slapd - dn normalize

["Scott Finnerty" <sfinnert@swbell.net>]

I've been using the LDAP distribution since its days at U. of Mich., but
haven't
been paying close attention to the development efforts to date - so I
apologize if this ends up being a stupid question.

I've got a backend that puts an LDAP front on NIS+ - so when new releases
come out I do the work to upgrade and make any adjustments.  With the latest
release of 1.2 I have run into a problem with the dn normalization...
specifically the case normalization.  When the LDAP operation interfaces in
slapd proper normalize the case of the dn it ignores the case sensitivity of
the attribute values in the dn and therefore breaks things in the backend
(such as matches).

>From what I can tell, the case sensitivity of attribute values is supported
in LDAP - thus the 'ces' and 'cis' distinctions in the attribute definition
file (i.e., slapd.at.conf)...

Perhaps the original, normalized dn (in terms of format with respect to
quotes, commas, whitespace, etc. - what dn_normalize does) is kept somewhere
that the backends can access... (Backend or Operation structs)?

If a particular backend wants to normalize case for efficiency reasons, then
that operation should occur in the backend and not in the core....

Or am I missing something about the nature of dn's and the attributes that
compose them....

I'm currently holding at 1.1.2 until I get a clear picture of what's going
on with the case of dn's....

Thanks,
Scott