IndexSetForLoadingStage.h

Go to the documentation of this file.

#ifndef IndexSetForIndexingStage_h
#define IndexSetForIndexingStage_h


#include "DDCLessOperators.h"



class CItemIndexForLoading
{ 
        vector< CTokenNo >*     m_pCurrOccurs;

        size_t                          m_IndexItemOffset;

public:

        size_t  GetIndexItemOffset() const
        {
                return m_IndexItemOffset;
        };
        void    SetIndexItemOffset(size_t Value) 
        {
                m_IndexItemOffset = Value;
        };

        vector<CTokenNo>* GetOccurs() 
        {
                assert (m_pCurrOccurs);
                return m_pCurrOccurs;
        };

        const vector<CTokenNo>* GetOccurs() const 
        {
                assert (m_pCurrOccurs);
                return m_pCurrOccurs;
        };
        //  return the number of occurrences
        size_t GetOccursSize() const 
        {
                return  m_pCurrOccurs->size();
        };
        bool InitOccurs();

        void FreeOccurs();

        bool WriteOccurrences(FILE* fp) const;

        bool CheckOccurrences(CTokenNo EndTokenNo) const;
        
        void ClearOccurrences();
        
        bool ReadFromTemporalFile (FILE* fp);

        void WriteToTemporalFile (FILE* fp) const;
};






class CIndexSetForLoadingStage
{
        LessIndexString2< CItemIndexForLoading> m_LoadLess2;
        LessIndexString1< CItemIndexForLoading> m_LoadLess1;


        string  m_CurrOccurTempFileName;

        virtual string GetName() const = 0;

        vector<CItemIndexForLoading  >                  m_MemoryLoadIndexHash[256] ;

        vector<CItemIndexForLoading  >                  m_InputLoadIndexHash[256] ;
        FILE*           m_TempBigramsFile;
    
    deque<string>  m_LeftContext;

        bool FindIndexItemInVector (const char* Item, vector<CItemIndexForLoading>::iterator& it, vector<CItemIndexForLoading>& V);

        // finds an item in the swap index set, if it is not found, finds the item in the file index  set
        bool FindIndexItem (const char* Item, vector<CItemIndexForLoading>::iterator& it);

        bool    AddToMemoryLoadIndexAndClear(vector<CItemIndexForLoading>& Body, vector<CItemIndexForLoading>& FileIndexSet);
    int     GetHashNo(const char* Str) const;

protected:
        string      m_TempBigramsFileName;
        FILE*           m_TempStorageFile;
        string      m_TempStorageFileName;
        string  m_MainOccurTempFileName;

        size_t                  AddItemStrToBuffer(const char* Str, size_t StrLen);

    bool        IndexOneBigram(const char* Word1,  BYTE Word1Len, const char* Word2, BYTE Word2Len,const CTokenNo& TokenOffset, char SecondWordAddress);


public:
        bool                    m_bUseItemStorage;

    int                 m_MaxBigramWindowSize;

    int         m_BigramBorder;

        vector<char>    m_StringBuffer;


        CIndexSetForLoadingStage ();
        ~CIndexSetForLoadingStage ();

        bool    CreateTempFiles (string Path);
        bool    DeleteTempFiles();
        size_t  GetMemoryLoadIndexItemsCount() const;
        bool    SaveMemoryLoadIndex();
        bool    AddInputLoadIndexToMemoryLoadIndex();
        void    SortInputAndMemoryIndices();
        bool    AddMemoryLoadIndexToMainLoadIndex();
        void    InsertToInputLoadIndex(const char* Str, size_t StrLen, const vector<CTokenNo>& occurrences);
    void    CloseTempBigramsFile();
    bool    UseBigrams() const;
    void    ProcessBigramBorder(CTokenNo occurrence);
};



#endif