Google GitGoogle Git
Sign in
apache/datafusion-comet/HEAD
  1. 8320ae4 fix: Delete completed tasks' data files when an Iceberg write job fails (#5663) by Andy Grove · 11 hours ago main
  2. 5c5dda8 perf: give collect_list and collect_set a native GroupsAccumulator (#5803) by Andy Grove · 11 hours ago
  3. 149a1db feat: support max_by and min_by aggregate expressions (#4817) by Andy Grove · 11 hours ago
  4. 9388ecb ci: move the job routing policy out of ci.yml expressions and into compute-changes.py (#5850) by Andy Grove · 13 hours ago
  5. 4e9ab67 feat: expose native Parquet scan I/O and read-amplification metrics (#5453) by Chao Sun · 13 hours ago
  6. 5e302d9 feat: route translate and to_csv through codegen dispatch by default (#5032) by Andy Grove · 15 hours ago
  7. bdd4b90 test: name the whole dispatched subtree in the decimal promotion assertion (#5849) by Andy Grove · 18 hours ago
  8. 99237e1 feat: support native aggregate function `mode` (#4782) by Andy Grove · 19 hours ago
  9. bd1aa66 fix: isolate object-store registration by backend and configuration (#5503) by Chao Sun · 21 hours ago
  10. 8e489ea fix: list a fanout Iceberg write's data files in a stable order (#5810) by Andy Grove · 22 hours ago
  11. bf0558a feat: support Spark encode expression via codegen dispatch (#5037) by Andy Grove · 22 hours ago
  12. f4b8efd test: add helpers to assert whether an expression ran natively or via codegen dispatch (#5610) by Andy Grove · 22 hours ago
  13. 2d3eca2 fix: decide libhdfs routing from the scheme as written (#5825) by Oleks V · 23 hours ago
  14. 01f9c0c fix: roll native Iceberg data files on iceberg-java's 1000-row grid (#5780) by Andy Grove · 24 hours ago
  15. 392da2c ci: gate the Delta contrib build on symbols, not on libcomet size (#5827) by Andy Grove · 36 hours ago
  16. 80c8ec3 chore: Improve network retry configuration for maven and artifact upload (#5782) by Oleks V · 2 days ago
  17. df632c3 test: strengthen ANSI exception assertions (#5800) by KUAN-HAO HUANG · 2 days ago
  18. c152051 fix: align string to timestamp parsing with Spark's segment rules (#5682) by Peter Lee · 2 days ago
  19. c0b0b5e test: expand ANSI coverage for round, conv and elt (#5799) by KUAN-HAO HUANG · 2 days ago
  20. 9b608cc ci: shard Iceberg Spark tests across four runners (#5459) by Chao Sun · 2 days ago
  21. 26a70bd fix: ignore structural tags when lifting expression coverage (#5471) by Chao Sun · 2 days ago
  22. c8ee6ae fix: preserve aggregate result identity during exchange reuse (#5470) by Chao Sun · 2 days ago
  23. 04f7444 feat: route `abs` on interval types through the codegen dispatcher (#5622) by Kazantsev Maksim · 2 days ago
  24. fd8e09e bench: measure nested types as native shuffle hash partitioning keys (#5788) by Liang-Chi Hsieh · 2 days ago
  25. 4989b5e perf: avoid repeated decimal promotion in expression serialization (#5736) by Peter Lee · 2 days ago
  26. 424c31a fix: enable FIRST/LAST partial merge (#5041) by Peter Lee · 2 days ago
  27. 5627ab8 perf: reduce allocations when collecting cache statistics (#5734) by Peter Lee · 2 days ago
  28. 79d1be1e fix: match iceberg-java's exception for unclustered input to a clustered Iceberg write (#5779) by Andy Grove · 2 days ago
  29. e256ddb chore: run Preflight mandatory (#5784) by Oleks V · 2 days ago
  30. 4eeb1f8 fix: check nested TIMESTAMP_MILLIS overflow in unfiltered scans (#5740) by Peter Lee · 2 days ago
  31. 17f54da test: cover ambiguous exact nested Parquet field matches (#5751) by Peter Lee · 2 days ago
  32. fefee03 ci: label pull requests by changed paths and title prefix (#5762) by dustin · 3 days ago
  33. f6714a6 fix: attach tokio runtime threads to the JVM as daemon threads (#5748) by Feng Zhang · 3 days ago
  34. b886032 perf: use Arrow cast for date to timestamp NTZ (#5735) by Peter Lee · 3 days ago
  35. 0d1348f feat: address remaining issues for `CreateArray` (#5766) by Oleks V · 3 days ago
  36. 49a4b6c feat: enable codegen dispatch for `lpad` and `rpad` (#5764) by Daipayan Mukherjee · 3 days ago
  37. fd5592f fix: dispatch Iceberg system functions wrapped as ApplyFunctionExpression (#5773) by Andy Grove · 3 days ago
  38. b56268d fix: let AQE optimize queries over Comet caches (#5733) by Peter Lee · 3 days ago
  39. 4c14448 refactor: share one helper for pushing a struct's null mask into its children (#5769) by Liang-Chi Hsieh · 3 days ago
  40. 03875d4 fix: native Iceberg write panics on an evolved partition spec and on a timestamptz partition path (#5729) by Andy Grove · 3 days ago
  41. 8c7b706 feat: native dynamic filter pushdown for hash join into Parquet scans (#5699) by Ping Zhang · 3 days ago
  42. 8e68468 feat: support native concat_ws with string arrays (#5725) by Peter Lee · 3 days ago
  43. 3734905 test: restore Spark 4.1 Variant shredding suites (#5745) by KUAN-HAO HUANG · 3 days ago
  44. 8d25379 bench: add a benchmark for the Spark hash kernels (#5765) by Liang-Chi Hsieh · 3 days ago
  45. e696f53 feat: normalize marked Variant arrays at the native Parquet boundary (#5715) by Peter Lee · 3 days ago
  46. 6f25df7 perf: cache expected schemas for remote shuffle decoding (#5722) by Ping Zhang · 3 days ago
  47. cb4a261 chore: Add benches for datetime funcs (#5767) by Bhargava Vadlamani · 3 days ago
  48. cbe3b39 perf: evaluate posexplode array expressions once per batch (#5737) by KUAN-HAO HUANG · 3 days ago
  49. 31a69ee test: restore Parquet V2 writer and delta encoding coverage (#5760) by KUAN-HAO HUANG · 3 days ago
  50. 5af898b fix: read Iceberg tables partitioned by an unknown transform (#5759) by Andy Grove · 3 days ago
  51. bb9e740 fix: apply the parent struct's null mask before hashing its fields (#5754) by Liang-Chi Hsieh · 3 days ago
  52. 6368c4e test: restore Comet coverage for recursive HAVING and ORDER BY (#5755) by KUAN-HAO HUANG · 3 days ago
  53. 99d3100 fix: preserve ANSI errors for rejected TIMESTAMP_NTZ casts (#5752) by Peter Lee · 3 days ago
  54. 92ad99e fix: keep the dictionary hash fast path off nested and reseeded buffers (#5757) by Liang-Chi Hsieh · 3 days ago
  55. 7f1e001 test: cover Spark-to-Arrow batch conversion edge cases (#5713) by Peter Lee · 4 days ago
  56. 7f45ead add_benches_hash_agg (#5730) by Bhargava Vadlamani · 4 days ago
  57. 8a67fe0 fix: propagate Arrow array copy errors (#5747) by KUAN-HAO HUANG · 4 days ago
  58. bc74cc7 feat: route next_day and levenshtein collated input through the codegen dispatcher (#5720) by ChenChen Lai · 5 days ago
  59. 0b4549f perf: slice the child instead of gathering it when unnesting (#5667) by Andy Grove · 5 days ago
  60. 820b2d5 perf: validate shuffle IPC context reuse savings (#5727) by Peter Lee · 5 days ago
  61. 135ee53 fix: distinguish "nothing spilled" from a spill backend with no local path (#5726) by Andy Grove · 5 days ago
  62. 7e19843 chore: add benches array functions (#5700) by Bhargava Vadlamani · 5 days ago
  63. b5a39c3 feat: support nested types as native shuffle hash partitioning keys (#5567) by Liang-Chi Hsieh · 6 days ago
  64. f97fb45 fix: make columnar-to-row benchmarks exercise Comet (#5718) by KUAN-HAO HUANG · 6 days ago
  65. 0fc6779 fix: restore columnar transitions under the native Iceberg write (#5696) by Andy Grove · 6 days ago
  66. c348f77 fix: correctly rounded decimal to double/float cast matching BigDecimal.doubleValue/floatValue (#5684) by Peter Lee · 6 days ago
  67. 75fdddc deps: bump DataFusion 55.0 and Arrow/Parquet 59.2 (#5262) by Matt Butrovich · 6 days ago
  68. a88288b fix: report native shuffle read metrics (#5554) by Peter Lee · 6 days ago
  69. 7190df6 feat: add native spark_sequence kernel for integral element types (#5614) by ChenChen Lai · 6 days ago
  70. 1c38a3e fix: recover native Celeborn shuffle from oversized rows (#5668) by Ping Zhang · 6 days ago
  71. a1f8bd3 feat: support S3 compliant filesystems (#5314) by Oleks V · 6 days ago
  72. 93c3234 feat: route unrecognized StaticInvoke and Invoke through the codegen dispatcher (#5692) by Andy Grove · 6 days ago
  73. 719cba1 test: add explode operator microbenchmark (#5381) by Andy Grove · 7 days ago
  74. 2da3291 fix: prevent silent overflow when reading Parquet TIMESTAMP_MILLIS values (#5177) by Peter Lee · 7 days ago
  75. 296470c fix: fall back for concat_ws with array arguments instead of failing natively (#5679) by Peter Lee · 7 days ago
  76. 72fcc6f chore(deps): bump the codeql-actions group with 2 updates (#5669) by dependabot[bot] · 7 days ago
  77. 1119deb perf: reuse per-partition scratch in the shuffle write path (#5568) by dustin · 7 days ago
  78. 165e2c6 feat: carry VariantType identity through schema serialization (#5631) by Peter Lee · 7 days ago
  79. a1352ca fix: return NULL from rpad/lpad when the length column is NULL instead of panicking (#5680) by Peter Lee · 7 days ago
  80. 683a021 fix: match Spark's ANSI bound check for float/double to integral casts (#5683) by Peter Lee · 7 days ago
  81. 81d637b feat: implement regr_slope, regr_intercept, regr_r2, regr_sxx, regr_syy, regr_sxy aggregates (#4775) by Andy Grove · 7 days ago
  82. 9cac716 test: exercise the Iceberg write split-operator plan in Iceberg's own suites (#5640) by Andy Grove · 7 days ago
  83. eaf6426 feat: Support Iceberg system functions (bucket, truncate, years/months/days/hours) natively (#5638) by Andy Grove · 7 days ago
  84. 55ae4f2 fix: expand object store option references, uniquify constant metadata names, drop dead parquet JNI (#5653) by dustin · 7 days ago
  85. c052ea4 fix: make native shuffle spill metrics independent of input batching (#5628) by Chao Sun · 7 days ago
  86. 4219fc7 fix: match Spark's null short-circuiting in array_join and enable it natively (#5558) by Viacheslav Inozemtsev · 8 days ago
  87. e2ded76 fix(celeborn): reject unsafe native push completion tracking (#5665) by Ping Zhang · 8 days ago
  88. a089669 fix: Bump iceberg-rust so native Iceberg writes URL-escape partition paths (#5651) by Andy Grove · 8 days ago
  89. 7b1833e chore: move dead and defensive serde guards out of convert (#5595) by Andy Grove · 8 days ago
  90. 90c1dd9 fix: make CometDiskBlockWriter spill registry per-task instead of executor-global (#5493) by Peter Lee · 8 days ago
  91. ef62b46 fix: accept dictionary encodings in remote shuffle (#5650) by Ping Zhang · 8 days ago
  92. 10537e1 fix: skip codegen dispatcher null short-circuit when a foldable subtree can raise (#5623) by Andy Grove · 9 days ago
  93. 5b332b9 feat: implement native Iceberg V2 writer via iceberg-rust (#5361) by Jordan Epstein · 9 days ago
  94. 2011eb9 feat: support unicode case sensitive field names for reading parquet (#5602) by Oleks V · 9 days ago
  95. 6e366ac fix: scope Celeborn bootstrap hooks to Comet clients (#5627) by Ping Zhang · 9 days ago
  96. 5c41c06 chore: add benches datetime (#5620) by Bhargava Vadlamani · 9 days ago
  97. 88176ea fix: rebase map offsets in mapsort so sliced maps do not overrun entries (#5630) by Liang-Chi Hsieh · 9 days ago
  98. e98a2fb fix: preserve Arrow Field metadata across C Data exports (#5552) by Peter Lee · 9 days ago
  99. 8729f6e add_comet_meeting_link (#5621) by Bhargava Vadlamani · 10 days ago
  100. 252b7c8 test: enable native path in lower/upper_enabled sql fixtures (#5619) by Cestercian · 10 days ago