The GitHub Actions job "CI" on fluss.git/feat/spark-scan-consolidation-pr1 has 
failed.
Run started by GitHub user YannByron (triggered by YannByron).

Head commit for run:
ef46c007bfb28a9f99801c26c23e6b11bdc43614 / Yann <[email protected]>
[spark] Restore lake predicate pushdown for PK tables in consolidated scan 
builder

The scan builder consolidation (e16af58c) merged FlussSupportsPushDownV2Filters
and FlussLakeSupportsPushDownV2Filters into a single trait but only preserved
the ARROW log-table branch. This left PK tables with lake enabled unable to push
predicates to the lake source, causing pk filter pushdown tests to report empty
pushed predicates.

Add an `else if (isDataLakeEnabled)` branch that probes the lake source via
FlussLakeUtils.applyLakeFilters — mirroring the former 
FlussLakeSupportsPushDown-
V2Filters behavior — so that accepted predicates are reported back to Spark and
passed to the planner for lake-path filtering.

🤖 Generated with [Qoder][https://qoder.com]

Report URL: https://github.com/apache/fluss/actions/runs/28576597303

With regards,
GitHub Actions via GitBox

Reply via email to