|
Boost : |
From: Daryle Walker (darylew_at_[hidden])
Date: 2001-12-01 14:55:31
on 11/30/01 7:48 PM, David Abrahams at david.abrahams_at_[hidden] wrote:
> I'm showing some interest, but also ignorance: why wouldn't you just do the
> finalization step "on demand", when the checksum is requested? Does that
> invalidate the MD5 process so that it can't be continued?
I came up with a similar idea right when I wrote this message, but I don't
know yet how feasible it is. Maybe I can cache the result if I get the
checksum multiple times.
> ----- Original Message -----
> From: "Daryle Walker" <darylew_at_[hidden]>
> To: "Boost" <boost_at_[hidden]>
> Sent: Friday, November 30, 2001 7:23 PM
> Subject: [boost] Need interface advice
[SNIP]
>> I want to add a similar class that works the MD5 standard. But there's an
>> interface problem. CRC and Adler-32 can process bytes, and the user can
>> check the checksum of what has been processed so far. Unfortunately, MD5
>> requires a finalization step before getting the checksum. I could use an
>> interface like CRC, and just add a finalization method to a MD5 class.
>> However, this will add a state so I can make sure that all byte processing
>> occurs before finalization and the checksum calculation happens after
>> finalization. How should I have the MD5 (class) interface?
-- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT mac DOT com
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk