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 = ;
 int charnum = ;
 int linenum = ;

 struct wordsdata                //存放单词信息
 {
     char words[];           //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) < ) && (wordcmp(p1.two, p2.two) < ));
     }

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

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

         return size_t(__h);
     }

 };

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

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

     getfiles(path, &fileinfo, handle);

     gettop(fourletter);

     system("pause");
     return ;
 }

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

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

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

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

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

                 _findclose(temphandle);
             }//递归完毕

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

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

     _findclose(handle);       //关闭句柄

     return ;

 }

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

     char present[] = "";  //存储当前单词

     char address[] = "";
     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!!! \n", path);
         return ;
     }
     linenum++;
     while (temp != -)
     {
         //读取字符串
         temp = fgetc(fp);
         if (temp >  && temp < )
             charnum++;
         if (temp == '\n' || temp == '\r')
             linenum++;

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

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

             //计算首四位代表地址
             num = address[] *  + address[] *  + address[] *  + address[];

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

                 else
                 {
                     neword = new wordsdata;
                     neword->number = ;
                     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, ));
                 }
             }
             previous = now;

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

     fclose(fp);
     return ;
 }

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

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

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

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

         q1 = *p1;
         q2 = *p2;

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

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

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

     if (*p1 ==  && *p2 == )           //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 ;

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

     return ;
 }

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

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