Re: Instance level authorization in Pyramid

2011-09-05 Thread Michael Merickel
It's not incorrect, I just merged two thoughts which probably made it unclear. If he doesn't specify a ``traverse`` parameter then traversal will not happen, which will use a similar RootFactory to what I showed except that he might want to raise a HTTPNotFound if the ``page`` is None. The rest of

Re: Instance level authorization in Pyramid

2011-09-05 Thread Chris McDonough
On Mon, 2011-09-05 at 15:50 -0500, Michael Merickel wrote: > Brian, I just want to clarify some points from your original email. > > > Specifying the ``factory`` on the route is telling the traversal > system how to get the root of your resource tree for that specific > route. Thus in your exampl

Re: Instance level authorization in Pyramid

2011-09-05 Thread Michael Merickel
Brian, I just want to clarify some points from your original email. Specifying the ``factory`` on the route is telling the traversal system how to get the root of your resource tree for that specific route. Thus in your example you might do: def PageFactory(request): pagename = request.matchd

Re: Instance level authorization in Pyramid

2011-09-05 Thread Chris McDonough
On Mon, 2011-09-05 at 12:44 -0700, Brian wrote: > Chris, > > Thanks for the reply. One more question... > > Is it acceptable for __acl__ to be a callable associated with an > instance? > > def __acl__(self): > return [ > (Allow, 'user:%s' % self.owner, 'edit'), > ] No, it must be an at

Re: Instance level authorization in Pyramid

2011-09-05 Thread Brian
Chris, Thanks for the reply. One more question... Is it acceptable for __acl__ to be a callable associated with an instance? def __acl__(self): return [ (Allow, 'user:%s' % self.owner, 'edit'), ] Thanks, Brian -- You received this message because you are subscribed to the Google Gro

Re: Instance level authorization in Pyramid

2011-09-05 Thread Chris McDonough
On Thu, 2011-09-01 at 06:30 -0700, Brian wrote: > I'm in the early stages of designing a my first Pyramid app and I was > hoping for some verification on my approach to instance level > authorization. Most of the stock documentation discusses global ACLs > which apply to an entire class, not indivi