记一个男默女泪的 BUG

姗姗来迟的词频统计代码 BUG 的发现

1. 此前提交的第一次代码作业总结博客

http://www.cnblogs.com/ustczwq/p/8680704.html

2. BUG 本天成，妙手偶得之

虽然代码已经提交，但总是感觉哪个地方不太对，bug 存在得过于莫名其妙。然后，随手打开代码，稍微调试了一下，当我发现 bug 的时候，不知道该说些什么好，只想讲脏话。

 

出现 bug 的地方：

 

 

改过之后：

 

 

 

看出来了吧，妈卖批，三目运算符没赋值。改完之后，输出结果立马正确。怪不得用 unordered_map 的时候哈希表的查询出问题了，我 TM 定义的哈希函数有问题。虽然迟了，但那种优化是对的，简单补一篇，算是对原博客的完善。

 

3. 加了几个等于号之后的源代码

#include "io.h"
#include "math.h"
#include "stdio.h"
#include "string.h"
#include "stdlib.h"
#include "unordered_map"

using namespace std;
   
#define small 2

int wordnum = 0;
int charnum = 0;
int linenum = 0;

struct wordsdata                //存放单词信息
{
    char words[1024];           //单词字符串
    int number;                 //出现次数
    wordsdata *next;
};
struct phrases
{
    char *one;
    char *two;
    int num;
};

int wordcmp(char *str1, char *str2);
int gettop(struct wordsdata **word);
int getwords(char *path, struct wordsdata **word);
int getfiles(char *path, struct _finddata_t *fileinfo, long handle);

struct phrase_cmp
{
    bool operator()(const phrases &p1, const phrases &p2) const
    {
        return ((wordcmp(p1.one, p2.one) < 2) && (wordcmp(p1.two, p2.two) < 2));
    }

};
struct phrase_hash
{
    size_t operator()(const phrases &ph) const
    {
        unsigned long __h = 0;
        int temp;
        size_t i;
        for (i = 0; ph.one[i]; i++)
        {
            temp = ph.one[i];
            if (temp > 64)
            {
                (temp > 96) ? (temp -= 96) : (temp -= 64);
                __h += (29 * __h + temp);
                __h %= 2147483647;
            }

        }
        for (i = 0; ph.two[i]; i++)
        {
            temp = ph.two[i];
            if (temp > 64)
            {
                (temp > 96) ? (temp -= 96) : (temp -= 64);
                __h += (29 * __h + temp);
                __h %= 2147483647;
            }
        }

        return size_t(__h);
    }

};

typedef unordered_map<phrases, int, phrase_hash, phrase_cmp> Char_Phrase;
Char_Phrase phrasemap;
struct wordsdata *fourletter[26 * 26 * 26 * 26] = {}; //按首四字母排序

int main()
{
    int j = 0;                            
    long handle = 0;                           // 用于查找的句柄 
    struct _finddata_t fileinfo;               // 文件信息的结构体 
    char *path = __argv[1];
    
    getfiles(path, &fileinfo, handle);

    gettop(fourletter);

    system("pause");
    return 1;
}

int getfiles(char *path, struct _finddata_t *fileinfo, long handle)
{                                    
    handle = _findfirst(path, fileinfo);            //第一次打开父目录
    if (handle == -1)
        return -1;


    do
    {
        //printf("> %s
", path);           //显示目录名

        if (fileinfo->attrib & _A_SUBDIR)           //如果读取到子目录
        {
            if (strcmp(fileinfo->name, ".") != 0 && strcmp(fileinfo->name, "..") != 0)
            {
                char temppath[1024] = "";              //记录子目录路径
                long temphandle = 0;
                struct _finddata_t tempfileinfo;
                strcpy(temppath, path);
                strcat(temppath, "/*");

                temphandle = _findfirst(temppath, &tempfileinfo);  //第一次打开子目录
                if (temphandle == -1)
                    return -1;

                do                              //对子目录所有文件递归
                {
                    if (strcmp(tempfileinfo.name, ".") != 0 && strcmp(tempfileinfo.name, "..") != 0)
                    {
                        strcpy(temppath, path);
                        strcat(temppath, "/");
                        strcat(temppath, tempfileinfo.name);
                        getfiles(temppath, &tempfileinfo, temphandle);
                    }
                } while (_findnext(temphandle, &tempfileinfo) != -1);

                _findclose(temphandle);
            }//递归完毕

        } //子目录读取完毕
        else
            getwords(path, fourletter);


    } while (_findnext(handle, fileinfo) != -1);

    _findclose(handle);       //关闭句柄

    return 1;

}

int getwords(char *path, struct wordsdata **word)
{
    FILE *fp;
    int j = 0;
    int cmp = 0;
    int num = 0;               //计算首四位地址
    char temp = 0;             //读取一个字符 ACSII 码值
    int length = 0;

    char present[1024] = "";  //存储当前单词

    char address[4] = "";
    struct wordsdata *q = NULL;
    struct wordsdata *pre = NULL;
    struct wordsdata *neword = NULL;
    struct wordsdata *now = NULL;
    struct wordsdata *previous = NULL;
    struct phrases *newphrase = NULL;

    if ((fp = fopen(path, "r")) == NULL)
    {
        //printf("error!!! 
", path);
        return 0;
    }
    linenum++;
    while (temp != -1)
    {
        //读取字符串
        temp = fgetc(fp);
        if (temp > 31 && temp < 127)
            charnum++;
        if (temp == '
' || temp == '
')
            linenum++;

        while ((temp >= '0' && temp <= '9') || (temp >= 'a' && temp <= 'z') || (temp >= 'A' && temp <= 'Z'))
        {
            if (length != -1 && length < 4)
            {
                if (temp >= 'A')  //是字母
                {
                    present[length] = temp;
                    address[length] = (temp >= 'a' ? (temp - 'a') : (temp - 'A'));
                    length++;
                }
                else            //不是字母
                    length = -1;
            }
            else if (length >= 4)
            {
                present[length] = temp;
                length++;
            }
            temp = fgetc(fp);
            if (temp > 31 && temp < 127)
                charnum++;
            if (temp == '
' || temp == '
')
                linenum++;
        } // end while

          //判断是否为单词
        if (length >= 4)
        {
            wordnum++;

            //计算首四位代表地址
            num = address[0] * 17576 + address[1] * 676 + address[2] * 26 + address[3];

            //插入当前单词
            if (word[num] == NULL)
            {
                word[num] = new wordsdata;
                neword = new wordsdata;
                neword->number = 1;
                neword->next = NULL;
                strcpy(neword->words, present);
                word[num]->next = neword;
                now = neword;
            }
            else
            {
                pre = word[num];
                q = pre->next;
                cmp = wordcmp(q->words, present);

                while (cmp == small)
                {
                    pre = q;
                    q = q->next;
                    if (q != NULL)
                        cmp = wordcmp(q->words, present);
                    else
                        break;
                }
                if (q != NULL && cmp <= 1)
                {
                    now = q;
                    q->number++;
                    if (cmp == 1)
                        strcpy(q->words, present);                
                }

                else
                {
                    neword = new wordsdata;
                    neword->number = 1;
                    strcpy(neword->words, present);
                    pre->next = neword;
                    neword->next = q;
                    now = neword;
                }
            }

            if (previous != NULL)
            {
                newphrase = new phrases;

                newphrase->one = previous->words;
                newphrase->two = now->words;
                
                unordered_map<phrases, int>::const_iterator got = phrasemap.find( *newphrase);
                if (got != phrasemap.end())
                {
                    phrasemap[*newphrase]++;
                }
                else
                {
                    phrasemap.insert(pair<phrases, int>(*newphrase, 1));
                }
            }
            previous = now;

            //当前单词置空
            for (int j = 0; present[j] && j < 1024; j++)
                present[j] = 0;
        }
        length = 0;
    }

    fclose(fp);
    return 1;
}

int wordcmp(char *str1, char *str2)
{
    char *p1 = str1;
    char *p2 = str2;
    char q1 = *p1;
    char q2 = *p2;

    if (q1 >= 'a' && q1 <= 'z')
        q1 -= 32;

    if (q2 >= 'a' && q2 <= 'z')
        q2 -= 32;

    while (q1 && q2 && q1 == q2)
    {
        p1++;
        p2++;

        q1 = *p1;
        q2 = *p2;

        if (q1 >= 'a' && q1 <= 'z')
            q1 -= 32;

        if (q2 >= 'a' && q2 <= 'z')
            q2 -= 32;
    }

    while (*p1 >= '0' && *p1 <= '9')
        p1++;
    while (*p2 >= '0' && *p2 <= '9')
        p2++;

    if (*p1 == 0 && *p2 == 0)           //两单词等价    
        return strcmp(str1, str2);       //等价前者字典顺序小返回-1，大返回1，完全相等返回0

    if (q1 < q2)                   //前者小
        return 2;

    if (q1 > q2)                   //后者小
        return 3;

    return 4;
}

int gettop(struct wordsdata **word)
{
    int i = 0, j = 0;
    struct wordsdata *topw[12] = {};
    struct phrases *toph[12] = {};
    struct wordsdata *w = NULL;
    FILE *fp;
    fp = fopen("result.txt", "w");
    fprintf(fp,"characters:%d 
words:%d 
lines:%d
",  charnum,wordnum, linenum);

    for (j = 0; j < 12; j++)
    {        
        toph[j] = new struct phrases;
        toph[j]->num = 0;
        topw[j] = new struct wordsdata;
        topw[j]->number = 0;
    }
    for (i = 0; i < 456976; i++)
    {
        if (word[i] != NULL)
        {
            w = word[i]->next;
            while (w != NULL)
            {
                topw[11]->number = w->number;
                topw[11]->next = w;
                j = 11;
                while (j > 1 && topw[j]->number > topw[j - 1]->number)
                {
                    topw[0] = topw[j];
                    topw[j] = topw[j - 1];
                    topw[j - 1] = topw[0];
                    j--;
                }
                w = w->next;
            }
        }
    }
    for (j = 1; j < 11; j++)
    {
        if (topw[j]->number)
            fprintf(fp,"
%s :%d", topw[j]->next->words, topw[j]->number);
    }
    for (Char_Phrase::iterator it = phrasemap.begin(); it != phrasemap.end(); it++)
    {
        toph[11]->one = it->first.one;
        toph[11]->two = it->first.two;
        toph[11]->num = it->second;
        j = 11;
        while (j > 1 && toph[j]->num > toph[j - 1]->num)
        {
            toph[0] = toph[j];
            toph[j] = toph[j - 1];
            toph[j - 1] = toph[0];
            j--;
        }
    }
    fprintf(fp, "
");
    for (j = 1; j < 11; j++)
    {
        if (toph[j]->num)
            fprintf(fp,"
%s %s :%d", toph[j]->one, toph[j]->two, toph[j]->num);
    }
    fclose(fp);
    return 1;
}