orgtagdlg.cpp

分词词典软件

// OrgTagDlg.cpp : implementation file
//

#include "stdafx.h"
#include "OrgTag.h"
#include "OrgTagDlg.h"

#include "ICTCLAS.h"
#include "PreProcess.h"

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

/////////////////////////////////////////////////////////////////////////////
// CAboutDlg dialog used for App About

class CAboutDlg : public CDialog
{
public:
	CAboutDlg();

// Dialog Data
	//{{AFX_DATA(CAboutDlg)
	enum { IDD = IDD_ABOUTBOX };
	//}}AFX_DATA

	// ClassWizard generated virtual function overrides
	//{{AFX_VIRTUAL(CAboutDlg)
	protected:
	virtual void DoDataExchange(CDataExchange* pDX);    // DDX/DDV support
	//}}AFX_VIRTUAL

// Implementation
protected:
	//{{AFX_MSG(CAboutDlg)
	//}}AFX_MSG
	DECLARE_MESSAGE_MAP()
};

CAboutDlg::CAboutDlg() : CDialog(CAboutDlg::IDD)
{
	//{{AFX_DATA_INIT(CAboutDlg)
	//}}AFX_DATA_INIT
}

void CAboutDlg::DoDataExchange(CDataExchange* pDX)
{
	CDialog::DoDataExchange(pDX);
	//{{AFX_DATA_MAP(CAboutDlg)
	//}}AFX_DATA_MAP
}

BEGIN_MESSAGE_MAP(CAboutDlg, CDialog)
	//{{AFX_MSG_MAP(CAboutDlg)
		// No message handlers
	//}}AFX_MSG_MAP
END_MESSAGE_MAP()

/////////////////////////////////////////////////////////////////////////////
// COrgTagDlg dialog

COrgTagDlg::COrgTagDlg(CWnd* pParent /*=NULL*/)
	: CDialog(COrgTagDlg::IDD, pParent)
{
	//{{AFX_DATA_INIT(COrgTagDlg)
	m_label = _T("");
	//}}AFX_DATA_INIT
	// Note that LoadIcon does not require a subsequent DestroyIcon in Win32
	m_hIcon = AfxGetApp()->LoadIcon(IDR_MAINFRAME);
}

void COrgTagDlg::DoDataExchange(CDataExchange* pDX)
{
	CDialog::DoDataExchange(pDX);
	//{{AFX_DATA_MAP(COrgTagDlg)
	DDX_Control(pDX, IDC_RICHEDIT_TEXT, m_edit);
	DDX_Control(pDX, IDC_PROGRESS_TAG, m_progress);
	DDX_Text(pDX, IDC_STATIC1, m_label);
	//}}AFX_DATA_MAP
}

BEGIN_MESSAGE_MAP(COrgTagDlg, CDialog)
	//{{AFX_MSG_MAP(COrgTagDlg)
	ON_WM_SYSCOMMAND()
	ON_WM_PAINT()
	ON_WM_QUERYDRAGICON()
	ON_BN_CLICKED(IDC_BUTTON_IMPORT, OnButtonImport)
	ON_BN_CLICKED(IDC_BUTTON_TAG, OnButtonTag)
	ON_BN_CLICKED(IDC_BUTTON_CLOSE, OnButtonClose)
	ON_COMMAND(ID_MENUITEM_ABOUT, OnMenuitemAbout)
	ON_COMMAND(ID_MENUITEM_ANNOTATOR, OnMenuitemAnnotator)
	ON_COMMAND(ID_MENUITEM_TRAIN, OnMenuitemTrain)
	//}}AFX_MSG_MAP
END_MESSAGE_MAP()

/////////////////////////////////////////////////////////////////////////////
// COrgTagDlg message handlers

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

	// Add "About..." menu item to system menu.

	// IDM_ABOUTBOX must be in the system command range.
	ASSERT((IDM_ABOUTBOX & 0xFFF0) == IDM_ABOUTBOX);
	ASSERT(IDM_ABOUTBOX < 0xF000);

	CMenu* pSysMenu = GetSystemMenu(FALSE);
	if (pSysMenu != NULL)
	{
		CString strAboutMenu;
		strAboutMenu.LoadString(IDS_ABOUTBOX);
		if (!strAboutMenu.IsEmpty())
		{
			pSysMenu->AppendMenu(MF_SEPARATOR);
			pSysMenu->AppendMenu(MF_STRING, IDM_ABOUTBOX, strAboutMenu);
		}
	}

	// 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
	m_operation = ANNOTATOR;

	m_nList = 0;

	m_senList = new CSentenceList[10];
	m_annotator = new CAnnotator(&m_progress);
	
	return TRUE;  // return TRUE  unless you set the focus to a control
}

void COrgTagDlg::OnSysCommand(UINT nID, LPARAM lParam)
{
	if ((nID & 0xFFF0) == IDM_ABOUTBOX)
	{
		CAboutDlg dlgAbout;
		dlgAbout.DoModal();
	}
	else
	{
		CDialog::OnSysCommand(nID, lParam);
	}
}

// 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 COrgTagDlg::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 COrgTagDlg::OnQueryDragIcon()
{
	return (HCURSOR) m_hIcon;
}

void COrgTagDlg::OnButtonImport() 
{
	// TODO: Add your control notification handler code here
	m_edit.SetSel(0, -1);
	m_edit.ReplaceSel("");
	m_strInput = "";

	CFileDialog fileDialog(TRUE, "*.txt", "",
		OFN_HIDEREADONLY | OFN_PATHMUSTEXIST | OFN_OVERWRITEPROMPT,
		"Text Files(*.txt)|*.txt|", NULL);

//	if(m_operation == TRAIN)
//		fileDialog.m_ofn.Flags |= OFN_EXPLORER|OFN_ALLOWMULTISELECT;
	
	m_nList = 0;

	POSITION pos;

	if(fileDialog.DoModal() == IDOK )
	{
		pos = fileDialog.GetStartPosition();
		while(pos)
		{
			m_fileName = fileDialog.GetNextPathName(pos);

			if(m_operation == TRAIN)
			{
				FileToSenList(m_senList[m_nList], m_fileName);
				m_strInput += SenListToString(m_senList[m_nList++]) + "\r\n\r\n";
			}
			else
			{
/*				CFile file;

				if(!file.Open(m_fileName,CFile::modeRead))
				{
					AfxMessageBox("Fail to open the file!");
					return;
				}
				
				BYTE buffer[4096];
				DWORD dwRead;
				while(dwRead = file.Read(buffer, 4096))
					for(unsigned int i=0; i<dwRead; i++)
					
						m_strInput += buffer[i];
				
				file.Close();*/

				char* oldFile = new char[strlen(m_fileName)];
				char* newFile = new char[strlen(m_fileName) + 4];
				sprintf(oldFile, "%s", m_fileName);
				CString m_newFileName = m_fileName;
				m_newFileName.Replace(".txt", "_org.txt");
				sprintf(newFile, "%s", m_newFileName);

				CFileFind find;
				if (!find.FindFile(newFile))
				{
					ICTCLAS_Init();
					ICTCLAS_FileProcess(oldFile, newFile);
					ICTCLAS_Exit();
				}
			
				FileToSenList(m_senList[0], m_newFileName);

				m_strInput += SenListToString(m_senList[0]);

				m_nList ++ ;
			}

			m_edit.SetWindowText(m_strInput);
		}

		if(m_operation == TRAIN)
			this->DisplayResult(m_senList);
	}
}

void COrgTagDlg::OnButtonTag() 
{
	// TODO: Add your control notification handler code here
	CSentenceList senList;
	m_progress.SetPos(0);

	if(m_operation == ANNOTATOR)
	{
//		m_label = "分词中……";
//		UpdateData(false);

		m_label = "标注中……";
		UpdateData(false);
		// annotate
		// waiting to be modified by Xu-YuShi
		// m_senList[0] := CSentenceList
		// &m_progress := pointer to CProgressCtrl
		//CAnnotator anno(&m_progress);
		m_senList[0] = m_annotator->Annotate(m_senList[0]);
		
		m_label = "更新显示中……";
		UpdateData(false);
		this->DisplayResult(m_senList);

		m_label = "完成";
		UpdateData(false);
	}

	else // TRAIN
	{
		// waiting to be modifid by Xu-YuShi
		// m_senList[cnt] := CSentenceList
		// &m_progress := pointer to CProgressCtrl

		CAnnotator anno(&m_progress);
		for(int cnt = 0; cnt < m_nList; cnt ++ )
		{
			anno.Train(m_senList[cnt]);
		}
	}
}

void COrgTagDlg::OnButtonClose() 
{
	// TODO: Add your control notification handler code here
	CDialog::OnCancel();
}

void COrgTagDlg::OnMenuitemAbout() 
{
	// TODO: Add your command handler code here
	CAboutDlg* about = new CAboutDlg();
	about->DoModal();
}

void COrgTagDlg::OnMenuitemAnnotator() 
{
	// TODO: Add your command handler code here
	m_operation = ANNOTATOR;
	
	CheckMenuItem(this->GetMenu()->m_hMenu, ID_MENUITEM_TRAIN, MF_UNCHECKED);
	CheckMenuItem(this->GetMenu()->m_hMenu, ID_MENUITEM_ANNOTATOR, MF_CHECKED);

	((CButton*)GetDlgItem(IDC_BUTTON_TAG))->SetWindowText("Tag");
}

void COrgTagDlg::OnMenuitemTrain() 
{
	// TODO: Add your command handler code here
	m_operation = TRAIN;

	CheckMenuItem(this->GetMenu()->m_hMenu, ID_MENUITEM_ANNOTATOR, MF_UNCHECKED);
	CheckMenuItem(this->GetMenu()->m_hMenu, ID_MENUITEM_TRAIN, MF_CHECKED);

	((CButton*)GetDlgItem(IDC_BUTTON_TAG))->SetWindowText("Train");
}

void COrgTagDlg::DisplayResult(CSentenceList* pSenList)
{
	CHARFORMAT2  cf;  
	m_edit.GetSelectionCharFormat(cf);  
	cf.cbSize=sizeof(cf);  
	cf.dwMask=CFM_BACKCOLOR;  
	cf.crBackColor  =  255;
	cf.dwEffects  =  0;  

	POSITION p, p2;

	CSentence sentence;
	CWord word;

	int cnt = 0;
	for(int i = 0; i < m_nList; i++)
	{
		p = pSenList[i].GetHeadPosition();
		for(int k = 0; k < pSenList[i].GetCount(); k++)
		{
			sentence.RemoveAll();
			sentence = pSenList[i].GetNext(p);
			p2 = sentence.GetHeadPosition();
			for(int k2 = 0; k2 < sentence.GetCount(); k2++)
			{
				word = sentence.GetNext(p2);
				int temp = strlen(word.m_word);
				cnt += temp;
				
				if(!word.m_ne)
				{
					m_edit.SetSel(cnt - temp, cnt);
					m_edit.SendMessage(EM_SETCHARFORMAT,SCF_SELECTION,(LPARAM)&cf);  
				}
			}
		}

		cnt += 4;
	}
}


//**************************************************************
CODE PosTranlate(CString pos)
{
	if(pos == "tt") return tt;
	else if(pos == "w") return w;
	else if(pos == "Ag") return Ag;
	else if(pos == "a") return a;
	else if(pos == "ad") return ad;
	else if(pos == "an") return an;
	else if(pos == "b") return b;
	else if(pos == "c") return c;
	else if(pos == "Dg") return Dg;	
	else if(pos == "d") return d;
	else if(pos == "e") return e;
	else if(pos == "f") return f;
	else if(pos == "g") return g;
	else if(pos == "h") return h;
	else if(pos == "i") return i;
	else if(pos == "j") return j;
	else if(pos == "k") return k;
	else if(pos == "l") return l;
	else if(pos == "m") return m;
	else if(pos == "Ng") return Ng;
	else if(pos == "n") return n;
	else if(pos == "nr") return nr;
	else if(pos == "ns") return ns;
	else if(pos == "nt") return nt;
	else if(pos == "nz") return nz;
	else if(pos == "o") return o;
	else if(pos == "p") return p;	
	else if(pos == "q") return q;
	else if(pos == "r") return r;
	else if(pos == "s") return s;
	else if(pos == "Tg") return Tg;
	else if(pos == "t") return t;
	else if(pos == "u") return u;
	else if(pos == "Vg") return Vg;
	else if(pos == "v") return v;
	else if(pos == "vd") return vd;
	else if(pos == "vn") return vn;
	else if(pos == "x") return x;
	else if(pos == "y") return y;
	else return z;
}

//**************************************************************
void FileToSenList(CSentenceList& senList, CString fileName)
{
	CFile file;
	file.Open(fileName, CFile::modeRead);

	char pbuf[1024];
	UINT nBytesRead;
	CString str = "";

	while(nBytesRead = file.Read( pbuf, 1024 ))
	{
		for(unsigned int i=0; i<nBytesRead; i++)
			str += pbuf[i];
	}

	file.Close();

	int length = strlen(str);
	char* input = new char[length];
	sprintf(input, "%s", str);

	senList.RemoveAll();

	CWord word;
	CSentence sentence;

	char* seps = "/";
	int cnt = 0;
	bool flag = false;

	char* w_word;
	char* w_pos;

	char* temp = strtok(input, seps);
	while(temp!=NULL)
	{
		if(cnt == 0)
		{
			cnt = strlen(temp);
			w_word = new char[cnt+1];
			w_word = temp;
			w_word[cnt] = '\0';
		}

		else
		{
			cnt = strlen(temp);
			flag = false;

			int j = 0;

			while((temp[j] >= 'A' && temp[j] <= 'Z')
				|| (temp[j] >= 'a' && temp[j] <= 'z'))
				j++;
			
			w_pos = new char[j];
			w_pos = temp;
			w_pos[j] = '\0';
			word.m_ne = NONNE;

			if(j == 3) 
			{
				if(w_pos[1] == 'y') w_pos[1] = '\0';
				else if(w_pos[2] == 'y') w_pos[2] = '\0';

				word.m_ne = ORG;
			}

			word.m_word.Format("%s", w_word);
			word.m_pos = PosTranlate(w_pos);
			sentence.AddTail(word);
			if(word.m_pos == w)
			{
				senList.AddTail(sentence);
				sentence.RemoveAll();
			}

			j+=2;

			int t1 = 0;
			int t2 = 0;
			if(j<cnt) w_word = new char[cnt-j+1];
			while(j<cnt)
			{
				while(temp[j]==13 || temp[j]==10)
				{
					w_word[t1++] = temp[j++];
					flag = true;
				}
				if(flag)
				{
					if(temp[j] == ' ')
						while(temp[j] == ' ')
							w_word[t1++] = temp[j++];
					else if(temp[j]!=-95)
						flag = false;
				}

				if(flag)
				{
					w_word[t1] = '\0';
					w_pos = "tt";
					word.m_word.Format("%s", w_word);
					word.m_pos = PosTranlate(w_pos);
					word.m_ne = NONNE;
					sentence.AddTail(word);
					flag = false;
				}

				w_word[t2++] = temp[j++];
			}		
			w_word[t2] = '\0';			
		}

		temp = strtok(NULL, seps);
	}
}

//**************************************************************
CString SenListToString(CSentenceList senList)
{
	CString output;

	POSITION p, p2;
	p = senList.GetHeadPosition();

	CSentence sentence;
	CWord word;

	for(int k = 0; k < senList.GetCount(); k++)
	{
		sentence.RemoveAll();
		sentence = senList.GetNext(p);
		p2 = sentence.GetHeadPosition();
		for(int k2 = 0; k2 < sentence.GetCount(); k2++)
		{
			word = sentence.GetNext(p2);
			int temp = strlen(word.m_word);
			
			for(int k3 = 0; k3 < temp; k3++)
				output += word.m_word[k3];
		}
	}

	return output;
}