i think Lenny probably meant he was checking first.
but while we're on it, your code isn't much better because you're not
checking if the character is a letter. maybe both you and Lenny are
assuming a check first to make sure the character is appropriate to modify;
in which case it actually generates more code with less efficient execution
to do the masking on letters that are already uppercase as you suggest. as
long as Lenny checks to make sure he's only modifying lowercase characters,
his code is better.
-----Original Message-----
From: Kerry W. Podolsky <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
Cc: Lenny Palozzi <[EMAIL PROTECTED]>
Date: Wednesday, March 24, 1999 5:53 AM
Subject: Re: StrToUpper() ?
>The code you posted is Extremely bad.
>
>You are blindly subtracting 32 from every character.
>What if the character is ALREADY upper case.
>
>i.e. Letter 'A' = 0x41 (65) - 0x20(32) = 0x21(33) or '!'
>
>What you want to do is "mask" the bit off.
>
>i.e. *c &= 0xdf
>
>this way if 'a'(0x61) ANDed with 0x20 = 'A'(0x41)
>
>Kerry
>
>
>At 11:25 PM 3/23/99 -0500, Lenny wrote:
>>Is there not a StrToUpper API function? I could only find a StrToLower().
>>
>>I rolled my own, running through each character doing a "*c -= 32;" for
the
>>time being.
