I have a use-case where it would be handy to use the deref control to request attributes from an entry's parent entry. But this would require kind of a pseudo-attribute 'parentDN' (similar to 'entryDN'). Anybody aware whether such a thing is already available?

--On Thursday, October 01, 2015 10:48 PM +0200 Ivan Nejgebauer <ian(a)uns.ac.rs&gt; wrote: > On 09/30/2015 09:19 PM, Michael Ströder wrote: >> I have a use-case where it would be handy to use the deref control to >> request attributes from an entry's parent entry. But this would require >> kind of a pseudo-attribute 'parentDN' (similar to 'entryDN'). >> >> Anybody aware whether such a thing is already available? > > I've attached my implementation -- hope it helps. The patch is against > REL_ENG_2_4. > > (More details: the parentDN part is fairly straightforward, given the > entryDN implementation as a model. However, the deref overlay can't use > dynamic attributes [you can try the unmodified slapd with deref and > entryDN], so I wrote a nasty hack to make it work with parentDN. Not > extensively tested, e.g., I'm not sure it would work "" as the parent.) If you want it considered for inclusion in OpenLDAP, please file an ITS as per <http://www.openldap.org/devel/contributing.html>

On 01.10.2015. 22:45, Quanah Gibson-Mount wrote: > --On Thursday, October 01, 2015 10:48 PM +0200 Ivan Nejgebauer > <ian(a)uns.ac.rs&gt; wrote: >> On 09/30/2015 09:19 PM, Michael Ströder wrote: >>> I have a use-case where it would be handy to use the deref control to >>> request attributes from an entry's parent entry. But this would require >>> kind of a pseudo-attribute 'parentDN' (similar to 'entryDN'). >>> >>> Anybody aware whether such a thing is already available? >> >> I've attached my implementation -- hope it helps. The patch is against >> REL_ENG_2_4. >> [...] > > If you want it considered for inclusion in OpenLDAP, please file an > ITS as per <http://www.openldap.org/devel/contributing.html> I will, once we have an I-D with the design -- the existing patch is just a quick hack. Off the top of my head, the I-D should address questions like: - Does the rootDSE have a parent (I think not, and the code behaves accordingly); - What to do about alias dereferencing/DN rewriting (IMO nothing, but it should be noted in the design.)