mrhhsg commented on PR #68279:
URL: https://github.com/apache/doris/pull/68279#issuecomment-6269994692
<!-- autopilot-response -->
Re-checked the full review history at HEAD
`ef01c98e4dcfe07e279fde164037bf539e476dc6`.
**Previous inline findings — fixed; both threads are resolved:**
- Sparse-accounting overcount in `segment_iterator.cpp`:
`_get_row_ranges_from_conditions()` now derives each stage's delta through
`_count_alive_rows()` (`be/src/storage/segment/segment_iterator.cpp:1162`),
which uses `roaring_bitmap_range_cardinality` against the surviving
`_row_bitmap` instead of the dense min/max envelope.
`DisjointKeyRangesOnlyCountAliveRows`
(`be/test/storage/segment/segment_iterator_key_range_stats_test.cpp:204`)
covers key ranges `{1}` and `{5000}` and asserts `RowsStatsFiltered == 1`.
- Trailing `DROP TABLE` in `point_lookup_zone_map_profile.groovy`: removed;
both tables are dropped only before creation (lines 21 and 161).
**Latest review:** the most recent `github-actions` review (2026-10-19, at
`ac69ac76305`) reports `NO_NEW_VALUABLE_FINDINGS` and states both earlier
concerns are addressed with no remaining P0/P1. The only delta since that
reviewed commit is `ef01c98e4dc`, which adapts BE unit-test fixtures to the
`check_duplicate_json_path` option removed on master (test-only, four files).
**Other top-level entries:** `/review` and `run buildall` commands, two
stale "Codex review failed" notices, a local PASS receipt, and TPC-H / TPC-DS /
ClickBench / coverage payloads. None contains a code finding.
The `CHANGES_REQUESTED` decision comes from the 2026-09-20 automated review;
I cannot dismiss it. Could a maintainer take another look and clear it if the
remaining concerns are satisfied?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
