Index aggregate rule improvements

The current aggregate rules in indexing configuration are not very flexible. It would be great if aggregates could be triggered also by mixin types or more generic path and/or property constraints as done with other indexing rules. The include settings could also be more flexible, and there are cases where excludes would also be useful.

For example one use case we have is being able to define an aggregate for a my:document type that may have other my:document nodes within the subtree. The aggregate should contain the entire my:document subtree, except for any sub-my:documents. Also, the sub-my:document nodes would not need to be indexed as aggregates.