makelibdlg.cpp

通过VC编程实现收词程序,并且通过编程实现数据库的建立,并且建立了索引

// MakeLibDlg.cpp : implementation file
//

#include "stdafx.h"
#include "MakeLib.h"
#include "MakeLibDlg.h"

#include <direct.h>  // chdor()
#include "afxtempl.h"



#ifdef _DEBUG
#define new DEBUG_NEW
#undef THIS_FILE
static char THIS_FILE[] = __FILE__;
#endif

WORD  TranCode(WORD);
void GetGB2312Count(CStringArray * pArray);
CString StringFiltter(CString);
char *   StringFiltter(char *);


UINT  WordThreadProcess( LPVOID lpVoid);



/////////////////////////////////////////////////////////////////////////////
// CMakeLibDlg dialog

CMakeLibDlg::CMakeLibDlg(CWnd* pParent /*=NULL*/)
	: CDialog(CMakeLibDlg::IDD, pParent)
{
	//{{AFX_DATA_INIT(CMakeLibDlg)
		// NOTE: the ClassWizard will add member initialization here
	//}}AFX_DATA_INIT
	// Note that LoadIcon does not require a subsequent DestroyIcon in Win32
	m_hIcon = AfxGetApp()->LoadIcon(IDR_MAINFRAME);
}

void CMakeLibDlg::DoDataExchange(CDataExchange* pDX)
{
	CDialog::DoDataExchange(pDX);
	//{{AFX_DATA_MAP(CMakeLibDlg)
	DDX_Control(pDX, IDC_STATIC_HINT, m_Hint);
	DDX_Control(pDX, IDC_PROGRESS1, m_Prog);
	//}}AFX_DATA_MAP
}

BEGIN_MESSAGE_MAP(CMakeLibDlg, CDialog)
	//{{AFX_MSG_MAP(CMakeLibDlg)
	ON_WM_PAINT()
	ON_WM_QUERYDRAGICON()
	ON_BN_CLICKED(IDC_BUTTON_BROWSE, OnButtonBrowse)
	ON_BN_CLICKED(IDC_BUTTON_BWBROWS_FILE, OnButtonBwbrowsFile)
	ON_BN_CLICKED(IDC_BUTTON_RESTORE, OnButtonRestore)
	//}}AFX_MSG_MAP
END_MESSAGE_MAP()

/////////////////////////////////////////////////////////////////////////////
// CMakeLibDlg message handlers

BOOL CMakeLibDlg::OnInitDialog()
{
	CDialog::OnInitDialog();

	// Set the icon for this dialog.  The framework does this automatically
	//  when the application's main window is not a dialog
	SetIcon(m_hIcon, TRUE);			// Set big icon
	SetIcon(m_hIcon, FALSE);		// Set small icon
	
	// TODO: Add extra initialization here

	memset((void*) &ti, 0, sizeof(THRAD_INFOR));
	lpti = &ti;

	return TRUE;  // return TRUE  unless you set the focus to a control
}

// If you add a minimize button to your dialog, you will need the code below
//  to draw the icon.  For MFC applications using the document/view model,
//  this is automatically done for you by the framework.

void CMakeLibDlg::OnPaint() 
{
	if (IsIconic())
	{
		CPaintDC dc(this); // device context for painting

		SendMessage(WM_ICONERASEBKGND, (WPARAM) dc.GetSafeHdc(), 0);

		// Center icon in client rectangle
		int cxIcon = GetSystemMetrics(SM_CXICON);
		int cyIcon = GetSystemMetrics(SM_CYICON);
		CRect rect;
		GetClientRect(&rect);
		int x = (rect.Width() - cxIcon + 1) / 2;
		int y = (rect.Height() - cyIcon + 1) / 2;

		// Draw the icon
		dc.DrawIcon(x, y, m_hIcon);
	}
	else
	{
		CDialog::OnPaint();
	}
}

// The system calls this to obtain the cursor to display while the user drags
//  the minimized window.
HCURSOR CMakeLibDlg::OnQueryDragIcon()
{
	return (HCURSOR) m_hIcon;
}

void CMakeLibDlg::OnButtonBrowse() 
{
	BROWSEINFO      bi; 
	TCHAR			szDisplayName[MAX_PATH] = {'\0'}; 
	LPITEMIDLIST    pidl; 
	LPMALLOC		pMalloc = NULL;  
	ZeroMemory(&bi, sizeof(BROWSEINFO));  
	bi.hwndOwner = GetSafeHwnd();
	bi.pszDisplayName = szDisplayName; 
	bi.lpszTitle = TEXT("选择文件夹!"); 
	bi.ulFlags = BIF_RETURNONLYFSDIRS;  
	pidl = SHBrowseForFolder(&bi);  
	if(pidl)
	{
		SHGetPathFromIDList(pidl, szDisplayName); 
		SetDlgItemText(IDC_EDIT_FOLDER, szDisplayName);
	}
	if (SHGetMalloc(&pMalloc) == NOERROR) 
		pMalloc->Free((void *)pidl); 			

	
}

void CMakeLibDlg::OnOK() 
{
	GetDlgItemText(IDC_EDIT_FOLDER, strFolder);
	GetDlgItemText(IDC_EDIT_OUT,    strStdFile);
	if( strFolder.IsEmpty() )  return ;
	if( strStdFile.IsEmpty() )  return ;
	lpti->lpFolderPath   = (LPCTSTR) strFolder;
	lpti->lpStdFileName  = (LPCTSTR) strStdFile;
    lpti->pParent = this;
	lpti->pProg   = &m_Prog;
	AfxBeginThread(WordThreadProcess, lpti);
}


UINT  WordThreadProcess( LPVOID lpVoid)
{
	LPTHRAD_INFOR  lpti = ( LPTHRAD_INFOR ) lpVoid;
	CStringArray array;
	FindWordFile( CString( lpti->lpFolderPath),  &array);
	CStringArray arrWord;
	GetGB2312Count( & arrWord);
	int i=0,k=0,j=0, size = arrWord.GetSize(); 
	LPINDEX_ONE pOne = new INDEX_ONE [ size ];
	ZeroMemory( pOne, sizeof( INDEX_ONE ) * size);
	for( i=0;i<size;i++)
	{
		CString str = arrWord.GetAt(i);
		BYTE h = str.GetAt(0);
		BYTE l = str.GetAt(1); 
		WORD w = MAKEWORD(l, h);
		(*(pOne+i)).w  = w;
	}
 
	
	
	BYTE b1=0,b2=0 ;
	CString   strLine;
	int FileCount = array.GetSize();
	int WordCount = arrWord.GetSize();
	CStdioFile file;     
	
	HEAD_INFOR  head;
	memset((void*) &head, 0, sizeof(HEAD_INFOR));
	memcpy((void*)head.Flag, (const BYTE *) "WWWW", 4); 
	head.Unit[0] = 0xC055;
	head.Unit[0] = 0xC055;
	head.Unit[0] = 0xC055;
	head.Unit[0] = 0xC055;	
	head.size = size;
	CFile fout;
	fout.Open(lpti->lpStdFileName, CFile::modeCreate|CFile::modeWrite|CFile::typeBinary); 
	fout.Write((void*) &head, sizeof(HEAD_INFOR));
	fout.Seek(  sizeof(HEAD_INFOR) + sizeof(INDEX_ONE) * WordCount, CFile::begin);
	lpti->pProg->SetRange32(0, WordCount);
	lpti->pProg->SetStep(1);
	lpti->pProg->SetPos(0);

	
	for(i=0;i< WordCount;i++)
	{
        for(k=0; k<FileCount; k++)
		{
			CString str = array.GetAt(k);
			file.Open((LPCTSTR) str, CFile::modeRead);
			UINT len =0;
			file.ReadString(strLine);
			file.ReadString(strLine);
			strLine.TrimLeft();
			strLine.TrimRight();
			len = strLine.GetLength();
			switch(len)
			{
				case 4:
					(*(pOne+i)).FirstOffset2  =  fout.GetPosition(); 
					break;
				case 6:
					(*(pOne+i)).FirstOffset3  =  fout.GetPosition(); 
					break;
				case 8:
					(*(pOne+i)).FirstOffset4  =  fout.GetPosition(); 
					break;
			}

			file.SeekToBegin();
			while( file.ReadString(strLine) ) 
			{
				b1 = (BYTE) strLine.GetAt(0);
				b2 = (BYTE) strLine.GetAt(1);
				WORD w2 = MAKEWORD(b2, b1);
				if (w2 == (*(pOne+i)).w )
				{
					strLine = strLine;
					fout.Write((void*) (LPCTSTR) strLine, strLine.GetLength()  );
					if(len == 4) (*(pOne+i)).Num2 ++;
					if(len == 6) (*(pOne+i)).Num3 ++;
					if(len == 8) (*(pOne+i)).Num4 ++;
				}
				strLine.Empty();
			}

			file.Close(); 
		}
		lpti->pProg->SetPos(i); 
	}

	fout.Seek(  sizeof(HEAD_INFOR), CFile::begin);	
	for(i=0;i< WordCount;i++)
	{	
		fout.Write((void*) (pOne+i),  sizeof(INDEX_ONE) );	
	}

	if (pOne)  delete[] pOne;
	fout.Close(); 
	lpti->pProg->SetPos(0); 
	AfxMessageBox("OK");
	return 1;
}


void CMakeLibDlg::OnButtonBwbrowsFile() 
{
	// TODO: Add your control notification handler code here
	CFileDialog dlg(FALSE);
	if( dlg.DoModal() == IDCANCEL ) return ;
	SetDlgItemText(IDC_EDIT_OUT, (LPCTSTR)dlg.GetPathName() );
}

WORD TranCode( WORD w)
{
	BYTE h = HIBYTE(w);
	BYTE l = LOBYTE(w);
	return (WORD) MAKEWORD(h, l);
}


void GetGB2312Count(CStringArray * pArray)
{
	TCHAR szW[3] = {'\0'};
	BYTE i = 0x00, k=0x00;
	BOOL Flag = FALSE;
    for( i = 0xb0; i<0xff; i++)
	{
		for(k=0xa1; k<0xff; k++)
		{
			szW[0] = i;
			szW[1] = k;
			szW[2] ='\0';
			pArray->Add (CString (szW));
			WORD w = MAKEWORD(k, i);
			if( w == 0xf7fe) 
			{
				Flag = TRUE;
			}
		}
		if(Flag == TRUE) break;
	}
}


void CMakeLibDlg::OnButtonRestore() 
{
	CArray  <INDEX_ONE, INDEX_ONE &> arr;

	CFile fw;
	CFile fr;
	fr.Open("d:\\out.dat", CFile::modeRead|CFile::typeBinary); 
	fw.Open("d:\\out.txt", CFile::modeCreate|CFile::modeWrite);

	UINT i=0, x=0, y=0, z=0;
	HEAD_INFOR  head;
	INDEX_ONE   one;
	memset((void*) &head, 0, sizeof(HEAD_INFOR));
	fr.Read((void*) &head, sizeof(HEAD_INFOR));
	for(i=0;i< head.size ; i++)
	{
		memset((void*) &one, 0, sizeof(INDEX_ONE));
		fr.Read((void*) &one, sizeof(INDEX_ONE));
		arr.Add(one); 
	}
	fr.SeekToBegin();


	BYTE b4 [4] = {0};
	BYTE b6 [6] = {0};
	BYTE b8 [8] = {0};
	BYTE kg = 0x20;
	WORD hh= 0x0a0d;

	for( i=0; i< head.size ; i++)
	{
		memset((void*) &one, 0, sizeof(INDEX_ONE));
		one = arr.GetAt(i); 
		fr.Seek(one.FirstOffset2, CFile::begin) ;
		for(x=0;x<one.Num2 ;x++)
		{
			fr.Read((void*) b4, sizeof(BYTE)*4);
			fw.Write((void*) b4, sizeof(BYTE)*4);
			fw.Write((void*) &kg, sizeof(BYTE));
		} 
		fr.Seek(one.FirstOffset3, CFile::begin) ;
		for(y=0;y<one.Num3 ;y++)
		{
			fr.Read((void*) b6, sizeof(BYTE)*6);
			fw.Write((void*) b6, sizeof(BYTE)*6);
			fw.Write((void*) &kg, sizeof(BYTE));
		}

		fr.Seek(one.FirstOffset4, CFile::begin) ;
		for(z=0;z<one.Num4 ;z++)
		{
			fr.Read((void*) b8, sizeof(BYTE)*8);
			fw.Write((void*) b8, sizeof(BYTE)*8);
			fw.Write((void*) &kg, sizeof(BYTE));
		} 
		fw.Write((void*) &hh, sizeof(WORD));
	}

	fw.Close();
	fr.Close();
	AfxMessageBox("OK");
}