Speed up requests for many rows [LUCENE-6828]

Standard relevance ranked searches for top-X results uses the HitQueue class to keep track of the highest scoring documents. The HitQueue is a binary heap of ScoreDocs and is pre-filled with sentinel objects upon creation.

Binary heaps of Objects in Java does not scale well: The HitQueue uses 28 bytes/element and memory access is scattered due to the binary heap algorithm and the use of Objects. To make matters worse, the use of sentinel objects means that even if only a tiny number of documents matches, the full amount of Objects is still allocated.

As long as the HitQueue is small (< 1000), it performs very well. If top-1M results are requested, it performs poorly and leaves 1M ScoreDocs to be garbage collected.

An alternative is to replace the ScoreDocs with a single array of packed longs, each long holding the score and the document ID. This strategy requires only 8 bytes/element and is a lot lighter on the GC.

Some preliminary tests has been done and published at https://sbdevel.wordpress.com/2015/10/05/speeding-up-core-search/
These indicate that a long[]-backed implementation is at least 3x faster than vanilla HitDocs for top-1M requests.

For smaller requests, such as top-10, the packed version also seems competitive, when the amount of matched documents exceeds 1M. This needs to be investigated further.

Going forward with this idea requires some refactoring as Lucene is currently hardwired to the abstract PriorityQueue. Before attempting this, it seems prudent to discuss whether speeding up large top-X requests has any value? Paging seems an obvious contender for requesting large result sets, but I guess the two could work in tandem, opening up for efficient large pages.

---
Migrated from [LUCENE-6828](https://issues.apache.org/jira/browse/LUCENE-6828) by Toke Eskildsen (@tokee), updated Oct 07 2015

Start by examining the HitQueue, ScoreDocs, and PriorityQueue abstractions described in the issue, then review the linked preliminary performance tests. Determine whether large top-X requests justify the proposed refactoring and how correctness and performance would be evaluated; the work is done when the approach is agreed and validated against relevant request sizes.

Written by the indexing model from the issue text.