Messages by Thread
-
-
[PR] ci: label pull requests from workflow_run instead of pull_request_target [datafusion-ballista]
codekongdev (via GitHub) via github
-
Re: [I] Substrait: the grouping set column holds `__grouping_id`, not the set's index [datafusion]
namanjain24-sudo (via GitHub) via github
-
Re: [PR] build(deps): bump the codeql-actions group across 1 directory with 2 updates [datafusion-comet]
comphead (via GitHub) via github
-
[I] Incorrect regex-to-LIKE simplification for literal backslashes [datafusion]
lonless9 (via GitHub) via github
-
[PR] perf: [EXPERIMENT] prototype faster native columnar-to-row conversion [datafusion-comet]
andygrove (via GitHub) via github
-
Re: [PR] Derive: pass dialect method signatures through macro tokens [datafusion-sqlparser-rs]
alamb (via GitHub) via github
-
Re: [PR] Do not broadcast when co-partitioned. [datafusion]
ShayanGho (via GitHub) via github
-
[PR] fix: TopK dynamic filter returns wrong rows when one sort key contain… [datafusion]
shahwazreza (via GitHub) via github
-
Re: [PR] perf: unroll PrimitiveArray hashing for the no-null path [datafusion]
mbutrovich (via GitHub) via github
-
[PR] fix: keep constant filters above grand totals [datafusion]
Phoenix500526 (via GitHub) via github
-
[I] PostgreSQL: unary minus binds more loosely than `^`, so `-1 ^ 0` parses as `-(1 ^ 0)` [datafusion-sqlparser-rs]
kumarUjjawal (via GitHub) via github
-
[PR] fix: prevent panic in ceil and floor on decimal scalar when number_rows is 0 [datafusion]
SPY-Github22 (via GitHub) via github
-
[PR] fix: count the largest capacity seen for a shared buffer address [datafusion]
andygrove (via GitHub) via github
-
Re: [I] Use null counts when estimating IS NULL and IS NOT NULL [datafusion]
gabotechs (via GitHub) via github
-
Re: [I] Potential revamp of broadcast compression policy [datafusion-comet]
comphead (via GitHub) via github
-
[PR] chore: prepare for DataFusion 56 [datafusion-iceberg]
Rich-T-kid (via GitHub) via github
-
Re: [I] Sqlparser Binary Size [datafusion-sqlparser-rs]
srv1n (via GitHub) via github
-
[PR] chore(deps): bump parquet from 59.3.0 to 59.3.2 [datafusion-iceberg]
dependabot[bot] (via GitHub) via github
-
Re: [PR] perf: factor common guards out of pruning predicates [datafusion]
github-actions[bot] (via GitHub) via github
-
Re: [PR] feat: give PhysicalExtensionCodec the full decode context [datafusion]
namanjain24-sudo (via GitHub) via github
-
[PR] fix: ship Iceberg scan pool entries per partition so task memory does… [datafusion-comet]
parthchandra (via GitHub) via github
-
[I] Offset overflow - Spark-to-Arrow conversion cuts batches by row count only [datafusion-comet]
andygrove (via GitHub) via github
-
Re: [I] MinMaxStructState::emit_to(EmitTo::First(n)) subtracts row counts from a byte counter [datafusion]
dragonknightk (via GitHub) via github
-
[PR] fix: subtract emitted memory size in MinMaxStructState::emit_to(EmitTo::First) [datafusion]
dragonknightk (via GitHub) via github
-
[I] Aggregates with `FILTER (WHERE ...)` copy every input column once per aggregate [datafusion]
xudong963 (via GitHub) via github
-
Re: [I] Wrong results: with `pushdown_filters = true`, a filter on a volatile subquery column is evaluated again in the Parquet scan [datafusion]
matthewmturner (via GitHub) via github
-
[PR] fix: reject a multi-key null-aware join without a NOT IN value key count [datafusion]
viirya (via GitHub) via github
-
[PR] chore: remove the native columnar-to-row implementation [datafusion-comet]
andygrove (via GitHub) via github
-
[PR] fix: wrong aggregate functional dependencies when a GROUP BY column is not determined by the key [datafusion]
asolimando (via GitHub) via github
-
Re: [PR] docs(optimizer): add generated optimizer rules reference [datafusion]
alamb (via GitHub) via github
-
[PR] chore(ci): bump taiki-e/install-action from 2.88.1 to 2.89.0 [datafusion-ballista]
dependabot[bot] (via GitHub) via github
-
Re: [PR] fix: read Hive's null partition placeholder as NULL [datafusion]
xanderbailey (via GitHub) via github
-
Re: [I] Native Iceberg write silently falls back on Iceberg 1.12: ManifestFiles.read(ManifestFile, FileIO) was removed [datafusion-comet]
tlfua (via GitHub) via github
-
[PR] fix(physical-plan): return an error when an Unnest or Explain node has unset fields [datafusion]
geoHeil (via GitHub) via github
-
Re: [I] Incorrect statistics read for struct array in parquet [datafusion]
matthewmturner (via GitHub) via github
-
[I] Offset overflow - Native columnar-to-row conversion wraps row offsets past 2 GiB [datafusion-comet]
andygrove (via GitHub) via github
-
[I] Offset overflow - Native Parquet scan expands flat dictionary columns and reads twice Spark's batch rows [datafusion-comet]
andygrove (via GitHub) via github
-
[I] Offset overflow - String kernels can produce more than 2 GiB per batch, and some reserve it up front [datafusion-comet]
andygrove (via GitHub) via github
-
[I] Offset overflow - Codegen dispatcher writes each batch's output into one Arrow vector [datafusion-comet]
andygrove (via GitHub) via github
-
[I] Offset overflow - JVM columnar shuffle cuts row-to-Arrow batches by row count only [datafusion-comet]
andygrove (via GitHub) via github
-
[I] Offset overflow - Explode copies columns past 2 GiB and never splits one long row [datafusion-comet]
andygrove (via GitHub) via github
-
[I] Offset overflow - Shuffle read panics expanding dictionary blocks from the JVM columnar shuffle [datafusion-comet]
andygrove (via GitHub) via github
-
Re: [PR] fix: give zstd the real destination capacity when compressing cached batches [datafusion-comet]
parthchandra (via GitHub) via github
-
[PR] prototype: capture Parquet scan operations with tracing in I/O tests [datafusion]
adriangb (via GitHub) via github
-
[I] Support disabling output batch coalescing in HashJoinExec (and LimitedBatchCoalescer) [datafusion]
stuhood (via GitHub) via github
-
Re: [PR] feat: experiment with `RealUsagePool` [datafusion-comet]
comphead (via GitHub) via github
-
Re: [PR] feat: support multi-column NOT IN subqueries with null-aware joins [datafusion]
alamb (via GitHub) via github
-
[I] Native Iceberg scan copies the whole scan's delete-file metadata into every task, causing executor OOM on large merge-on-read tables [datafusion-comet]
parthchandra (via GitHub) via github
-
[PR] fix: [branch-1.2] preserve float collect_set semantics in distinct aggregate chains (#7220) [datafusion-comet]
andygrove (via GitHub) via github
-
[PR] fix: `ballista.planner.propagate_empty.enabled` has no effect: `PropagateEmptyExecRule` always runs [datafusion-ballista]
milenkovicm (via GitHub) via github
-
Re: [I] TopK dynamic filter returns wrong rows when one sort key contains another (DynamicFilterPhysicalExpr remaps the inner column first) [datafusion]
shahwazreza (via GitHub) via github
-
Re: [PR] fix: bound parallel batch sorting in SortExec [datafusion]
geoffreyclaude (via GitHub) via github
-
Re: [I] Support DATETIME SQL type in the planner [datafusion]
alamb (via GitHub) via github
-
Re: [PR] docs: move optimizer rule reference into rustdoc and remove the markdown file [datafusion]
alamb (via GitHub) via github
-
Re: [I] `BoundedWindowAggExec` can emit batches larger than `batch_size` [datafusion]
comphead (via GitHub) via github
-
Re: [PR] refactor: return Result from fallible catalog and schema methods [datafusion]
gratus00 (via GitHub) via github
-
[I] perf: probe tasks wait for the shared broadcast decode, and every wave of tasks decodes again [datafusion-comet]
comphead (via GitHub) via github
-
[PR] PostgreSQL: Support `ONLY` in `CREATE INDEX` [datafusion-sqlparser-rs]
LucaCappelletti94 (via GitHub) via github
-
Re: [PR] chore(deps): bump pbjson-types from 0.8.0 to 0.9.0 in the proto group across 1 directory [datafusion]
comphead (via GitHub) via github
-
[PR] fix(proto): return an error instead of panicking when a scan needs a disabled feature [datafusion]
geoHeil (via GitHub) via github
-
Re: [PR] fix: give a multi-set aggregate the grouping set index Substrait defines [datafusion]
namanjain24-sudo (via GitHub) via github
-
Re: [I] Substrait producer emits AGGREGATION_PHASE_UNSPECIFIED for every aggregate and window function [datafusion]
namanjain24-sudo (via GitHub) via github
-
[PR] PostgreSQL: Support `[NOT] LEAKPROOF` in `CREATE FUNCTION` [datafusion-sqlparser-rs]
LucaCappelletti94 (via GitHub) via github
-
Re: [PR] fix: return a query error for correlated scalar subquery with outer-column HAVING over count aggregates [datafusion]
namanjain24-sudo (via GitHub) via github
-
[PR] fix: reject multiple UPDATE assignments to the same column [datafusion]
vshalpnjabi (via GitHub) via github
-
[PR] perf(core): split sort-shuffle rows by partition on arrival when it fits [datafusion-ballista]
NoahKusaba (via GitHub) via github
-
Re: [PR] fix: support binaryAsString in the native Parquet scan [datafusion-comet]
rich7420 (via GitHub) via github
-
[I] Reduce repeated work across filtered aggregates [datafusion]
yashrb24 (via GitHub) via github
-
Re: [I] Aggregations on partition columns may fail if `wrap_partition_value_in_dict` is used. [datafusion]
akupchinskiy (via GitHub) via github
-
Re: [I] CreateFunction statement can't represent input or output types that require metadata [datafusion]
cblp (via GitHub) via github
-
Re: [I] Support MIN/MAX over StructType in native aggregates [datafusion-comet]
parthchandra (via GitHub) via github
-
[PR] PostgreSQL: Support `WINDOW` in `CREATE FUNCTION` [datafusion-sqlparser-rs]
LucaCappelletti94 (via GitHub) via github
-
[I] AQE research: broaden skew-join applicability (dynamic filters, multi-join stages, default-on criteria) [datafusion-ballista]
andygrove (via GitHub) via github
-
[PR] stream ROW_NUMBER output by input batch [datafusion-comet]
nat-openai (via GitHub) via github
-
[PR] PostgreSQL: Support `SUPPORT` in `CREATE FUNCTION` [datafusion-sqlparser-rs]
LucaCappelletti94 (via GitHub) via github
-
[PR] PostgreSQL: Support `COST` in `CREATE FUNCTION` [datafusion-sqlparser-rs]
LucaCappelletti94 (via GitHub) via github
-
[PR] PostgreSQL: Support `ROWS` in `CREATE FUNCTION` [datafusion-sqlparser-rs]
LucaCappelletti94 (via GitHub) via github
-
[PR] fix: count shared buffers once in RecursiveQueryExec's memory reservation [datafusion]
akupchinskiy (via GitHub) via github
-
[PR] chore(deps): bump the all-other-cargo-deps group with 3 updates [datafusion]
dependabot[bot] (via GitHub) via github
-
Re: [I] Incorrect results: Bloom filters on `UInt8`, `Int8`, `UInt16` and `Int16` columns always return false negatives [datafusion]
alamb (via GitHub) via github
-
[PR] perf: evaluate the child of `size` once when it needs no NULL guard [datafusion-comet]
comphead (via GitHub) via github
-
Re: [I] Drop support for Spark 3.4 [datafusion-comet]
manuzhang (via GitHub) via github
-
Re: [I] [DISCUSSION] How should we test datafusion-spark against Spark? [datafusion]
alamb (via GitHub) via github