Hi everyone,
I am wondering if it makes sense to still extend the Jackrabbit API given that 
Jackrabbit 2 is mostly in maintenance mode.

In the context of https://issues.apache.org/jira/browse/JCRVLT-515 I would need 
some API to get all(!) bound principals for one session 
(https://issues.apache.org/jira/browse/JCR-4689)
Neither JCR nor Jackrabbit API have a method for that, but OTOH Oak API has 
(https://jackrabbit.apache.org/oak/docs/apidocs/org/apache/jackrabbit/oak/api/AuthInfo.html#getPrincipals--).

Unfortunately there is no easy way to get the Oak API entities from the JCR 
objects like Session (when running in Oak). This is often necessary as custom 
module running inside Sling or AEM use JCR entities exclusively (as access to 
the repository is managed by Sling).

Is there somewhere an overview what should be exposed by Jackrabbit API vs by 
Oak API?
To prevent extending the Jackrabbit API for all those edge cases I would 
recommend to rather introduce a simple way to convert a Oak-based JCR Session 
object into a Content Session 
(https://jackrabbit.apache.org/oak/docs/apidocs/org/apache/jackrabbit/oak/api/ContentSession.html).
 That way one could easily use Oak API if Oak is used (without being forced to 
pass both Oak and JCR objects around).

WDYT?

Konrad

Reply via email to