improve CursorIterator

Consider making CursorIterator more extensible. It would be reasonable to be able to subclass to provide a range iterator (i.e. a forward iterator that checks an upper bound key, or a reverse iterator checking a lower bound). Since the class is final and tryToComputeNext is private, this is not currently feasible.

Another minor point is that the state machine should probably include a CLOSED state (in which hasNext returns false, and repeated calls to close are idempotent).

[benalexau]

I just pushed the final commits. LmdbJava now supports every range type offered by Guava (in both forward and backward cursor directions) and method names now reflect its terminology. There is also an example of using a Guava comparator, plus I added Guava to the ComparatorTest to further validate the project’s comparators against third-party references. Are we happy to close this ticket?

[benalexau]

The only thing that feels a bit unclear is whether keys are inclusive or exclusive. Ideally we should let users choose for themselves.

I too found the names unclear, which is why I put examples in the KeyRangeType JavaDocs. I intend to add some more KeyRangeType values to cover the exclusive cases.

I suppose that we intend to let users provide their own comparator by overriding the BufferProxy.compare method, right? I think this is good since there are quite a few comparators out there that users might want to use for performance reasons.

Users can provide their own Comparator by invoking CursorIterator<T> Dbi.iterate(Txn<T>, KeyRange<T>, Comparator<T>). Is that sufficient?

Maybe we could add a test using (for example) the lexicographical comparators in UnsignedBytes and SignedBytes?

Good idea. I’ll tackle that tomorrow.