test.cpp

对中科院分词程序的简化版本

// Test.cpp : Defines the entry point for the console application.
//

// test.cpp : Defines the entry point for the console application.
//

#include "stdafx.h"
#include "..\\Word.h"
#include <stdio.h>
#include <string.h>
#pragma comment(lib, "..\\debug\\Word.lib")
#define MAX_CHAR 35500
#define MAX_PATH 256

int findStr(char *array,int len,char des)
{
	for(int i=0;i<len;i++)
	{
		if(array[i]==des) return i;		
	}
	return -1;
}

void deleteStr(char *array,int len,int a)
{
	for(int i=a;i<len;i++)
	{
		array[i]=array[i+1];
	}
}

int main(int argc, char* argv[])
{	
	char *m_result;
	bool m_bDisable=IsDataExists();
	if(!m_bDisable)
	{
		printf("No Data Dictionay!\n");
		return -1;
	}

	//读取文本
	char content[MAX_CHAR]="";
	char    sSourceFile[MAX_PATH]=".\\word.txt";
	FILE *fpSource;
	char *sParagraph;
	
	sParagraph=new char[4*1024];

    if((fpSource=fopen(sSourceFile,"rt"))==NULL)
	{
		printf("No file!\n");
		return -1;  
	}
	while(!feof(fpSource))
	{
		if(fgets(sParagraph,4*1024,fpSource)==0)
			continue;	
		strcat(content,sParagraph);	
	}
	fclose(fpSource);
	//读完
	//清除回车
	int i=findStr(content,strlen(content),10);
	while(i!=-1)
	{
		deleteStr(content,strlen(content),i);
		i=findStr(content,strlen(content),10);
	}	
	//分词	

	m_result=RunFenci(content);
	printf("%s\n",m_result);
	
	return 0;
}