add vector searching from lucene 9.0 using KnnVectorQuery and KnnVectorField

[mdavis95]

Replaces old, unused, and slow implementation using Superbit projects and minimum should match queries with cosine similarity

Examples searches:

Alone

search = new Search("vectorTestIndex");
search.addQuery(new VectorTopNQuery(new float[] { 1.0f, 0.0f, 0.0f, 0.0f }, 3, "v"));

Post Filtering (After the topN)

search = new Search("vectorTestIndex");
search.addQuery(new VectorTopNQuery(new float[] { 1.0f, 0.0f, 0.0f, 0.0f }, 3, "v"));
search.addQuery(new FilterQuery("red").addQueryField("description"));

Pre Filtering

search = new Search("vectorTestIndex");
VectorTopNQuery vectorTopNQuery = new VectorTopNQuery(new float[] { 1.0f, 0.0f, 0.0f, 0.0f }, 3, "v");
vectorTopNQuery.addPreFilterQuery(new FilterQuery("blue").addQueryField("description"));
search.addQuery(vectorTopNQuery);

Creating Index

ClientIndexConfig indexConfig = new ClientIndexConfig();
indexConfig.addDefaultSearchField("description");
indexConfig.addFieldConfig(FieldConfigBuilder.createString("description").indexAs(DefaultAnalyzers.STANDARD).sort());
indexConfig.addFieldConfig(FieldConfigBuilder.createUnitVector("v").index()); // use createVector if not unit normalized
indexConfig.setIndexName("vectorTestIndex");
zuliaWorkPool.createIndex(indexConfig);

Indexing

float[] vector = { 0, 0, 0.70710678f, 0.70710678f } // needs to be unit normalized if type is UNIT_VECTOR
Document mongoDocument = new Document();
mongoDocument.put("description", "great description");
mongoDocument.put("v", vector != null ? Floats.asList(vector) : null);
Store s = new Store("someId", "vectorTestIndex");