directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alex Karasulu <>
Subject Re: Some more thoughts about the Dn class
Date Thu, 17 Feb 2011 10:59:27 GMT
On Thu, Feb 17, 2011 at 12:11 PM, Emmanuel Lecharny <> wrote:
> On 2/17/11 11:07 AM, Stefan Seelmann wrote:
>> On Thu, Feb 17, 2011 at 10:46 AM, Emmanuel Lecharny<>
>>  wrote:
>>> On 2/17/11 10:28 AM, Stefan Seelmann wrote:
>>>> One concern: What is the use case of the getAncestorOf/getDescendantOf
>>>> from the user's point of view? In the server and in studio such
>>>> methods are useful (e.g. calculating the local name for ACI handling).
>>>> But does the end user of the API really need those methods? I mean we
>>>> should avoid to put methods into the Dn class that is only/mostly used
>>>> in server and studio. Maybe a DnUtils class with static methods for
>>>> advanced DN processing would help?
>>> Good point. I can see some case where you need to split the DN in two
>>> parts,
>>> for instance if you access two servers looking for an entry for which you
>>> have the name. For instance, if you have two ldap server, one in US and
>>> one
>>> in FR, you may have an entry present as "cn=John Doe, dc=acme, c=us" or
>>> "cn=John Doe, dc=acme, c=fr", you may want to get the ancestor to know in
>>> which server the entry is stored into.
>> Hm, in that case you need to know the exact left-part (in that case
>> "cn=John Doe, dc=acme") in order to call the getAnchestor() method,
>> right? In that case a getAnchestor(1) would be more useful.
> Yes, but we don't want to use the indexes for such operation, because users
> *never* now if the index is from the left or from the right. Add to that the
> fact that you also have to remember if it's a 0 based or 1 based index ...

Yeah but unfortunately there are going to be times when you just need
to slice and dice a DN without knowing it's contents. I'm afraid we're
going to need something along this line.


View raw message