·
2 commits
to main
since this release
Highlights
- Streamed execution in write and merged (lower memory pressure)
- Overall improved memory allocation
- Schema evolution in
MERGE
- Unity catalog support through
uc://
uri-prefix
Performance improvements
- feat: streamed execution in MERGE by @ion-elgreco in #3145
- feat: streamed execution, writer refactor, simplified generated columns and schema evolution by @ion-elgreco in #3229
- refactor: logical plans in writer by @ion-elgreco in #3141
- chore(python): choose performance over compile time in release mode by @ion-elgreco in #3177
- chore(python): use custom global alloc by @ion-elgreco in #3176
New features
- feat: add schema evolution to merge statement by @JustinRush80 in #3136
- feat(python): add capability to read unity catalog (uc://) uris by @omkar-foss in #3113
- feat: support reading tables via Unity Catalog provided credentials by @hntd187 in #3078
- feat: added workspace oauth source for UC by @hntd187 in #3152
- feat: update metadata field builder by @Nordalf in #3149
- feat: set column metadata from python by @umartin in #3189
- feat(disable conflict checker): start counting concurrent retries at 0 and disable conflict checker with maxRetries=0 by @daphnenhuch-at in #3204
- feat: make Add:get_stats public by @jkylling in #3216
- feat: configurable column encoding for parquet checkpoint files by @dmunch in #3214
- feat: return metrics about the commit by @HawaiianSpork in #3202
- feat: cdf tableprovider with predicate pushdown support by @ion-elgreco in #3220
Bug Fixes
- fix: properly encode metadata field values when creating actions by @rtyler in #3183
- fix: only use stats for required cols by @ion-elgreco in #3210
- fix: load cdf latest version by @ion-elgreco in #3218
- fix: adhere write stats configuration by @ion-elgreco in #3209
- fix(rust, python): treat FSCK files_removed as strings by @liamphmurphy in #3219
- fix: update snapshot after conflict check by @ion-elgreco in #3205
Other Changes
- chore: preparing the 0.24 release of the Rust crates` by @rtyler in #3132
- chore: fix typos by @roeap in #3138
- refactor: combine normal and cdf plan until write for merge by @ion-elgreco in #3142
- chore: remove lazy_static and once_cell dependencies by @nyurik in #3148
- docs: Fix broken markdown syntax by @hattajr in #3134
- chore: minor changes to allow 0.24.0 to be published by @rtyler in #3158
- chore: prevent the pr actions from filing on merge_group by @rtyler in #3161
- docs: fix broken links by @ion-elgreco in #3179
- chore: update to DataFusion 45.0.0 / arrow/parquet
54.0.0
by @alamb in #3175 - chore: adopt the released datafusion 45 release by @rtyler in #3195
- fix: removing openssl from the dependency tree by @rtyler in #3160
- chore: remove the unnecessary hashbrown dependency by @rtyler in #3198
- chore: switch to cache@v4 by @rtyler in #3230
- fix(typo): s/implemtnation/implementation/ by @akesling in #3227
- chore: bump python 0.25 by @ion-elgreco in #3233
- chore(python): set streaming opt-in by @ion-elgreco in #3178
New Contributors
- @nyurik made their first contribution in #3148
- @hattajr made their first contribution in #3134
- @JustinRush80 made their first contribution in #3136
- @umartin made their first contribution in #3189
- @daphnenhuch-at made their first contribution in #3204
- @dmunch made their first contribution in #3214
- @akesling made their first contribution in #3227
Full Changelog: python-v0.24.0...python-v0.25.0