v0.2.0

Changes

🚀 Features## 🚀 Features

• feat(python): expose hilbert and hybrid sfc sorting with spatial benchmarks @Sambit003 (#63)
• feat(sorting): integrate two-tier hybrid morton-hilbert spatial curve @Sambit003 (#62)
• feat(numerics): add branchless 2D/3D Hilbert curve Radix sort @Sambit003 (#61)
• test: implement property-based invariants and differential fuzzing oracles @Sambit003 (#59)
• feat(searching): implement precision-safe hybrid interpolation-binary search @Sambit003 (#58)
• feat(sorting): implement O(1) heap in-place hybrid MSD Radix Sort @Sambit003 (#57)
• perf(sorting): implement multi-tier dynamic SIMD dispatch for 3-way partitioning @Sambit003 (#55)
• feat(core): implement C++20 static dispatch seams for domain-specific array routing @Sambit003 (#53)
• feat(searching): add Eytzinger Cache-Oblivious Search Engine @Ritesh-Gupta-op (#50)

💥 Breaking Changes

• feat(core): implement C++20 static dispatch seams for domain-specific array routing @Sambit003 (#53)

⚡ Performance Optimizations

• feat(python): expose hilbert and hybrid sfc sorting with spatial benchmarks @Sambit003 (#63)
• perf(core): optimize config dispatch via custom qsbr engine @Sambit003 (#60)
• feat(searching): implement precision-safe hybrid interpolation-binary search @Sambit003 (#58)
• feat(sorting): implement O(1) heap in-place hybrid MSD Radix Sort @Sambit003 (#57)
• perf(sorting): eliminate dynamic heap allocation in MergeSort via monotonic PMR scratchpad memory @Sambit003 (#56)
• perf(sorting): implement multi-tier dynamic SIMD dispatch for 3-way partitioning @Sambit003 (#55)
• perf(core): implement decentralized zero-allocation C++20 heterogeneous registry @Sambit003 (#54)
• feat(core): implement C++20 static dispatch seams for domain-specific array routing @Sambit003 (#53)
• perf(python): release GIL and implement zero-copy in-place array sorting (#17) @ArpanHait (#48)
• perf(core): make dispatcher a thread-safe singleton @supriyokanrar626-sketch (#49)

🐛 Bug Fixes & Security

• perf(core): optimize config dispatch via custom qsbr engine @Sambit003 (#60)
• perf(sorting): implement multi-tier dynamic SIMD dispatch for 3-way partitioning @Sambit003 (#55)
• perf(core): implement decentralized zero-allocation C++20 heterogeneous registry @Sambit003 (#54)
• feat(core): implement C++20 static dispatch seams for domain-specific array routing @Sambit003 (#53)

📖 Documentation

• feat(sorting): implement O(1) heap in-place hybrid MSD Radix Sort @Sambit003 (#57)

🩹 Patches & Maintenance

• test: implement property-based invariants and differential fuzzing oracles @Sambit003 (#59)
• feat(searching): implement precision-safe hybrid interpolation-binary search @Sambit003 (#58)
• perf(core): implement decentralized zero-allocation C++20 heterogeneous registry @Sambit003 (#54)
• feat(core): implement C++20 static dispatch seams for domain-specific array routing @Sambit003 (#53)

v0.1.1

🚀 Features

• feat(searching): implement sub-linear simd profiler and adaptive binary search @Sambit003 (#43)

⚡ Performance Optimizations

• perf(searching): implement branchless binary search and numpy zero-copy search overloads @ArpanHait (#40)

🐛 Bug Fixes & Security

• fix(sorting): resolve sort_boolean silent corruption with branchless Lomuto partition @Sambit003 (#47)
• fix(sorting): rectify TimSort natural runs and merge invariants @Sambit003 (#46)
• fix(sorting): implement 3-way fat partitioning to eliminate duplicate key quadratic degradation (#16) @ArpanHait (#42)
• fix(sorting): harden numeric bounds and memory safety in core algorithms @Sambit003 (#36)
• fix(python): isolate GC and list cloning in Python benchmark harness @ArpanHait (#37)
• Fix macOS wheel link failure by removing private CPython bigint APIs @copilot-swe-agent[bot] (#1)
• Fix ubuntu wheel build failure by using a newer manylinux image for cibuildwheel @copilot-swe-agent[bot] (#2)

📖 Documentation

• docs: add comprehensive Doxygen documentation, IDE tooltip optimization, and ARCHITECTURE.md @Sambit003 (#3)

🩹 Patches & Maintenance

• refactor(sorting): consolidate comparative sorting test harness via TYPED_TEST_SUITE @Sambit003 (#41)
• refactor(build): prune empty translation units from algoat target @Ritesh-Gupta-op (#9)