Add an S3-based directory.

There are so many projects (latest one I've heard of is Nixiesearch, [presented at Haystack](https://shuttie.github.io/haystack24-nixiesearch-slides/)) trying to read Lucene indexes from S3, let's provide a S3-based (and other object stores) directory in Lucene directly?

The issue names no files, tests, or entry points. Start by reviewing Lucene's existing directory abstractions and clarifying the scope for S3 and other object stores; done should include an agreed design and a working directory implementation with tests.

Written by the indexing model from the issue text.