C++ MCollect::size方法代码示例

// Ищем в справочнике набор тегов, заданный списком tags. При необходимости
// вносим в БД новую запись. Возвращается ID найденной или созданной записи.
int TagSets::Register( const lem::MCollect< std::pair<int,int> > &tags )
{
 if( tags.empty() )
  {
   return 0;
  }

 #if defined LEM_THREADS
 lem::Process::CritSecLocker lock(&cs); 
 #endif

 // Для устранения вариантов записи одного и того же набора тегов отсортируем элементы по id_tag.
 lem::MCollect< std::pair<int,int> > *sorted_tags = new lem::MCollect< std::pair<int,int> >(tags);
 std::sort( sorted_tags->begin(), sorted_tags->end(), tags_sorter );

 // Такой кортеж есть?
 const int i = tag_ptr.find(*sorted_tags);
 
 if( i==UNKNOWN )
  { 
   // Нет.
   // Поищем в БД.
   lem::UFString s;
   if( tags.size()==1 )
    {
     s = lem::format_str( L"%d %d", tags.front().first, tags.front().second );
    }
   else if( tags.size()==2 )
    {
     s = lem::format_str( L"%d %d %d %d", sorted_tags->get(0).first, sorted_tags->get(0).second, sorted_tags->get(1).first, sorted_tags->get(1).second );
    }
   else
    {
     for( lem::Container::size_type i=0; i<sorted_tags->size(); ++i )
      {
       if(i>0) s += L' ';
       s += lem::format_str( L"%d %d", sorted_tags->get(i).first, sorted_tags->get(i).second );
      }
    }
   
   const int id = db->AddTagSet(s);

   id2tags.insert( std::make_pair(id,sorted_tags) );
   tag_ptr.push_back( sorted_tags );
   tagset_id.push_back(id);

   return id;
  }
 else
  {
   delete sorted_tags;
   return tagset_id[i];
  }
}

static bool x_contains_any_of_y( const lem::MCollect<int> &x, const lem::MCollect<int> &y )
{
 if( y.size()==1 )
  return x.find(y.front())!=UNKNOWN;
 else if( y.size()==2 )
  return x.find(y.front())!=UNKNOWN || x.find(y.back())!=UNKNOWN;
 else
  {
   for( lem::Container::size_type i=0; i<y.size(); ++i )
    if( x.find(y[i])!=UNKNOWN )
     return true;

   return false;
  }
}

void LEMM_Compiler::LoadNGram( lem::Iridium::Macro_Parser & txtfile, Dictionary & dict, lem::MCollect<int> & terms, int order ) const
{
 lem::Iridium::BSourceState beg = txtfile.tellp();

 while( !txtfile.eof() )
 {
  lem::Iridium::BethToken t = txtfile.read();
  if( lem::is_int(t.string()) )
   terms.push_back( lem::to_int(t.string()) );
  else
   {
    txtfile.seekp(t);
    break;
   }
 }

 if( terms.size() != order+1 )
  {
   dict.GetIO().merr().printf( "%vfDInvalid ngram%vn\n" );
   lem::Iridium::Print_Error( beg, txtfile );
   throw lem::E_ParserError();
  }

 return;
}

void Lemmatizator::Lemmatize( const lem::MCollect<lem::UCString> & words, lem::MCollect<lem::UCString> &lemmas )
{
 #if defined LEM_THREADS
 lem::Process::CritSecLocker lock(&cs);
 #endif

 if( !model_loaded )
  {
   bin->seekp( model_pos );
   model_loaded = true;
   model_available = bin->read_bool();
   if( model_available )
    {
     LoadModel();
    }
  }


 if( model_available )
  {
   LemmatizeViaModel( words, lemmas );
  }
 else
  {
   for( lem::Container::size_type i=0; i<words.size(); ++i )
    {
     lem::UCString lemma;
     Lemmatize( words[i], lemma );
     lemmas.push_back( lemma );
    }
  }

 return;
}

void SynPatternResult::FilterExportedNodes(const lem::MCollect< ExportNode > & must_be_exported)
{
    if (must_be_exported.empty())
    {
        exported_nodes.clear();
    }
    else
    {
        lem::MCollect< std::pair<const lem::UCString*, const Word_Form*> > filtered;
        for (lem::Container::size_type i = 0; i < exported_nodes.size(); ++i)
        {
            const lem::UCString & name = *exported_nodes[i].first;
            for (lem::Container::size_type j = 0; j < must_be_exported.size(); ++j)
            {
                if (must_be_exported[j].node_name == name)
                {
                    // Нашли ссылку, которую нужно перебросить в новый список, возможно уже под другим именем
                    filtered.push_back(std::make_pair(&must_be_exported[j].as_name, exported_nodes[i].second));
                    break;
                }
            }
        }

        exported_nodes = filtered;
    }

    return;
}

void SynPatternResult::SelectUnique_WithoutRemoval(lem::MCollect<const SynPatternResult*> & results)
{
    lem::MCollect<int> result_hash;
    lem::MCollect<const SynPatternResult*> unique_result;
    for (lem::Container::size_type k = 0; k < results.size(); ++k)
    {
        const SynPatternResult * result_k = results[k];
        const int h = result_k->CalcHash();

        bool found = false;
        for (lem::Container::size_type i = 0; i < unique_result.size(); ++i)
        {
            if (result_hash[i] == h)
            {
                if (SynPatternResult::Equals(result_k, unique_result[i]))
                {
                    found = true;
                    break;
                }
            }
        }

        if (!found)
        {
            result_hash.push_back(h);
            unique_result.push_back(result_k);
        }
    }

    results = unique_result;

    return;
}

void SynPatternResult::AppendDebugTrace(const lem::MCollect<SynPatternDebugTrace> & debug_trace2)
{
    for (lem::Container::size_type i = 0; i < debug_trace2.size(); ++i)
    {
        debug_trace.push_back(debug_trace2[i]);
    }

    return;
}

void TreeMatchingExperience::AddKBCheckerMatching( int id_facts, const lem::MCollect< const Solarix::Word_Form * > & arg_values, const KB_CheckingResult & res )
{
 LEM_CHECKIT_Z( id_facts!=UNKNOWN );
 LEM_CHECKIT_Z( arg_values.size()>0 );

 TME_KBChecker * y = new TME_KBChecker( arg_values, res );
 kbid2item.insert( std::make_pair( std::make_pair(id_facts,arg_values.front()), y ) );
 return;
}

void SyllabContext::Replace(int start_index, int count, lem::MCollect<SyllabContextPoint*> & new_points)
{
    for (int n = 0; n < count; ++n)
        points.Remove(start_index);

    for (lem::Container::size_type i = 0; i < new_points.size(); ++i)
    {
        const int new_index = start_index + CastSizeToInt(i);
        if (new_index == Count())
            points.push_back(new_points[i]);
        else
            points.Insert(new_index, new_points[i]);
    }

    return;
}

void NGramsStorage_SQLITE::UpdateNGrams(
                                        const lem::FString &suffix,
                                        int order, 
                                        const lem::MCollect< std::pair<int,int> > & list
                                       )
{
 char asql_buf[1000];

 sprintf( asql_buf, "UPDATE NGRAM%d%s SET w=w+? WHERE id=?", order, suffix.c_str() );

 sqlite3_stmt *stmt2=NULL;
 int res = sqlite3_prepare_v2( hdb, asql_buf, -1, &stmt2, NULL );
 if( res!=SQLITE_OK )
  {
   const char *errmsg = sqlite3_errmsg(hdb);
   throw E_BaseException( lem::format_str( L"SQLite error %S", errmsg ).c_str() );
  }

 for( lem::Container::size_type i=0; i<list.size(); ++i )
  {
   #if LEM_DEBUGGING==1
   int debug_id_ngram = list[i].first;
   int debug_add_w = list[i].second;
   if( debug_id_ngram==5 )
    {
//     printf( "add_w=%d\n", debug_add_w );
    }
   #endif

   res = sqlite3_bind_int( stmt2, 1, list[i].second );
   res = sqlite3_bind_int( stmt2, 2, list[i].first );
   res = sqlite3_step( stmt2 );

   if( res!=SQLITE_DONE )
    {
     const char *errmsg = sqlite3_errmsg(hdb);
     throw E_BaseException( lem::format_str( L"SQLite error %S", errmsg ).c_str() );
    }

   res = sqlite3_reset(stmt2);
  }

 sqlite3_finalize(stmt2);
 return;
}

void SynPatternResult::ResetCoordStates(int coord_id, const lem::MCollect<int> & state_ids)
{
    std::multimap< int /*id_coord*/, int /*id_state*/ > filtered;
    for (auto it = exported_coords.begin(); it != exported_coords.end(); ++it)
    {
        if (coord_id != it->first)
            filtered.insert(*it);
    }

    for (lem::Container::size_type i = 0; i < state_ids.size(); ++i)
    {
        filtered.insert(std::make_pair(coord_id, state_ids[i]));
    }

    exported_coords = filtered;

    return;
}

bool TreeMatchingExperience::FindKBCheckerMatching( int id_facts, const lem::MCollect< const Solarix::Word_Form * > & arg_values, KB_CheckingResult * res ) const
{
 LEM_CHECKIT_Z( id_facts!=UNKNOWN );
 LEM_CHECKIT_Z( arg_values.size()>0 );

 typedef KBID2ITEM::const_iterator IT;
 std::pair<IT,IT> pit = kbid2item.equal_range( std::make_pair( id_facts, arg_values.front() ) );

 for( IT it=pit.first; it!=pit.second; ++it )
  {
   if( arg_values == it->second->arg_values )
    {
     *res = it->second->res;
     return true;
    }
  }

 return false;
}

Word_Form::Word_Form( const lem::MCollect<const Word_Form*> &variants )
{
 LEM_CHECKIT_Z( !variants.empty() );

 // Первая версия становится основной, ее не копируем в альтернативы.
 for( lem::Container::size_type i=1; i<variants.size(); ++i )
  alt.push_back( new Word_Form(*variants[i]) );

 name = variants[0]->name;
 normalized = variants[0]->normalized;
 pair = variants[0]->pair;
 entry_key = variants[0]->entry_key;
 val = variants[0]->val;
 score=variants[0]->score;
 origin_pos = variants[0]->origin_pos;
 tokenizer_flags = variants[0]->tokenizer_flags;

 iversion = seq_iversion++;
 return;
}

static bool IsHtmlClosed( const lem::UFString &tag )
{
 if( tags1.empty() )
  {
   const wchar_t* stags[] = { L"br", L"hr", L"link", L"meta", L"img", L"input",
                              NULL
                            };

   int i=0;
   while(stags[i]!=NULL)
    tags1.push_back( lem::UFString(stags[i++]) );
  }

 for( lem::Container::size_type i=0; i<tags1.size(); ++i )
  {
   const lem::UFString &t = tags1[i];
   if( tag.eq_begi(t) && (tag.length()==t.length() || tag[ t.length() ]==L' ' ) ) 
    return true;
  }

 return false;
}

static bool IsTextDelimiterTag( const UFString &tag )
{
 if( tags2.empty() )
  {
   const wchar_t* stags[] = { L"p", L"br", L"table", L"td", L"tr", L"th",
                              L"ol", L"ul", L"li", L"dd", L"input", L"frame", L"div",
                              NULL
                            };

   int i=0;
   while(stags[i]!=NULL)
    tags2.push_back( lem::UFString(stags[i++]) );
  }

 for( lem::Container::size_type i=0; i<tags2.size(); ++i )
  {
   const lem::UFString &t = tags2[i];
   if( tag.eq_begi(t) && (tag.length()==t.length() || tag[ t.length() ]==L' ' ) ) 
    return true;
  }

 return false;
}