mainfrm.cpp

基于SVM的文本分类算法

// MainFrm.cpp : implementation of the CMainFrame class
//
//#include <StdAfx.h>
#include "stdafx.h"
#include "svmcls.h"
#include "svmclsDoc.h"
#include "Message.h"

#include "MainFrm.h"
#include "LeftView.h"
#include "svmclsView.h"

#include "ConvertDlg.h"
#include "SAXFileHandlers.h"
#include "svmparamspage.h"
#include "TrainParamsPage.h"
#include "TestParamsDlg.h"
#include "DocselDlg.h"
#include "classifier.h"
#include "SelectDocs.h"
#include "SegDocDlg.h"
#include "WordSegment.h"
#include <direct.h>

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

/////////////////////////////////////////////////////////////////////////////
// CMainFrame
CString CMainFrame::m_strResultDir;

IMPLEMENT_DYNCREATE(CMainFrame, CFrameWnd)

BEGIN_MESSAGE_MAP(CMainFrame, CFrameWnd)
	//{{AFX_MSG_MAP(CMainFrame)
	ON_WM_CREATE()
	ON_MESSAGE(WM_COMPUTATION_FINISH,OnComputationFinish)
	ON_WM_CLOSE()
	ON_WM_TIMER()
	ON_UPDATE_COMMAND_UI(ID_MAIN_CLASSIFY, OnUpdateMainClassify)
	ON_UPDATE_COMMAND_UI(ID_MAIN_OPENMODEL, OnUpdateMainOpenmodel)
	ON_COMMAND(ID_MAIN_OPENMODEL, OnMainOpenmodel)
	ON_UPDATE_COMMAND_UI(ID_MAIN_STOPRUNNING, OnUpdateMainStoprunning)
	ON_COMMAND(ID_MAIN_STOPRUNNING, OnMainStoprunning)
	ON_UPDATE_COMMAND_UI(ID_MAIN_PAUSE, OnUpdateMainPause)
	ON_COMMAND(ID_MAIN_PAUSE, OnMainPause)
	ON_UPDATE_COMMAND_UI(ID_MAIN_RESUME, OnUpdateMainResume)
	ON_COMMAND(ID_MAIN_RESUME, OnMainResume)
	ON_UPDATE_COMMAND_UI(ID_MAIN_RESULTS, OnUpdateMainResults)
	ON_COMMAND(ID_MAIN_RESULTS, OnMainResults)
	ON_COMMAND(ID_MAIN_TEST, OnMainTest)
	ON_UPDATE_COMMAND_UI(ID_MAIN_TEST, OnUpdateMainTest)
	ON_COMMAND(ID_TOOLS_RESULTS, OnToolsResults)
	ON_UPDATE_COMMAND_UI(ID_TOOLS_RESULTS, OnUpdateToolsResults)
	ON_COMMAND(ID_MAIN_TRAINKNN, OnMainTrainknn)
	ON_UPDATE_COMMAND_UI(ID_MAIN_TRAINKNN, OnUpdateMainTrainknn)
	ON_COMMAND(ID_MAIN_TRAINSVM, OnMainTrainsvm)
	ON_UPDATE_COMMAND_UI(ID_MAIN_TRAINSVM, OnUpdateMainTrainsvm)
	ON_COMMAND(ID_MAIN_CLASSIFY_SINGLEFILE, OnMainClassifySinglefile)
	ON_UPDATE_COMMAND_UI(ID_MAIN_CLASSIFY_SINGLEFILE, OnUpdateMainClassifySinglefile)
	ON_COMMAND(ID_TOOLS_CONVERT, OnToolsConvert)
	ON_COMMAND(ID_TOOLS_SELDOC, OnToolsSeldoc)
	ON_COMMAND(ID_TOOLS_WORDSEG, OnToolsWordseg)
	ON_COMMAND(ID_MAIN_READ_FEATURES, OnMainReadFeatures)
	ON_UPDATE_COMMAND_UI(ID_MAIN_READ_FEATURES, OnUpdateMainReadFeatures)
	//}}AFX_MSG_MAP
END_MESSAGE_MAP()

static UINT indicators[] =
{
	ID_SEPARATOR,           // status line indicator
	ID_INDICATOR_TIME,
};

UINT CMainFrame::TrainThreadProc(LPVOID pParam)
{	
	theClassifier.m_theSVM.com_param.running=1;
	theClassifier.Train(theClassifier.m_paramClassifier.m_nClassifierType);
	theClassifier.m_theSVM.com_param.running=0;
	::PostMessage((HWND)pParam,WM_COMPUTATION_FINISH,0,0);
	return 0;
}

UINT CMainFrame::TestThreadProc(LPVOID pParam)
{
	if(theClassifier.m_paramClassifier.m_nDocFormat==CClassifierParam::nDF_Directory)
	{
		if(theClassifier.Classify())
			CMainFrame::m_strResultDir = theClassifier.m_paramClassifier.m_strResultDir;
	}
	else if(theClassifier.m_paramClassifier.m_nDocFormat==CClassifierParam::nDF_Smart)
	{
		if(theClassifier.ClassifySmart())
			CMainFrame::m_strResultDir = theClassifier.m_paramClassifier.m_strResultDir;
	}
	::PostMessage((HWND)pParam,WM_COMPUTATION_FINISH,0,0);
	return 0;
}

UINT CMainFrame::ConvertThreadProc(LPVOID pParam)
{
	CTime startTime;
	CTimeSpan totalTime;

	CMessage::PrintInfo(_T("开始进行文档格式转换，请稍候..."));
	startTime=CTime::GetCurrentTime();
	theSaxFileHandler.Convert();
	totalTime=CTime::GetCurrentTime()-startTime;
	CMessage::PrintInfo(_T("文档格式转换结束，耗时")+totalTime.Format("%H:%M:%S"));
	CMessage::PrintStatusInfo("");

	::PostMessage((HWND)pParam,WM_COMPUTATION_FINISH,0,0);
	ExitThread(0);
	return 0;
}

UINT CMainFrame::SeldocsThreadProc(LPVOID pParam)
{
	CTime startTime;
	CTimeSpan totalTime;
	long lDocNum;
	CString str;

	CMessage::PrintInfo(_T("开始选择训练文档和测试文档，请稍候..."));
	startTime=CTime::GetCurrentTime();
	lDocNum=theSelectDocs.SelectDocs();
	totalTime=CTime::GetCurrentTime()-startTime;
	str.Format("共处理了%d篇文档!",lDocNum);
	CMessage::PrintInfo(str);
	CMessage::PrintInfo(_T("选择文档结束，耗时")+totalTime.Format("%H:%M:%S"));
	CMessage::PrintStatusInfo("");

	::PostMessage((HWND)pParam,WM_COMPUTATION_FINISH,0,0);
	ExitThread(0);
	return 0;
}

UINT CMainFrame::SegDocsThreadProc(LPVOID pParam)
{
	CTime startTime;
	CTimeSpan totalTime;
	
	startTime=CTime::GetCurrentTime();
	CMessage::PrintInfo(_T("分词程序初始化，请稍候..."));	
	if(!g_wordSeg.InitWorgSegment(theApp.m_strPath.GetBuffer(0)))
	{
		CMessage::PrintError(_T("分词程序初始化失败!"));
		return 0;
	}
	g_wordSeg.SetSegSetting(CWordSegment::uPlace);
	CMessage::PrintInfo(_T("正在进行分词，请稍候..."));	

	if(theSegDocsParam.m_bMultiDocs)
	{
		CString strTarget,strSource;
		HANDLE hFinder;
		LPWIN32_FIND_DATA lpFindFileData;	
		lpFindFileData  = new WIN32_FIND_DATA;
		hFinder = ::FindFirstFile(theSegDocsParam.m_strSource+"\\*.*",lpFindFileData );
		while(::FindNextFile(hFinder,lpFindFileData))
		{
			if( !strcmp(lpFindFileData->cFileName,".") || !strcmp(lpFindFileData->cFileName,"..") )
				continue;

			if(!(lpFindFileData->dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY))
			{
				strTarget=theSegDocsParam.m_strTarget+"\\";
				strTarget+=lpFindFileData->cFileName;
				strSource=theSegDocsParam.m_strSource+"\\";
				strSource+=lpFindFileData->cFileName;
				g_wordSeg.SegmentFile(strSource.GetBuffer(0),
									strTarget.GetBuffer(0),
									theSegDocsParam.m_bDelStopwords);
				CMessage::PrintStatusInfo(lpFindFileData->cFileName);
			}
		}
		delete	lpFindFileData;
	}
	else
		g_wordSeg.SegmentFile(theSegDocsParam.m_strSource.GetBuffer(0),
							theSegDocsParam.m_strTarget.GetBuffer(0),
							theSegDocsParam.m_bDelStopwords);

	g_wordSeg.FreeWordSegment();
	totalTime=CTime::GetCurrentTime()-startTime;
	CMessage::PrintInfo(_T("分词结束，耗时")+totalTime.Format("%H:%M:%S"));
	CMessage::PrintStatusInfo("");

	::PostMessage((HWND)pParam,WM_COMPUTATION_FINISH,0,0);
	ExitThread(0);
	return 0;
}
/////////////////////////////////////////////////////////////////////////////
// CMainFrame construction/destruction

CMainFrame::CMainFrame()
{
	m_pThread=NULL;
	m_strModel.Empty();
	m_strResultDir.Empty();
	m_nRunning=0;
	m_bPaused=false;
	m_nClassifierType=0;
}

CMainFrame::~CMainFrame()
{
}

int CMainFrame::OnCreate(LPCREATESTRUCT lpCreateStruct)
{
	if (CFrameWnd::OnCreate(lpCreateStruct) == -1)
		return -1;
	
	if (!m_wndToolBar.CreateEx(this, TBSTYLE_FLAT, WS_CHILD | WS_VISIBLE | CBRS_TOP
		| CBRS_GRIPPER | CBRS_TOOLTIPS | CBRS_FLYBY | CBRS_SIZE_DYNAMIC) ||
		!m_wndToolBar.LoadToolBar(IDR_MAINFRAME))
	{
		TRACE0("Failed to create toolbar\n");
		return -1;      // fail to create
	}

	if (!m_wndStatusBar.Create(this) ||
		!m_wndStatusBar.SetIndicators(indicators,
		  sizeof(indicators)/sizeof(UINT)))
	{
		TRACE0("Failed to create status bar\n");
		return -1;      // fail to create
	}

	CMessage::SetStatusWnd(&m_wndStatusBar);
	// TODO: Delete these three lines if you don't want the toolbar to
	//  be dockable
	m_wndToolBar.EnableDocking(CBRS_ALIGN_ANY);
	EnableDocking(CBRS_ALIGN_ANY);
	DockControlBar(&m_wndToolBar);

	return 0;
}

BOOL CMainFrame::OnCreateClient(LPCREATESTRUCT /*lpcs*/,
	CCreateContext* pContext)
{
	// create splitter window
	if (!m_wndSplitter.CreateStatic(this, 1, 2))
		return FALSE;

	if (!m_wndSplitter.CreateView(0, 0, RUNTIME_CLASS(CLeftView), CSize(150, 150), pContext) ||
		!m_wndSplitter.CreateView(0, 1, RUNTIME_CLASS(CSvmclsView), CSize(150, 150), pContext))
	{
		m_wndSplitter.DestroyWindow();
		return FALSE;
	}
	CMessage::SetOutputWnd((CWnd *)m_wndSplitter.GetPane(0, 1));
	return TRUE;
}

BOOL CMainFrame::PreCreateWindow(CREATESTRUCT& cs)
{
	cs.style&=~(LONG)FWS_ADDTOTITLE;
	if( !CFrameWnd::PreCreateWindow(cs) )
		return FALSE;
	// TODO: Modify the Window class or styles here by modifying
	//  the CREATESTRUCT cs

	return TRUE;
}

/////////////////////////////////////////////////////////////////////////////
// CMainFrame diagnostics

#ifdef _DEBUG
void CMainFrame::AssertValid() const
{
	CFrameWnd::AssertValid();
}

void CMainFrame::Dump(CDumpContext& dc) const
{
	CFrameWnd::Dump(dc);
}

#endif //_DEBUG

/////////////////////////////////////////////////////////////////////////////
// CMainFrame message handlers

CSvmclsView* CMainFrame::GetRightPane()
{
	CWnd* pWnd = m_wndSplitter.GetPane(0, 1);
	CSvmclsView* pView = DYNAMIC_DOWNCAST(CSvmclsView, pWnd);
	return pView;
}

CLeftView* CMainFrame::GetLeftView()
{
	CWnd* pWnd = m_wndSplitter.GetPane(0, 0);
	CLeftView* pView = DYNAMIC_DOWNCAST(CLeftView, pWnd);
	return pView;
}

void CMainFrame::OnUpdateViewStyles(CCmdUI* pCmdUI)
{
	// TODO: customize or extend this code to handle choices on the
	// View menu.

	CSvmclsView* pView = GetRightPane(); 

	// if the right-hand pane hasn't been created or isn't a view,
	// disable commands in our range

	if (pView == NULL)
		pCmdUI->Enable(FALSE);
	else
	{
		DWORD dwStyle = pView->GetStyle() & LVS_TYPEMASK;

		// if the command is ID_VIEW_LINEUP, only enable command
		// when we're in LVS_ICON or LVS_SMALLICON mode

		if (pCmdUI->m_nID == ID_VIEW_LINEUP)
		{
			if (dwStyle == LVS_ICON || dwStyle == LVS_SMALLICON)
				pCmdUI->Enable();
			else
				pCmdUI->Enable(FALSE);
		}
		else
		{
			// otherwise, use dots to reflect the style of the view
			pCmdUI->Enable();
			BOOL bChecked = FALSE;

			switch (pCmdUI->m_nID)
			{
			case ID_VIEW_DETAILS:
				bChecked = (dwStyle == LVS_REPORT);
				break;

			case ID_VIEW_SMALLICON:
				bChecked = (dwStyle == LVS_SMALLICON);
				break;

			case ID_VIEW_LARGEICON:
				bChecked = (dwStyle == LVS_ICON);
				break;

			case ID_VIEW_LIST:
				bChecked = (dwStyle == LVS_LIST);
				break;

			default:
				bChecked = FALSE;
				break;
			}

			pCmdUI->SetRadio(bChecked ? 1 : 0);
		}
	}
}


void CMainFrame::OnComputationFinish(WPARAM wParam,LPARAM lParam)
{
	CeaseComputation();
}

void CMainFrame::CeaseComputation()
{	
	m_wndStatusBar.SetPaneText(0,"运算全部结束!");
	m_nRunning=0;
	m_pThread=NULL;
	m_bPaused=false;
	KillTimer(m_nTimer);
}

void CMainFrame::BeginComputation()
{
	m_nRunning=1;
	m_nTimer=SetTimer(1001,500,NULL);
	m_tmBegin=CTime::GetCurrentTime();
}

void CMainFrame::OnTimer(UINT nIDEvent) 
{
	CTimeSpan t=CTime::GetCurrentTime()-m_tmBegin;
	m_wndStatusBar.SetPaneText(1,t.Format("耗时%H:%M:%S"),TRUE);
	CFrameWnd::OnTimer(nIDEvent);
}

void CMainFrame::OnClose() 
{
	GetActiveDocument()->SetModifiedFlag(false);
	if(m_nRunning>0)
	{
		if(AfxMessageBox("正在进行计算，真的要停止吗?",MB_YESNO)==IDYES)
		{
			StopComputation();
			CFrameWnd::OnClose();
		}
	}
	else
		CFrameWnd::OnClose();
}

void CMainFrame::Train()
{
	CMessage::ClearInfo();
	BeginComputation();
	m_pThread=AfxBeginThread(TrainThreadProc,GetSafeHwnd(),THREAD_PRIORITY_NORMAL);
}

void CMainFrame::Test() 
{
	CMessage::ClearInfo();
	BeginComputation();
	m_pThread=AfxBeginThread(TestThreadProc,GetSafeHwnd(),THREAD_PRIORITY_NORMAL);
}

void CMainFrame::OnMainClassifySinglefile() 
{
	static char BASED_CODE szTextFilter[] = "Text Files (*.txt)|*.txt|All Files (*.*)|*.*||";
	CFileDialog cfd(TRUE,szTextFilter,NULL,OFN_HIDEREADONLY | OFN_OVERWRITEPROMPT,szTextFilter,NULL);
	if (cfd.DoModal()==IDOK) 
	{
		CString strFile=cfd.GetPathName();
		short id;
		theClassifier.m_paramClassifier.m_dThreshold     =   60;
		theClassifier.m_paramClassifier.m_nKNN           =   35;
		theClassifier.m_paramClassifier.m_nClassifyType  =   CClassifierParam::nFT_Single;
		theClassifier.m_paramClassifier.m_bEvaluation    =   FALSE;
		theClassifier.m_paramClassifier.m_bCopyFiles     =   FALSE;