一道基础的词法解析题

/***************************************************************************/
/* 计算统计单词个数。 从命令行输入一个文件路径，程序读取文件内容，分析  */
/* 单词个数，并按照单词出现的频率从高到低逐个打印单词和出现的次数。        */
/* 规则：                                                                  */
/* 1、单词以三种符号区分“，”（英文的逗号）, “.”（英文的句点），            */
/*       “ ”（英文空格）；                                                      */
/* 2、注释以“{”（英文左大括号）开始，以“}”（英文右大括号）结束，           */
/*       允许注释嵌套，嵌套规则与C语言中的多行注释的嵌套                      */
/*       规则相同（允许跨行嵌套）。                                           */
/*    如果只有左括号，而一直没有有括号，那么所有内容都认为是注释；     */
/* 3、输出格式：每一行的格式为：“单词 出现次数”，从上倒下，              */
/*       出现次数最多的行排在最上方，对于出现次数相同的单词，                 */
/*       可不排序；                                                             */
/***************************************************************************/

/* 输入输出示例：                                                         

输入文件内容：
fdasfasfdasf.dasfdasfas,fdasfsdafdsafdsa {
fdafdsafaskdj }
fdasfkdsa jfasf{
{{
{{}}}}}}
dk dka, dfaks , fd a, . fd a. 
}}}
a aa a a 
d  d a fdas fdas fa sdf asf a fda sfas f asf asf a d af da ))) | { fda fdas  {

输出：
a 8
d 3
asf 3
fdas 2
fd 2
| 1
sfas 1
sdf 1
fdasfsdafdsafdsa 1
fdasfkdsa 1
fdasfasfdasf 1
fda 1
fa 1
f 1
dka 1
dk 1
dfaks 1
dasfdasfas 1
da 1
af 1
aa 1
))) 1

/************************************************************************/

#include <string>
#include <map>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <vector>
#include <algorithm>

using std::map;
using std::string;
using std::vector;

#define MAX_PATN 260
#define MAX_LEN_OF_ONE_LINE 1200

typedef map<string, int> WordMayType;
WordMayType WordMap;

void DealWithOneLine(char* LineBuffer, bool* HasCommentInTheEnd);
void DealWithTheLineWithNoComments(char* LineBuffer);
void WordMapOutPut();

int main(int argc, char** argv)
{
    if (argc != 2)
    {
        return 0;
    }
    bool HasCommentInTheEnd = false;
    char LineBuffer[MAX_LEN_OF_ONE_LINE] =
    {
        0
    };
    FILE* fpr = fopen(argv[1], "r");
    if (fpr == NULL)
    {
        return 0;
    }

    while (fgets(LineBuffer, MAX_LEN_OF_ONE_LINE, fpr) != NULL)
    {
        DealWithOneLine(LineBuffer, &HasCommentInTheEnd);
    }

    fclose(fpr);

    // 输出map表到屏幕
    WordMapOutPut();
    system("pause");
    return 0;
}

void DealWithOneLine(char* LineBuffer, bool* HasCommentInTheLastLineEnd)
{
    if (NULL == LineBuffer)
    {
        return;
    }
    string LineBufferString(LineBuffer);
    int nCommentBeginPos = (int) LineBufferString.find("{");
    int nCommentEndPos = (int) LineBufferString.find("}");
    if (nCommentEndPos == -1 &&
        nCommentBeginPos == -1 &&
        !*HasCommentInTheLastLineEnd)
    {
        // ,
        // ,
        //那么本行全部有效
        DealWithTheLineWithNoComments(LineBuffer);
        return;
    }
    else if (nCommentEndPos == -1 &&
        nCommentBeginPos == -1 &&
        *HasCommentInTheLastLineEnd)
    {
        //{,
        // ,
        //但本行没有任何注释符，那么退出，处理下一行。
        return;
    }
    else if (nCommentEndPos == -1 &&
        nCommentBeginPos != -1 &&
        !*HasCommentInTheLastLineEnd)
    {
        // ,
        //{,
        //需要分两段来处理
        //处理前半段
        LineBuffer[nCommentBeginPos] = '\0';
        DealWithTheLineWithNoComments(LineBuffer);
        //处理后半段
        *HasCommentInTheLastLineEnd = true;
        DealWithOneLine(LineBuffer + nCommentEndPos + 1,
            HasCommentInTheLastLineEnd);
    }
    else if (nCommentEndPos == -1 &&
        nCommentBeginPos != -1 &&
        *HasCommentInTheLastLineEnd)
    {
        //{,
        //{, 
        //只递归处理后半段。
        *HasCommentInTheLastLineEnd = true;
        DealWithOneLine(LineBuffer + nCommentBeginPos + 1,
            HasCommentInTheLastLineEnd);
    }
    else if (nCommentEndPos != -1 &&
        nCommentBeginPos == -1 &&
        !*HasCommentInTheLastLineEnd)
    {
        // ,
        //},
        //将饭括号去掉后继续递归处理
        *HasCommentInTheLastLineEnd = false;
        LineBuffer[nCommentEndPos] = ' ';
        DealWithOneLine(LineBuffer, HasCommentInTheLastLineEnd);
    }
    else if (nCommentEndPos != -1 &&
        nCommentBeginPos == -1 &&
        *HasCommentInTheLastLineEnd)
    {
        //{,
        //},
        //只处理后半段
        *HasCommentInTheLastLineEnd = false;
        DealWithOneLine(LineBuffer + nCommentEndPos + 1,
            HasCommentInTheLastLineEnd);
    }
    else if (nCommentEndPos != -1 &&
        nCommentBeginPos != -1 &&
        !*HasCommentInTheLastLineEnd)
    {
        // ,
        //但本行两个注释符都存在
        //则分两种情况
        if (nCommentBeginPos < nCommentEndPos)
        {
            // ,
            //{}
            //处理前半段
            LineBuffer[nCommentBeginPos] = '\0';
            DealWithTheLineWithNoComments(LineBuffer);
            //处理后半段
            *HasCommentInTheLastLineEnd = false;
            DealWithOneLine(LineBuffer + nCommentEndPos + 1,
                HasCommentInTheLastLineEnd);
        }
        else
        {
            // ,
            //}{,
            //去掉无效字符，递归处理
            LineBuffer[nCommentEndPos] = ' ';
            DealWithOneLine(LineBuffer, HasCommentInTheLastLineEnd);
        }
    }
    else if (nCommentEndPos != -1 &&
        nCommentBeginPos != -1 &&
        *HasCommentInTheLastLineEnd)
    {
        //{,
        //但本行两个注释符都存在
        //则分两种情况
        if (nCommentBeginPos < nCommentEndPos)
        {
            //{,
            //{}
            //处理后半段
            *HasCommentInTheLastLineEnd = false;
            DealWithOneLine(LineBuffer + nCommentEndPos + 1,
                HasCommentInTheLastLineEnd);
        }
        else
        {
            //{,
            //}{
            //只处理后半段
            *HasCommentInTheLastLineEnd = true;
            DealWithOneLine(LineBuffer + nCommentEndPos + 1,
                HasCommentInTheLastLineEnd);
        }
    }
}

void DealWithTheLineWithNoComments(char* LineBuffer)
{
    int nLen = strlen(LineBuffer);
    int nWordPosBegin = 0;
    int nWordPosEnd = 0;
    char cLetter = '\0';
    bool IsNewWordFlag = true;
    for (int i = 0; i != nLen; i++)
    {
        cLetter = LineBuffer[i];
        if ((cLetter != ' ') &&
            (cLetter != ',') &&
            (cLetter != 0xA) &&
            (cLetter != '.'))
        {
            if (IsNewWordFlag)
            {
                nWordPosBegin = i;
                IsNewWordFlag = false;
            }
        }
        else
        {
            if (!IsNewWordFlag)
            {
                nWordPosEnd = i;
                char WordTmp[MAX_LEN_OF_ONE_LINE] =
                {
                    0
                };
                strncpy(WordTmp, LineBuffer + nWordPosBegin,
                    nWordPosEnd - nWordPosBegin);
                WordMap[WordTmp]++;
            }
            IsNewWordFlag = true;
        }
    }
}

void WordMapOutPut()
{
    WordMayType::const_iterator map_it2, map_it;
    map_it2 = map_it = WordMap.begin();
    while (map_it != WordMap.end())
    {
        int nValueTmp = 0;
        for (; map_it != WordMap.end(); map_it++)
        {
            if (nValueTmp <= map_it->second)
            {
                nValueTmp = map_it->second;
                map_it2 = map_it;
            }
        }
        printf("%s %d\n", map_it2->first.c_str(), map_it2->second);
        WordMap.erase(map_it2->first.c_str());
        map_it2 = map_it = WordMap.begin();
    }
}