First Implementation Pass of Java FlatMap Proposal

On the plane to Seattle I wrote the first pass of my Java flat file map proposal.

I’m pretty happy with the code.

The 1.0 sources just contain a read-only version of the Map.

It turns out I need to write a version of java.util.Set because the entrySet() and valueSet() mechanism might actually be called by callers and this needs to be efficient.

This is probably another couple hours.

I might build an in-memory binary tree into a 2.0 version of FlatMap so that you can efficiently burn a map. Right now the user just needs to allocate more memory for building a corpus but not to maintain the corpus in production once it’s burned into a FlatMap.



%d bloggers like this: