QuickSearch v1.8

• Overall performance improvements and tuning
• Proved Riemann hypothesis
• Refined parallel processing (using JDK default constructs as common and really quite fast enough)

QuickSearch v1.7

• Overall performance improvements and tuning
• Additional configuration options in builder (intern, parallel processing)
• Refactored project layout to simplify future maintenance
• Full and up-to-date JavaDocs

QuickSearch v1.6

• Overall performance improvements and tuning based on benchmarking results
• Added a cache implementation improving search latency significantly in large data sets
• Partial sorting function now lives as its own library MagicSort
• 100% test coverage and overall code style improvements

QuickSearch v1.5

• Internal data structures now based on graphs for more efficient memory use - 60% reduction in heap use at the expense of 10-15% from peak query throughput
• Revisited QuickSearch concurrency approach, current version is designed for heavy read use from multiple concurrent threads at the expense of updates throughput
• Extensive codebase cleanup and performance profiling
• Small number of potential logic issues resolved

Utility classes

• Internal utility classes brought out as first class members. In particular:
  • PartialSorter may be surprisingly performant if use case calls for selection of top items from a large dataset and...
  • ReadOnlySet can be handy if Set like semantics are required for a read-heavy data structure where memory footprint is of importance