一天一个CRT函数 _memicmp

已经有两天没有发表这个系列了，我都快被失眠折磨疯了！连续两晚都是凌晨5点才睡着~

今天来看看_memicmp吧，很少使用的家伙！

5._memicmp

1: /****

2: Purpose:

3:     比较两缓冲区指定区间的字符，在比较前会转换为小写（忽略大小写）

4:

5: Entry:

6:        T *pBuffer1, *pBuffer2 - 所要比较的两缓冲区

7:        unsigned int uCnt - 缓冲区大小

8:

9: Exit:

10:        returns <0    if pBuffer1 <    pBuffer2

11:        returns 0    if pBuffer1 ==    pBuffer2

12:        returns >0    if pBuffer1 >    pBuffer2

13:

14: Exceptions:

15:

16: **/

17:

18: template<typename T>

19: inline int tMemICmp(T *pBuffer1, T *pBuffer2, unsigned int uCnt)

20: {

21:     if( uCnt == 0 )

22:         return 0;

23:

24:     while(--uCnt && ToLower(*pBuffer1++) == ToLower(*pBuffer2++) )

25:         ;

26:

27:     return (ToLower(*pBuffer1) - ToLower(*pBuffer2));

28: }

注：ToLower是我自己的转小写函数，同crt的tolower。

那和strcmp有什么关系呢？strcmp是专门针对字符串的函数，需要判断字符串’/0’终止符，而_memicmp是根据缓冲区大小来进行判断的。



测试：

1: int result = 0;

2: volatile tChar first[] = _T("Those Who Will Not Learn from History");

3: volatile tChar second[] = _T("THOSE WHO WILL NOT LEARN FROM their mistakes");

4: 

5: volatile const DWORD dwCount = 1000000;

6:

7: DWORD dwLast = 0;

8:{

9:     CCYPerformance timer(dwLast);

10:     for(DWORD i = 0; i < dwCount; ++i)

11:     {

12:         result = CY_CRT::tMemICmp( first, second, sizeof(first) / sizeof(first[0]) );

13:    }

14: }

15:cout << dwLast << endl;

16:

17: dwLast = 0;

18: {

19:     CCYPerformance timer(dwLast);

20:     for(DWORD i = 0; i < dwCount; ++i)

21:     {

22:          result = _memicmp( (const void *)first, (const void *)second, sizeof(first) / sizeof(first[0]) );

23:   }

24: }

25: cout << dwLast << endl;

结果肯定是_memicmp胜利！相差不大。