extractsentencedoc.cpp

本程序用vc编写

// ExtractSentenceDoc.cpp : implementation of the CExtractSentenceDoc class
//

#include "stdafx.h"
#include "ExtractSentence.h"

#include "LWORDSet.h"
#include "ExtractSentenceDoc.h"

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

#define MAXLENGTH 10000

/////////////////////////////////////////////////////////////////////////////
// CExtractSentenceDoc

IMPLEMENT_DYNCREATE(CExtractSentenceDoc, CDocument)

BEGIN_MESSAGE_MAP(CExtractSentenceDoc, CDocument)
	//{{AFX_MSG_MAP(CExtractSentenceDoc)
		// NOTE - the ClassWizard will add and remove mapping macros here.
		//    DO NOT EDIT what you see in these blocks of generated code!
	//}}AFX_MSG_MAP
END_MESSAGE_MAP()

/////////////////////////////////////////////////////////////////////////////
// CExtractSentenceDoc construction/destruction

CExtractSentenceDoc::CExtractSentenceDoc()
{
	// TODO: add one-time construction code here

}

CExtractSentenceDoc::~CExtractSentenceDoc()
{
}

BOOL CExtractSentenceDoc::OnNewDocument()
{
	if (!CDocument::OnNewDocument())
		return FALSE;

	// TODO: add reinitialization code here
	// (SDI documents will reuse this document)

	return TRUE;
}



/////////////////////////////////////////////////////////////////////////////
// CExtractSentenceDoc serialization

void CExtractSentenceDoc::Serialize(CArchive& ar)
{
	if (ar.IsStoring())
	{
		// TODO: add storing code here
	}
	else
	{
		// TODO: add loading code here
	}
}

/////////////////////////////////////////////////////////////////////////////
// CExtractSentenceDoc diagnostics

#ifdef _DEBUG
void CExtractSentenceDoc::AssertValid() const
{
	CDocument::AssertValid();
}

void CExtractSentenceDoc::Dump(CDumpContext& dc) const
{
	CDocument::Dump(dc);
}
#endif //_DEBUG

/////////////////////////////////////////////////////////////////////////////
// CExtractSentenceDoc commands

void CExtractSentenceDoc::splitParagraph(char *paragraph, char *sentence[MAXSENTENCE])
{
	int locateFullStop = 10000, locateQuestionMark = 10000, locateExclamationMark = 10000;
	int locate = 0;
	char *pLocate;
	int i, index = 0;
	int nLen = lstrlen(paragraph);
	char remainSentences[10000];	

	//get the remain sentences from paragraph, all of the operations base on the remainSentences[]
	for (i = 0; i < lstrlen(paragraph); i++)
	{
		*(remainSentences + i) = *(paragraph + i);
	}
	remainSentences[lstrlen(paragraph)] = '\0';

	//if there are no 。？！
	if (!strstr(paragraph, "。") && !strstr(paragraph, "？") && !strstr(paragraph, "！"))
	{
		//put the paragraph into sentence[0] as a single sentence
		sentence[0] = new char[lstrlen(paragraph) + 1];
		strncpy(sentence[0], paragraph, lstrlen(paragraph) + 1);
	}
	else
	{
		index = 0;
		while (*(remainSentences) != '\0')
		{
			//init
			locateFullStop = 10000;
			locateQuestionMark = 10000;
			locateExclamationMark = 10000;

			//get the position of 。？！
		    if ((pLocate = strstr(remainSentences, "。")) != NULL)
			{
				locateFullStop = (int)(pLocate - remainSentences + 1);
			}

			if ((pLocate = strstr(remainSentences, "？")) != NULL)
			{
				locateQuestionMark = (int)(pLocate - remainSentences + 1);
			}

			if ((pLocate = strstr(remainSentences, "！")) != NULL)
			{
				locateExclamationMark = (int)(pLocate - remainSentences + 1);	
			}

			//if there is no 。？！in the current sentences
			if (locateFullStop == 10000 && locateQuestionMark == 10000 && locateExclamationMark == 10000)
			{
				sentence[index] = new char[lstrlen(remainSentences) + 1];
				strncpy(sentence[index], remainSentences, lstrlen(remainSentences));
				sentence[index][lstrlen(remainSentences)] = '\0';
				*(remainSentences) = '\0';
			}
			else
			{
				//find the position of the first sentence in the paragraph
				if (locateFullStop < locateQuestionMark)
					locate = locateFullStop;
				else
					locate = locateQuestionMark;

				if (locateExclamationMark < locate)
					locate = locateExclamationMark;

				//get s sentence in the paragraph 
				sentence[index] = new char[locate + 2];
				strncpy(sentence[index], remainSentences, locate + 1);
				sentence[index][locate + 1] = '\0';
				
				//get the remain sentences
				int j = 0;
				while (remainSentences[j + locate + 1] != '\0')
				{
					remainSentences[j] = remainSentences[j + locate + 1];
					j++;
				}
				remainSentences[j] = '\0';

				index++;
			}
		}
	}
}

bool CExtractSentenceDoc::searchTermsInSentence(char *sen, CString term)
{
	char word[50];
	for (int i = 0; i < term.GetLength(); i++)
	{
		word[i] = term[i];
	}
	word[i] = '\0';

	if (strstr(sen, word) == NULL)
		return false;
	return true;
}