Well, we pay O(n) to encode/decode, then O(n) again to copy XMLByte* to XMLCh*.  Per 
Peiyong's explanation, I see now that we've strictly implemented the RFC, then added 
helper functions to get us to XMLCh*.  The copying O(n) is small relative to the 
encoding O(n), so I think that it makes more sense to stick to the RFC, than to encode 
directly to/from XMLCh* to XMLByte*, even though we do potentially have to new[] a 
large region of memory an extra time.

Adam Heinz
Development Consultant

Exstream Software
[EMAIL PROTECTED]
317.879.2831

connecting with the eGeneration
www.exstream.com

-----Original Message-----
From: Jeffrey Rodriguez [mailto:[EMAIL PROTECTED]
Sent: Wednesday, September 03, 2003 1:51 AM
To: [EMAIL PROTECTED]
Subject: Re: Base64 efficiency

About efficiency I don't know where you would gain efficiency, could you 
elaborate? You worry about having to treat XMLCh ( a double byte) as a two bytes? But 
you would have to do that anyway because your data could have one or two bytes.

Cheers,

                     Jeff Rodriguez

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to