Word查询1

// word_dict.cpp : 定义控制台应用程序的入口点。
#include "stdafx.h"
#include <assert.h>
#include <ctype.h>
#include <stdio.h>
#include <string.h>
#include <time.h>
#include <string>
#include <vector>

typedef struct
{
	size_t uLineNo;
	size_t uWordSeq;
} term_index_t;

#if defined(_WIN32)
#include <Windows.h>
#include <hash_map>

using namespace std;
using namespace stdext;

typedef hash_map< string, vector<term_index_t> > map_term_index;

#if !defined(_WINSOCK2API_) && !defined(_WINSOCKAPI_)
struct timeval
{
	long tv_sec;
	long tv_usec;
};
#endif

int gettimeofday(struct timeval *tv, void *tzp)
{
	time_t clock;
	struct tm tm;
	SYSTEMTIME wtm;

	GetLocalTime(&wtm);
	tm.tm_year = wtm.wYear - 1900;
	tm.tm_mon  = wtm.wMonth - 1;
	tm.tm_mday = wtm.wDay;
	tm.tm_hour = wtm.wHour;
	tm.tm_min  = wtm.wMinute;
	tm.tm_sec  = wtm.wSecond;
	tm.tm_isdst= -1;
	clock = mktime(&tm);
	tv->tv_sec = (long)clock;
	tv->tv_usec = wtm.wMilliseconds * 1000;

	return 0;
}
#else
#include <sys/time.h>
#include <ext/hash_map>

using namespace std;
using namespace __gnu_cxx;

struct str_hash
{
	size_t operator()(const string &str) const
	{
		return __stl_hash_string(str.c_str());
	}
};

typedef hash_map<string, vector<term_index_t>, str_hash> map_term_index;

#endif

map_term_index g_term_index;

size_t time_stat(bool bBegin = true)
{
	static struct timeval st_time_begin, st_time_end;
	static char szInfo[128];

	if (bBegin)
	{
		gettimeofday(&st_time_begin, NULL);
		return 0;
	}
	else
	{
		gettimeofday(&st_time_end, NULL);
		return (st_time_end.tv_sec - st_time_begin.tv_sec) * 1000
			+ (st_time_end.tv_usec - st_time_begin.tv_usec) / 1000;
	}
}

void Str_Lower(char *pStr)
{
	assert(pStr);

	while (*pStr)
	{
		*pStr++ = tolower(*pStr);
	}
}

// 返回true表示还未读完，否则表示读完
bool GetWord(const char *pBuffer, string &strWord, size_t &uCurPos)
{
	assert(pBuffer);

	strWord = "";

	// 跳过前面的空白以及标点符号等
	for (;;)
	{
		if (!pBuffer[uCurPos])
		{
			return false;
		}
		else if (isalnum(pBuffer[uCurPos]))
		{
			break;
		}
		uCurPos++;
	}

	// 读取单词
	for (;;)
	{
		if (!pBuffer[uCurPos])
		{
			return false;
		}
		else if (isalnum(pBuffer[uCurPos]))
		{
			strWord.append(1, tolower(pBuffer[uCurPos]));
		}
		else
		{
			break;
		}
		uCurPos++;
	}

	return true;
}

int CreateRevIndex(const char *pInFile)
{
	assert(pInFile);

	FILE *pFp = fopen(pInFile, "r");
	if (!pFp)
	{
		printf("Open file:%s failed!\n", pInFile);
		return -1;
	}

	char szInBuffer[4096];
	bool bProcEnd;
	string strWord;
	size_t uProcPos, uLineNo = 0, uWordSeq;
	term_index_t word_index;
	while (fgets(szInBuffer, sizeof(szInBuffer), pFp))
	{
		uProcPos = 0;
		bProcEnd = false;
		uLineNo++;
		uWordSeq = 0;
		while (!bProcEnd)
		{
			bProcEnd = !GetWord(szInBuffer, strWord, uProcPos);
			if (0 == strWord.size())
			{
				continue;
			}
			uWordSeq++;

			word_index.uLineNo = uLineNo;
			word_index.uWordSeq= uWordSeq;

			g_term_index[strWord.c_str()].push_back(word_index);
		}
	}
	fclose(pFp);

	return 0;
}

int main(int argc, char *argv[])
{
	const char *pInFile = "kjv.txt";
	if (argc > 1)
	{
		pInFile = argv[1];
	}

	size_t uInteval = 0;
	time_stat(true);
	if (0 != CreateRevIndex(pInFile))
	{
		printf("Call CreateRevIndex failed!\n");
		return -2;
	}
	uInteval = time_stat(false);
	printf("construct search use: %ums\n", uInteval);

	char szBuffer[512];
	string strOutInfo;
	while (true)
	{
		printf("Please Enter word which you want to search:");
		scanf("%s", szBuffer);
		Str_Lower(szBuffer);

		if (!strncmp("exit", szBuffer, 5))
		{
			printf("exit..., good bye!\n");
			break;
		}
		printf("The word you search is : %s\nSearch result: ", szBuffer);
		time_stat(true);
		map_term_index::iterator iter
			= g_term_index.find(szBuffer);
		if (g_term_index.end() == iter)
		{
			uInteval = time_stat(false);
			printf("Not found!\n");
		}
		else
		{
			strOutInfo = "";
			for (vector<term_index_t>::iterator it = iter->second.begin()
				; it != iter->second.end(); ++it)
			{
				::_snprintf(szBuffer, sizeof(szBuffer), "{%u,%u},", it->uLineNo, it->uWordSeq);
				// printf("{%u,%u},", it->uLineNo, it->uWordSeq);
				strOutInfo += szBuffer;
			}
			uInteval = time_stat(false);
			printf("%s\nTotal num is: %u\n", strOutInfo.c_str(), iter->second.size());
		}
		printf("This search use: %ums\n", uInteval);
	}

	return 0;
}