0.26.0

What's Changed

• Resolve repeated columns in AGG error #2781 by @joocer in #2783
• update rust dependencies by @joocer in #2785
• improved error message by @joocer in #2784
• support EXCEPT and ORDER BY #2786 by @joocer in #2787
• raw data statistic [parquet] #2789 by @joocer in #2790
• embed sample datasets as files #2791 by @joocer in #2792
• unpin cython version by @joocer in #2796
• Don't try to read stats from empty iceberg tables #2797 by @joocer in #2798
• Fix documentation issues: misleading comments, docstrings, and typos by @Copilot in #2799
• Fix README inaccuracies and inconsistencies for potential users by @Copilot in #2800
• Replace pyarrow for parquet metadata #2804 by @joocer in #2805
• Fix file descriptor leak and improve buffer handling in connectors and decoders by @Copilot in #2806
• sqlparser bump by @joocer in #2809
• Use Rugo for Schemas by @joocer in #2811
• fix-version by @joocer in #2812
• sync-version-across-files by @joocer in #2813
• Consolidate list_ops Cython files into single .so file with auto-generation by @Copilot in #2814
• Optimize import performance with lazy loading - 73% faster initialization by @Copilot in #2816
• Fix fuzzing test unreliability by making random value generation deterministic by @Copilot in #2817
• COUNT DISTINCT on empty results fails #2818 by @joocer in #2819
• [rugo] Use conditional metadata reading by @joocer in #2820
• fix compiler errors by @joocer in #2821
• refactor DATEDIFF function #2823 by @joocer in #2824
• improve external DBMS interactions by @joocer in #2826
• Performance optimizations: LRU caching and regex precompilation for SQL utilities by @Copilot in #2827
• SQL Connector doesn't support schemas #2829 by @joocer in #2830
• Consolidate join code into single .so file by @Copilot in #2828
• changed test structure by @joocer in #2833
• manual rig of tests after AI's first pass by @joocer in #2834
• create get started tools by @joocer in #2837
• Add INITCAP and REPLACE by @joocer in #2838
• reinstate examples by @joocer in #2839
• Simplify third-party library structure: consolidate from 3 locations to 2 by @Copilot in #2822
• compiled code housekeeping by @joocer in #2841
• Split test_shapes_and_errors_battery.py into 7 modular test files by @Copilot in #2840
• Parallelize selection and projection operations with free-threading support by @Copilot in #2832
• Enhance test coverage with 381 new edge cases and variations across 11 test files by @Copilot in #2842
• openssl headers by @joocer in #2843
• Support wildcards in file paths for SQL queries by @Copilot in #2845
• Support protocol prefixes for paths with wildcards (gs://, s3://) and permission controls by @Copilot in #2846
• Return column names as per SELECT #2849 by @joocer in #2850
• Implement fast JSONL decoder with Cython-based value extraction by @Copilot in #2848
• improve jsonl performance by @joocer in #2852
• improve jsonl performance by @joocer in #2853
• cython-performance by @joocer in #2854
• performance-tweaks by @joocer in #2856
• all tests passing again by @joocer in #2866
• Clickbench performance regression investigation 1 by @joocer in #2867
• changes to parquet reader by @joocer in #2868
• Changes to aggregate functions by @joocer in #2869
• jsonl reads draken by @joocer in #2870
• Clickbench performance regression investigation 1 by @joocer in #2871
• custom disk reader by @joocer in #2872
• fix disk io errors by @joocer in #2873
• rewrite memory view stream by @joocer in #2875
• fix agg bug by @joocer in #2876
• linux disk performance by @joocer in #2879
• group by performance by @joocer in #2880
• group by improvements by @joocer in #2881
• Performance Review by @joocer in #2883
• limit pushdown by @joocer in #2884
• Predicate Compaction Strategy by @joocer in #2886
• separate nested loop join by @joocer in #2889
• 0.26.0-release by @joocer in #2890
• Bump version from 0.26.1-beta.1726 to 0.26.0 by @joocer in #2891
• release by @joocer in #2893

New Contributors

• @Copilot made their first contribution in #2799

Full Changelog: 0.25.1...0.26.0-r