From 4c81fe84aa71895ee99b2543e8d483ed96d6e45e Mon Sep 17 00:00:00 2001 From: Alessandro Solimando Date: Thu, 19 Mar 2026 13:28:56 +0100 Subject: [PATCH] fix: update clickbench expected plan for NDV-aware optimization #19957 introduced NDV extraction from Parquet metadata. The optimizer now sees NDV=1 for HitColor, BrowserCountry, BrowserLanguage in the clickbench test file and short-circuits COUNT(DISTINCT) to a constant projection, skipping the full table scan. --- datafusion/sqllogictest/test_files/clickbench.slt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/datafusion/sqllogictest/test_files/clickbench.slt b/datafusion/sqllogictest/test_files/clickbench.slt index 42f066a80de7..4e9849e3650a 100644 --- a/datafusion/sqllogictest/test_files/clickbench.slt +++ b/datafusion/sqllogictest/test_files/clickbench.slt @@ -1203,8 +1203,8 @@ logical_plan 02)--SubqueryAlias: hits 03)----TableScan: hits_raw projection=[HitColor, BrowserLanguage, BrowserCountry] physical_plan -01)AggregateExec: mode=Single, gby=[], aggr=[count(DISTINCT hits.HitColor), count(DISTINCT hits.BrowserCountry), count(DISTINCT hits.BrowserLanguage)] -02)--DataSourceExec: file_groups={1 group: [[WORKSPACE_ROOT/datafusion/core/tests/data/clickbench_hits_10.parquet]]}, projection=[HitColor, BrowserLanguage, BrowserCountry], file_type=parquet +01)ProjectionExec: expr=[1 as count(DISTINCT hits.HitColor), 1 as count(DISTINCT hits.BrowserCountry), 1 as count(DISTINCT hits.BrowserLanguage)] +02)--PlaceholderRowExec query III SELECT COUNT(DISTINCT "HitColor"), COUNT(DISTINCT "BrowserCountry"), COUNT(DISTINCT "BrowserLanguage") FROM hits;