directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Emmanuel Lecharny <elecha...@gmail.com>
Subject Re: [ASN.1] Thoughts about ASN.1 code
Date Fri, 21 Jan 2011 18:49:36 GMT
On 1/21/11 7:56 AM, Alex Karasulu wrote:
> ASN1 BER =>  shared-asn1-ber
>     o again has some old remnant code that needs cleanup
I checked all the classes in shared-asn1-ber, and I don't see any that 
is not necessary. May be you have some tool that prove me wrong though, 
and then, we will have to remove the spurious classes.

>     o some interfaces are here that can be exposed as SPI
The interface we have in this module are :
- Action
- Grammar
- States
- TLVBerDecoderMBean (probably useless at this point)
- Asn1Container

Let's check what has to be exposed here.
>     o some stuff seems reusable regarding of encoding used like Value etc
Value is really depending on the encoding. It's for BER/DER only
>     o need to check dependencies, how much are we exposing and using
> from outside how much can we actually hide here
Sure. Those classes are only necessary for those who want to implement a 
codec, and who know how to do that. Ie, atm, Kiran and me :)

It would be great to document this part...

>     o BER/DER/CER flag for changing mode of operation?
Not necessary at this point, as we only use BER and DER, and they are 
not really different. For Kerberos and LDAP, what we have is enough (in 
fact, LDAP is not even using the full BER spec, it extends it a bit).


-- 
Regards,
Cordialement,
Emmanuel L├ęcharny
www.iktek.com


Mime
View raw message