Signed-off-by: Robert L. Carpenter <robacarp@gmail.com>

Signed-off-by: Robert L. Carpenter <robacarp@gmail.com>

Signed-off-by: Robert L. Carpenter <robacarp@gmail.com>

Signed-off-by: Robert L. Carpenter <robacarp@gmail.com>

I'm dumping a bunch of queries and such through this logging mechanism -
it's useful to be able to filter those back down.

Trust me, this hurts me more than it does you.

This is kind of un-pretty, but functional.

Just a bit of interface definition. No real mechanics live here, nor
should they, in all likelihood.

Internally, we're using this for some ORM-y stuff.

Have I mentioned this isn't exactly stable yet?

I still don't really like this - what about the case where an Iterator
yields some falsy value? Might mandate hasNext() on \SparkLib\Iterator
descendants. It's already provided by MongoCursor and some others.

A base class for SPL Iterators. Adds an each(), a weak sort of map(), and
the expectation of a getNext().

\SparkLib\Iterator\Wrapper is a child class that takes another iterator
and wraps its result by instantiating a given class. At the moment, we're
using it to turn arrays from Mongo into instances of a model class.

More useful info _and_ easier to parse. What's not to like? (Apart from
the fact that the webservices stuff doesn't just expose this feature...)