refactor: improve execution pipeline and TPCC backends#313

Merged

KKould merged 27 commits intomainfrom

refactor--hep_match

Mar 28, 2026

Commits on Mar 27, 2026

refactor: optimizer normalization passes
KKould
committed
refactor: normalization scheduling to avoid matcher scans
KKould
committed
chore: output_columns reduces the number of Vecs constructed
KKould
committed
refactor: derive column positions during binding instead of BindPosition fallback
KKould
committed
refactor: avoid redundant operator clones in normalization rules
KKould
committed
feat: add LMDB storage and fix scalar subquery semantics
KKould
committed
refactor: borrow range bounds and reuse delete buffers
KKould
committed
refactor: make key encoding paths borrowed and reusable
KKould
committed
refactor: storage buffers and trim optimizer column-reference overhead
KKould
committed
refactor: column pruning to reuse borrowed column summaries
KKould
committed
refactor: remove memo and choose physical impls directly
KKould
committed
refactor: optimizer hint propagation and preindex implementation rules
KKould
committed
refactor: optimizer annotate pass to merge post-rules and restore borrowed scan hints
KKould
committed
optimizer: avoid schema diff in column pruning remap
KKould
committed
optimizer: avoid recursive expr walks in column pruning
KKould
committed
perf: prebind range comparison evaluators
KKould
committed
refactior: new ExecArena-based executor model
KKould
committed
fix: tpcc segfault caused by bump-backed executor drop order
KKould
committed
refactor: reuse arena result tuple across execution pipeline
KKould
committed
chore: uncheck for decode tuple primary key when query
KKould
committed
refactor(tpcc): split backends and improve benchmark tooling
KKould
committed
refactor(execution): unify executor node API and trim test helpers
KKould
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

refactor: improve execution pipeline and TPCC backends#313

refactor: improve execution pipeline and TPCC backends#313
KKould merged 27 commits intomainfrom
refactor--hep_match

Commits on Mar 27, 2026

refactor: optimizer normalization passes

refactor: normalization scheduling to avoid matcher scans

chore: output_columns reduces the number of Vecs constructed

refactor: derive column positions during binding instead of BindPosition fallback

refactor: avoid redundant operator clones in normalization rules

feat: add LMDB storage and fix scalar subquery semantics

refactor: borrow range bounds and reuse delete buffers

refactor: make key encoding paths borrowed and reusable

refactor: storage buffers and trim optimizer column-reference overhead

refactor: column pruning to reuse borrowed column summaries

refactor: remove memo and choose physical impls directly

refactor: optimizer hint propagation and preindex implementation rules

refactor: optimizer annotate pass to merge post-rules and restore borrowed scan hints

optimizer: avoid schema diff in column pruning remap

optimizer: avoid recursive expr walks in column pruning

perf: prebind range comparison evaluators

refactior: new ExecArena-based executor model

fix: tpcc segfault caused by bump-backed executor drop order

refactor: reuse arena result tuple across execution pipeline

chore: uncheck for decode tuple primary key when query

refactor(tpcc): split backends and improve benchmark tooling

refactor(execution): unify executor node API and trim test helpers

Commits on Mar 28, 2026

chore: codefmt

refactor(stats): load sketches alongside histograms in StatisticsMeta

feat: improve normalization rules and add tpcc benchmark runner

chore(release): bump kite_sql and macros crate to 0.2.0

feat(cargo): gate native backends behind rocksdb/lmdb features