Sql benchmark sync updated queries

benchmarks/sql_benchmarks/clickbench_extended/benchmarks/q13.benchmark

@@ -0,0 +1,19 @@
+name Q13
+group clickbench_extended
+subgroup ${CLICKBENCH_TYPE:-single}
+
+init sql_benchmarks/clickbench/init/set_config.sql
+
+load sql_benchmarks/clickbench/init/load-${CLICKBENCH_TYPE:-single}.sql
+
+assert I
+SELECT COUNT(*) > 0 from hits;
+----
+true
+
+run
+SELECT SUM("CounterID") AS counter_id_sum
+FROM hits
+WHERE "URL" < 'zzzz';
+
+result sql_benchmarks/clickbench_extended/results/q13.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q13.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q13
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (ROW_NUMBER top-2 per partition)
+SELECT id2, largest2_v2 FROM (
+    SELECT id2, v2 AS largest2_v2,
+           ROW_NUMBER() OVER (PARTITION BY id2 ORDER BY v2 DESC) AS order_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE order_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q13.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q14.benchmark

@@ -0,0 +1,26 @@
+subgroup window
+
+name Q14
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N partition cardinality sweep (id3 % N gives N distinct partitions).
+-- These exercise PartitionedTopKExec across cardinalities to validate it stays
+-- competitive with the SortExec+Filter baseline as partition count grows.
+-- Window Top-N: 100 partitions
+SELECT pk, largest2_v2 FROM (
+    SELECT id3 % 100 AS pk, v2 AS largest2_v2,
+           ROW_NUMBER() OVER (PARTITION BY id3 % 100 ORDER BY v2 DESC) AS order_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE order_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q14.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q15.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q15
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N: 1,000 partitions
+SELECT pk, largest2_v2 FROM (
+    SELECT id3 % 1000 AS pk, v2 AS largest2_v2,
+           ROW_NUMBER() OVER (PARTITION BY id3 % 1000 ORDER BY v2 DESC) AS order_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE order_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q15.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q16.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q16
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N: 10,000 partitions
+SELECT pk, largest2_v2 FROM (
+    SELECT id3 % 10000 AS pk, v2 AS largest2_v2,
+           ROW_NUMBER() OVER (PARTITION BY id3 % 10000 ORDER BY v2 DESC) AS order_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE order_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q16.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q17.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q17
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N: 100,000 partitions
+SELECT pk, largest2_v2 FROM (
+    SELECT id3 % 100000 AS pk, v2 AS largest2_v2,
+           ROW_NUMBER() OVER (PARTITION BY id3 % 100000 ORDER BY v2 DESC) AS order_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE order_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q17.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q18.benchmark

@@ -0,0 +1,27 @@
+subgroup window
+
+name Q18
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (RANK top-2 per partition, ~100 partitions)
+-- The RANK queries below mirror the ROW_NUMBER cardinality sweep
+-- above and add heavy-ties variants. RANK semantics retain boundary
+-- ties (`WHERE rk <= K` may keep more than K rows per partition), so
+-- this exercises PartitionedTopKRank's ties-Vec path.
+SELECT pk, largest_v2 FROM (
+    SELECT (id3 % 100) AS pk, v2 AS largest_v2,
+           RANK() OVER (PARTITION BY (id3 % 100) ORDER BY v2 DESC) AS rk_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE rk_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q18.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q19.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q19
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (RANK top-2 per partition, ~1K partitions)
+SELECT pkey, largest_v2 FROM (
+    SELECT (id3 % 1000) AS pkey, v2 AS largest_v2,
+           RANK() OVER (PARTITION BY (id3 % 1000) ORDER BY v2 DESC) AS rk_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE rk_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q19.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q20.benchmark

@@ -0,0 +1,25 @@
+subgroup window
+
+name Q20
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (RANK top-2 per partition, ~1K partitions, heavy ties)
+-- v2 % 10 forces 10 distinct OBY values, so most rows tie at the boundary
+-- and exercise PartitionedTopKRank's ties-Vec path.
+SELECT pkey, largest_v2 FROM (
+    SELECT (id3 % 1000) AS pkey, v2 AS largest_v2,
+           RANK() OVER (PARTITION BY (id3 % 1000) ORDER BY (v2 % 10) DESC) AS rk_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE rk_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q20.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q21.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q21
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (RANK top-2 per partition, ~10K partitions, low ties)
+SELECT id2, largest_v2 FROM (
+    SELECT id2, v2 AS largest_v2,
+           RANK() OVER (PARTITION BY id2 ORDER BY v2 DESC) AS rk_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE rk_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q21.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q22.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q22
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (RANK top-2 per partition, ~10K partitions, heavy ties)
+SELECT id2, largest_v2 FROM (
+    SELECT id2, v2 AS largest_v2,
+           RANK() OVER (PARTITION BY id2 ORDER BY (v2 % 10) DESC) AS rk_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE rk_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q22.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q23.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q23
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (RANK top-2 per partition, ~100K partitions)
+SELECT pk, largest_v2 FROM (
+    SELECT (id3 % 100000) AS pk, v2 AS largest_v2,
+           RANK() OVER (PARTITION BY (id3 % 100000) ORDER BY v2 DESC) AS rk_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE rk_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q23.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q24.benchmark

@@ -0,0 +1,28 @@
+subgroup window
+
+name Q24
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (DENSE_RANK top-2 per partition, ~100 partitions)
+-- The DENSE_RANK queries below mirror the RANK cardinality sweep above.
+-- DENSE_RANK semantics keep every row whose ORDER BY value is among the
+-- K distinct-greatest values in the partition, so total kept per partition
+-- is unbounded in rows-per-distinct-value — exercises PartitionedTopKDenseRank's
+-- HashMap-of-groups path.
+SELECT pk, largest_v2 FROM (
+    SELECT (id3 % 100) AS pk, v2 AS largest_v2,
+           DENSE_RANK() OVER (PARTITION BY (id3 % 100) ORDER BY v2 DESC) AS dr_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE dr_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q24.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q25.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q25
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (DENSE_RANK top-2 per partition, ~1K partitions)
+SELECT pkey, largest_v2 FROM (
+    SELECT (id3 % 1000) AS pkey, v2 AS largest_v2,
+           DENSE_RANK() OVER (PARTITION BY (id3 % 1000) ORDER BY v2 DESC) AS dr_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE dr_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q25.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q26.benchmark

@@ -0,0 +1,26 @@
+subgroup window
+
+name Q26
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (DENSE_RANK top-2 per partition, ~1K partitions, heavy ties)
+-- v2 % 10 forces 10 distinct OBY values; most rows share the top-2 distinct
+-- values so appends dominate — exercises the "Case A" append-to-existing-Vec
+-- fast path in PartitionedTopKDenseRank.
+SELECT pkey, largest_v2 FROM (
+    SELECT (id3 % 1000) AS pkey, v2 AS largest_v2,
+           DENSE_RANK() OVER (PARTITION BY (id3 % 1000) ORDER BY (v2 % 10) DESC) AS dr_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE dr_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q26.csv

benchmarks/sql_benchmarks/h2o/benchmarks/window/q27.benchmark

@@ -0,0 +1,23 @@
+subgroup window
+
+name Q27
+group h2o
+
+echo Loading ${H2O_BENCH_SIZE:-small} window ${H2O_FILE_TYPE:-csv} h2o data
+
+load sql_benchmarks/h2o/init/load_window_${H2O_BENCH_SIZE:-small}_${H2O_FILE_TYPE:-csv}.sql
+
+assert I
+SELECT COUNT(*) > 0 FROM x
+----
+true
+
+run
+-- Window Top-N (DENSE_RANK top-2 per partition, ~10K partitions, low ties)
+SELECT id2, largest_v2 FROM (
+    SELECT id2, v2 AS largest_v2,
+           DENSE_RANK() OVER (PARTITION BY id2 ORDER BY v2 DESC) AS dr_v2
+    FROM x WHERE v2 IS NOT NULL
+) sub_query WHERE dr_v2 <= 2;
+
+result sql_benchmarks/h2o/results/window/${H2O_BENCH_SIZE:-small}/q27.csv