I'm not sure about checkPermission with a String (or String[]). This
would imho only make sense if we see that we will have more
permissions in the future which I really don't see. And we don't need
an extension here as each additional permission would require changes
in the implementation to check those anyway.
So I would go with the separate methods - we could provide an abstract
class which already implements all methods, so implementing would be
just overwriting the required ones.
I also don't see a need to do it in the same way as JCR does - we
should do it what fits best in our resource api and what is the best
way to cover the use cases :)
Adding checks on the resource interface is problematic as this would
require each and every resource provider to implement these methods -
and this proposal is orthogonal so to speak to acl checks done by a
resource provider.
Carsten
2012/10/30 Ian Boston <i...@tfd.co.uk>:
> On 30 October 2012 20:18, Bertrand Delacretaz <bdelacre...@apache.org> wrote:
>> Hi,
>>
>> On Fri, Oct 26, 2012 at 10:43 AM, Mike Müller <mike...@mysign.ch> wrote:
>>> ...The main goal of this proposal is to provide a easy to use service in
>>> Sling to restrict (or grant) access to resources for special use cases
>>> (like giving access to some resources only between 8am and 5pm). The
>>> service should be very lightweight and should NOT be a replacement of
>>> ACLs...
>>
>> You have my bet that people *will* use this service as a replacement
>> for ACLs, and reinvent a few wheels in the process ;-)
>>
>> In the end, between this and the CRUD resource API changes we are
>> moving to two flavors of Sling: JCR-based, where the repository fully
>> handles such things, and non-JCR, where you need to implement your own
>> ResourceAccessGates and other niceties that JCR provides.
>>
>> I'm not against that, as long as it's a conscious decision and as long
>> as we're clear about best practices in either case.
>>
>>> ...I propose a new service interface ResourceAccessGateService...
>>
>> I'd call that just ResourceAccessGate.
>>
>>> enum GateResult { GRANTED, DENIED, NOTRESPONSIBLE };
>>
>> DONTCARE instead of NOTRESPONSIBLE maybe?
>>
>>> ...Alternatively, to make the interface more flexible, we could combine the
>>> canXXX() methods in a method named hasPermission with a parameter
>>> "operation" as String....
>>
>> I'd much prefer that, if this model can be closer to JCR's
>> Session.checkPermission(String absPath, String actions) that's helpful
>> IMO.
>>
>> (which makes me think that a new Resource.checkPermission method might
>> be more in line with Sling's resource-centric view on things).
>
> I agree with the single method checkPermissions as its a lot less
> overhead to those who have to implement, and the String actions allows
> for extension being out of band. What about String[] actions to allow
> one call to check many actions ?
>
> I am not certain about Resource.checkPermission (unless you mean a
> static), as that assumes "read" is granted ?
>
> Ian
>
>
>>
>> -Bertrand
--
Carsten Ziegeler
cziege...@apache.org
