jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Suhail M. Ahmed" <ilya...@mac.com>
Subject Re: jcr authentication and authorization
Date Sun, 06 Feb 2005 13:26:34 GMT
Hi David,

On Feb 6, 2005, at 11:17 AM, David Nuescheler wrote:
>> Build a repository adapter for LDAP to store A'nA data. I propose LDAP
>> because I  have experience with it. To this end I would like to
>> translate publicly available LDAP schemas to nodedefs. I have a few
>> days experience with JavaCC, if someone on this list can help me with
>> questions on using it to parse LDAP ldifs, I could have the whole 
>> bunch
>> translated in a jiffy.
> is there a reason why you want to go through ldif instead of directly
> going to the ldap server?

I think there will be two types of architecture variants of how LDAP 
and 170 will be used. In the first instance, LDAP servers being 
configured as a repository underneath jackrabbit and the other instance 
will be along the vision you elaborate below. I would like jackrabbit 
to support both. What I am currently trying to do is to implement the 
first variant. This would allow one to access LDAP using 170 API for 
AnA purposes.

> i think an ldap-content-importer for jackrabbit would be great,
> basically you could point an import script using something
> like http://docs.sun.com/source/816-5618-10/ (or jndi) directly
> at an ldap server.
> then, first all the nodetypes get generated (if necessary)
> and then all the information is sucked out of the ldap server and
> stored into the content repository.

By virtue of implementing the first variant the second set of use cases 
would be direct consequence. What I have in mind is the following:

View raw message