v0.2.1

Changes

🚀 Performance

• perf(encoding): reserve final matcher capacity for the dictionary (#34) @gargiulofrancesco
• perf(encoding): optimize pair counting and reserve training maps (#33) @gargiulofrancesco
• Replace long-bucket trie with length-grouped search (#31) @robert3005
• Add row-based training and encoding without copying (#30) @robert3005

🐛 Bug Fixes

• Fix partial shuffle training order (#28) @gargiulofrancesco

v0.2.0

Changes

✨ Features

• feat: separate dictionary safety and correctness validation (#24) @gargiulofrancesco
• feat: make CompactDictionary storage-backed (#23) @gargiulofrancesco

🧰 Maintenance

• Add rust-analyzer to pinned toolchain (#27) @gargiulofrancesco
• fix: remove vulnerable thrift dependency (#26) @gargiulofrancesco
• chore: remove cross-implementation benchmarks and rewrite README (#25) @gargiulofrancesco

v0.1.1

Changes

🐛 Bug Fixes

• fix(validation): reject unaddressable dictionaries (#22) @gargiulofrancesco
• fix(config): make default seed deterministic (#21) @gargiulofrancesco

v0.1.0

Changes

⚠️ Breaks

• Refactor/module layout (#20) @gargiulofrancesco

📖 Documentation

• Docs/fix imprecisions (#19) @gargiulofrancesco
• docs: shared in-memory ABI (plain interchange form) (#18) @gargiulofrancesco

v0.0.4

Added

• add back code_offsets to compressor (#15)

Other

• remove code boundaries (#13)

v0.0.3

Other

• fat-table layout with scalar copy and L2-indexed fallback (#12)
• Feat/decode fat table scalar (#7)
• update changelog (#10)

v0.0.2

Other

• automate releases with release-plz (#8)
• clean up benchmarks and decompression (#6)
• add benchmarks with onpair cpp (#5)
• c
• onpair
• gitignore