, id_coord4, id_state4,"
" affirm4, thesauruscheck_link, thesauruscheck_entry, is_positive,"
" is_regex, case_sensitive, id_metaentry"
" ) VALUES ("
" %d, %d, %d, '%us', %d,"
" %d, '%us', '%us', %d, %d,"
" %d, %d, %d, %d, %d,"
" %d, %d, %d, %d, %d,"
" %d, %d, %d, %d,"
" %d, %d, %d );\n",
id, id_fact, argument_index, word_text.c_str(), id_entry,
id_class, wordentryset_name.c_str(), wordset_name.c_str(), n_coords, id_coord1,
id_state1, affirm1, id_coord2, id_state2, affirm2,
id_coord3, id_state3, affirm3, id_coord4, id_state4,
affirm4, thesauruscheck_link, thesauruscheck_entry, is_positive,
is_regex, case_sensitive, id_metaentry );
}
rs_kb3.Delete();
out.printf( "\n\n" );
// ---- Выгружаем tree scorers
lem::Ptr<LS_ResultSet> ts_head_lemma( GetStorage().ListTreeScorerLemmas() );
while( ts_head_lemma->Fetch() )
{
int id_entry = ts_head_lemma->GetInt(0);
int id_head_point = ts_head_lemma->GetInt(1);
int score_type = ts_head_lemma->GetInt(2);
UFString score_expr = ts_head_lemma->GetUFString(3);
int root_node = ts_head_lemma->GetInt(4);
int id_group = ts_head_lemma->GetInt(5);
out.printf( "INSERT INTO ts_head_lemma( id_entry, id_head_point, score_type, score_expr, root_node, id_group ) VALUES ( %d, %d, %d, '%us', %d, %d );\n",
id_entry, id_head_point, score_type, score_expr.c_str(), root_node, id_group );
}
ts_head_lemma.Delete();
out.eol();
lem::Ptr<LS_ResultSet> ts_head_word( GetStorage().ListTreeScorerWords() );
while( ts_head_word->Fetch() )
{
lem::UFString word_text( sql_version.SqlStr(ts_head_word->GetUFString(0)) );
int id_head_point = ts_head_word->GetInt(1);
int score_type = ts_head_word->GetInt(2);
UFString score_expr = ts_head_word->GetUFString(3);
int root_node = ts_head_word->GetInt(4);
int id_group = ts_head_word->GetInt(5);
out.printf( "INSERT INTO ts_head_word( word, id_head_point, score_type, score_expr, root_node, id_group ) VALUES ( '%us', %d, %d, '%us', %d, %d );\n",
word_text.c_str(), id_head_point, score_type, score_expr.c_str(), root_node, id_group );
}
ts_head_word.Delete();
out.eol();
lem::Ptr<LS_ResultSet> ts_point = storage->ListTreeScorerPoints();
while( ts_point->Fetch() )
{
const int id = ts_point->GetInt(0);
const int id_parent = ts_point->GetInt(1);
lem::UFString word_text( sql_version.SqlStr(ts_point->GetUFString(2)) );
const int id_entry = ts_point->GetInt(3);
const int id_class = ts_point->GetInt(4);
lem::UFString wordentryset_name( sql_version.SqlStr(ts_point->GetUFString(5)) );
lem::UFString wordset_name( sql_version.SqlStr(ts_point->GetUFString(6)) );
lem::UFString wordformset_name( sql_version.SqlStr(ts_point->GetUFString(7)) );
const int n_coords = ts_point->GetInt(8);
const int id_coord1 = ts_point->GetInt(9);
const int id_state1 = ts_point->GetInt(10);
const int affirm1 = ts_point->GetInt(11);
const int id_coord2 = ts_point->GetInt(12);
const int id_state2 = ts_point->GetInt(13);
const int affirm2 = ts_point->GetInt(14);
const int id_coord3 = ts_point->GetInt(15);
const int id_state3 = ts_point->GetInt(16);
const int affirm3 = ts_point->GetInt(17);
const int id_coord4 = ts_point->GetInt(18);
const int id_state4 = ts_point->GetInt(19);
const int affirm4 = ts_point->GetInt(20);
const int thesauruscheck_link = ts_point->GetInt(21);
const int thesauruscheck_entry = ts_point->GetInt(22);
const int is_positive = ts_point->GetInt(23);
const int children_quantification_min_count = ts_point->GetInt(24);
const int children_quantification_max_count = ts_point->GetInt(25);
const int n_set_checkers = ts_point->GetInt(26);
lem::UFString set_name1( sql_version.SqlStr(ts_point->GetUFString(27)) );
const int set_type1 = ts_point->GetInt(28);
const int set_affirm1 = ts_point->GetInt(29);
lem::UFString set_name2( sql_version.SqlStr(ts_point->GetUFString(30)) );
const int set_type2 = ts_point->GetInt(31);
const int set_affirm2 = ts_point->GetInt(32);
lem::UFString set_name3( sql_version.SqlStr(ts_point->GetUFString(33)) );
const int set_type3 = ts_point->GetInt(34);
const int set_affirm3 = ts_point->GetInt(35);
const int link_type = ts_point->GetInt(36);
const int id_metaentry = ts_point->GetInt(37);
const int relative_position = ts_point->GetInt(38);
const int id_src = ts_point->GetInt(39);
const int id_group = ts_point->GetInt(40);
const int quantification_min_count = ts_point->GetInt(41);
const int quantification_max_count = ts_point->GetInt(42);
lem::UFString marker( sql_version.SqlStr(ts_point->GetUFString(43)) );
const int n_backref = ts_point->GetInt(44);
int br_affirm_0 = -1;
int br_coord0_0 = -1;
lem::UFString br_marker_0;
int br_coord1_0 = -1;
int br_for_group_0 = -1;
int br_affirm_1 = -1;
int br_coord0_1 = -1;
lem::UFString br_marker_1;
int br_coord1_1 = -1;
int br_for_group_1 = -1;
int br_affirm_2 = -1;
int br_coord0_2 = -1;
lem::UFString br_marker_2;
int br_coord1_2 = -1;
int br_for_group_2 = -1;
if( n_backref>0 )
{
br_affirm_0 = ts_point->GetInt(45);
br_coord0_0 = ts_point->GetInt(46);
br_marker_0 = sql_version.SqlStr(ts_point->GetUFString(47));
br_coord1_0 = ts_point->GetInt(48);
br_for_group_0 = ts_point->GetInt(49);
}
if( n_backref>1 )
{
br_affirm_1 = ts_point->GetInt(50);
br_coord0_1 = ts_point->GetInt(51);
br_marker_1 = sql_version.SqlStr(ts_point->GetUFString(52));
br_coord1_1 = ts_point->GetInt(53);
br_for_group_1 = ts_point->GetInt(54);
}
if( n_backref>2 )
{
br_affirm_2 = ts_point->GetInt(55);
br_coord0_2 = ts_point->GetInt(56);
br_marker_2 = sql_version.SqlStr(ts_point->GetUFString(57));
br_coord1_2 = ts_point->GetInt(58);
br_for_group_2 = ts_point->GetInt(59);
}
LEM_CHECKIT_Z( n_backref<=3 );
out.printf( "INSERT INTO ts_point("
" id, id_parent, word_text, id_entry,"
" id_class, wordentryset_name, wordset_name, wordformset_name, "
" n_coords, id_coord1, id_state1, affirm1,"
" id_coord2, id_state2, affirm2, id_coord3, id_state3,"
" affirm3, thesauruscheck_link, thesauruscheck_entry, is_positive, children_quantification_min_count, children_quantification_max_count,"
" id_coord4, id_state4, affirm4,"
" n_setcheckers, set_name1, set_type1, set_affirm1, set_name2,"
" set_type2, set_affirm2, set_name3, set_type3, set_affirm3,"
" link_type, id_metaentry, relative_position, id_src, id_group,"
" quantification_min_count, quantification_max_count, marker,"
"n_backref,"
"br_affirm_0, br_coord0_0, br_marker_0, br_coord1_0, br_for_group_0,"
"br_affirm_1, br_coord0_1, br_marker_1, br_coord1_1, br_for_group_1,"
"br_affirm_2, br_coord0_2, br_marker_2, br_coord1_2, br_for_group_2"
" ) VALUES ("
" %d, %d, '%us', %d,"
" %d, '%us', '%us', '%us',"
" %d, %d, %d, %d,"
" %d, %d, %d, %d, %d,"
" %d, %d, %d, %d, %d, %d,"
" %d, %d, %d,"
" %d, '%us', %d, %d, '%us',"
" %d, %d, '%us', %d, %d,"
" %d, %d, %d, %d, %d,"
" %d, %d, '%us',"
" %d,"
" %d, %d, '%us', %d, %d,"
" %d, %d, '%us', %d, %d,"
" %d, %d, '%us', %d, %d"
" );\n",
id, id_parent, word_text.c_str(), id_entry,
id_class, wordentryset_name.c_str(), wordset_name.c_str(), wordformset_name.c_str(),
n_coords, id_coord1, id_state1, affirm1,
id_coord2, id_state2, affirm2, id_coord3, id_state3,
affirm3, thesauruscheck_link, thesauruscheck_entry, is_positive, children_quantification_min_count, children_quantification_max_count,
id_coord4, id_state4, affirm4,
n_set_checkers, set_name1.c_str(), set_type1, set_affirm1, set_name2.c_str(),
set_type2, set_affirm2, set_name3.c_str(), set_type3, set_affirm3,
link_type, id_metaentry, relative_position, id_src, id_group,
quantification_min_count, quantification_max_count, marker.c_str(),
n_backref,
br_affirm_0, br_coord0_0, br_marker_0.c_str(), br_coord1_0, br_for_group_0,
br_affirm_1, br_coord0_1, br_marker_1.c_str(), br_coord1_1, br_for_group_1,
br_affirm_2, br_coord0_2, br_marker_2.c_str(), br_coord1_2, br_for_group_2
);
}
ts_point.Delete();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> ts_group = storage->ListTreeScorerGroups();
while( ts_group->Fetch() )
{
const int id = ts_group->GetInt(0);
lem::UFString name( sql_version.SqlStr(ts_group->GetUFString(1)) );
const int allow_unmatched_children = ts_group->GetInt(2);
out.printf( "INSERT INTO ts_group(id,name,allow_unmatched_children) VALUES (%d,'%us',%d);\n", id, name.c_str(), allow_unmatched_children );
}
ts_group.Delete();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> ts_group2root = storage->ListTreeScoreGroupMembers();
while( ts_group2root->Fetch() )
{
const int id_group = ts_group2root->GetInt(0);
const int id_root = ts_group2root->GetInt(1);
out.printf( "INSERT INTO ts_group2root(id_group,id_root) VALUES (%d,%d);\n", id_group, id_root );
}
ts_group2root.Delete();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> ts_gts = storage->ListGenericTreeScorers();
while( ts_gts->Fetch() )
{
const int id = ts_gts->GetInt(0);
const int id_head_point = ts_gts->GetInt(1);
const int id_group = ts_gts->GetInt(2);
lem::UFString score_expr = sql_version.SqlStr(ts_gts->GetUFString(3));
const int score_type = ts_gts->GetInt(4);
out.printf( "INSERT INTO generic_tree_scorer(id,id_head_point,id_group,score_type,score_expr) VALUES (%d,%d,%d,%d,'%us');\n", id, id_head_point, id_group, score_type, score_expr.c_str() );
}
ts_gts.Delete();
out.printf( "\n\n" );
lem::Ptr<LS_ResultSet> ts_pr = storage->ListTreeScorerPredicates();
while( ts_pr->Fetch() )
{
const int id = ts_pr->GetInt(0);
const int id_point = ts_pr->GetInt(1);
UCString func_name = ts_pr->GetUCString(2);
const int narg = ts_pr->GetInt(3);
const int id_class1 = ts_pr->GetInt(4);
out.printf( "INSERT INTO ts_predicate(id,id_point,func_name,narg,id_class1) VALUES (%d,%d,'%us',%d,%d);\n", id, id_point, func_name.c_str(), narg, id_class1 );
}
ts_pr.Delete();
out.printf( "\n\n" );
out.printf( "\n\n" );
return;
}
bool lem::is_udelim( const lem::UCString &s )
{ return s.size()==1 && lem::is_udelim(s.front()); }
