Vision: To provide the fastest possible native primitives for file content indexing by aggressively bypassing bottlenecks in standard Java.
- Core Native Engine: Basic JNI implementation.
- Blueprint Standards: README, Reference, and Philosophy integration.
- N-Gram Index: 3-gram index over file content for substring filtering
- Basic Performance Suite: Initial benchmarks vs standard Java
- N-Gram Filter: Query → 3-grams → lookup in RAM → candidate filtering
- SIMD Substring Scan: AVX2/AVX-512 vectorized substring search
- File Type Support: Index .java, .kt, .xml, .md and other text files
- Zero-Copy Access: Off-heap mmap index, no full-text database
- Method Search: Pattern-based method finding with regex matching
- Code Visualization: Mini-UI for method extraction and display
- FastFileSearch Integration: Seamless integration with file name search
- Performance Optimization: 2-10ms on 200GB drives, 1-5ms on SSD
- Full Stability Audit: Long-run stress testing
- Enterprise Support: NUMA-awareness and Large Pages support
- CLI Integration: FastSearch command for instant file/content lookup
Focus: Performance is our USP. We optimize where Java stops.