v0.8.0

What's Changed

• Update version to v0.8.0 by @alexbaden in #563
• Use legacy syntax in Calcite by default in PyHDK. by @ienkovich in #562
• Avoid direct access to ResultSet in PyHDK by @ienkovich in #561
• [Warns] Reduce warnings amount by @Devjiu in #523
• Add common GPU runtime by @kurapov-peter in #566
• Add most of the arrow based tests to intel gpu CI flow by @kurapov-peter in #499
• Fix docker uuid by @leshikus in #560
• Fix arrow dictionary import. by @ienkovich in #585
• Fix uninitialized pointer. by @ienkovich in #586
• Fixup vectorized bitmap generator and add benchmark by @alexbaden in #582
• Fix taxi_bench when USE_HOT_DATA is enabled by @alexbaden in #587
• Add support for bitwise operations. by @ienkovich in #564
• Bump pygments from 2.7.4 to 2.15.0 in /docker by @dependabot in #589
• Refactor array loops generation. by @ienkovich in #592
• Heterogeneous demo notebook by @akroviakov in #576
• Ignore JVM signal handler for SIGINT. by @ienkovich in #593
• Add cardinality API to QueryBuilder. by @ienkovich in #591
• Add ostream operators for compilation options by @kurapov-peter in #598
• Calcite: Add support for the bitwise NOT (~) operator by @AndreyPavlenko in #580
• Fix warnings. by @ienkovich in #600
• Add TableStats to TableFragmentsInfo. by @ienkovich in #559
• Support UNNEST in projections. by @ienkovich in #595
• Remove accidental debug change. by @ienkovich in #602
• Support debug timers in config builder. by @ienkovich in #603
• Stop using chunk stats for expression ranges. by @ienkovich in #604
• [Bench] Reduce input cols number for fetch by @Devjiu in #590
• Invalidate row count cache on reduction. by @ienkovich in #609
• Fix default compilation options construction in arrow sql runner implemenation by @kurapov-peter in #599
• Re-enable ASAN build by @leshikus in #535
• Throw exception on empty projection. by @ienkovich in #611
• Add database id to TableGenerations. by @ienkovich in #612
• Copy table stats for shuffled data. by @ienkovich in #605
• Add support for GPerftools to taxi benchmark by @alexbaden in #614
• Assign unique id to Node copy. by @ienkovich in #619
• Add TopK/BottomK aggregate. by @ienkovich in #606
• Remove exceptions from decimal validator message by @alexbaden in #613
• Fix generate_custom_agg_target_infos type by @kurapov-peter in #620
• Add if_then_else to PyHDK query builder. by @ienkovich in #621
• Move to ubuntu 22.04 in docker files and update compute-runtime version by @kurapov-peter in #622
• Dumping IR after every optimization pass by @akroviakov in #528
• Introduce simplified string dictionary by @alexbaden in #615
• Support TopK aggregate in RelAlgDagBuilder. by @ienkovich in #626
• Move approx quantile code. by @ienkovich in #631
• Fix arrays fetch in ResultSetRegistry. by @ienkovich in #629
• Refactor GPU shared memory tests and add L0-specific ones [1/N] by @kurapov-peter in #618
• Make Logger free headers inlineNullValues.h and OpTypeEnums.h by @ienkovich in #632
• Remove implicit cast to size_t by @kurapov-peter in #634
• Update to LLVM 15 by @alexbaden in #633
• Support TopK aggregate in Calcite by @AndreyPavlenko in #641
• Update conda build env to openjdk 20 by @alexbaden in #638
• Disable IR file metadata by default by @alexbaden in #642
• Refactor GPU shared memory tests and add L0-specific ones [2/2] by @kurapov-peter in #637
• Forbid CompilationOptions default constructor by @kurapov-peter in #640
• Make FP64 capability check a GpuMgr responsibility by @kurapov-peter in #639
• [Join] Zero-copy storage column check. by @Devjiu in #623
• Add ENABLE_SQL build flag. by @ienkovich in #643
• Rename a runner label by @leshikus in #647
• Make L0 manager use all available GPU memory by @kurapov-peter in #644
• [Namings] Update job names by @Devjiu in #601
• Add quantile aggregate. by @ienkovich in #636
• Avoid unnecessary chunk stats recomputation on append. by @ienkovich in #646
• PVC CI test by @kurapov-peter in #645
• Fix lazy fetch for quantile. by @ienkovich in #651
• Enable multifrag results by default. by @ienkovich in #649
• Avoid unused values materialization in ResultSet::rowCount. by @ienkovich in #654
• Improve quantiles allocation. by @ienkovich in #653
• Introduce aggregate finalization. by @ienkovich in #650
• Use explicit types in LLVM Loads by @alexbaden in #655
• Refragmenting table by @akroviakov in #628
• Update to LLVM 15 in Docker image by @alexbaden in #659
• Remove OpenSSL by @alexbaden in #666
• Preserve interpolation on aggregate translation. by @ienkovich in #664
• Replace dictionary proxies with nested dictionaries 01/N by @ienkovich in #661
• [Join] InitHashTable optimisation by @Devjiu in #663
• Added QUANTILE agg function to Calcite by @AndreyPavlenko in #657
• Introduce per-thread memory pools for lock-free aloocations in quantile. by @ienkovich in #656
• Replace dictionary proxies with nested dictionaries 02/N by @ienkovich in #668

New Contributors

• @dependabot made their first contribution in #589

Full Changelog: v0.7.0...v0.8.0