Skip to content

Use octet_length for ClickBench Q27/Q28 byte-length semantics #23475

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
kosiew wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Use octet_length for ClickBench Q27/Q28 byte-length semantics #23475

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
9725356
refactor: update query and benchmark metrics to use octet_length() in…
kosiew Jul 11, 2026
420c306
feat(sql): update SQL logic tests to use octet_length function
kosiew Jul 11, 2026
08d730a
feat: add comments to Q27 and Q28 query copies for byte-length semant…
kosiew Jul 11, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion benchmarks/queries/clickbench/queries/q27.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,5 @@
-- Must set for ClickBench hits_partitioned dataset. See https://github.com/apache/datafusion/issues/16591
-- set datafusion.execution.parquet.binary_as_string = true

SELECT "CounterID", AVG(length("URL")) AS l, COUNT(*) AS c FROM hits WHERE "URL" <> '' GROUP BY "CounterID" HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
-- DataFusion length(...) counts characters; use octet_length(...) for ClickBench byte-length semantics.
SELECT "CounterID", AVG(octet_length("URL")) AS l, COUNT(*) AS c FROM hits WHERE "URL" <> '' GROUP BY "CounterID" HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
3 changes: 2 additions & 1 deletion benchmarks/queries/clickbench/queries/q28.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,5 @@
-- Must set for ClickBench hits_partitioned dataset. See https://github.com/apache/datafusion/issues/16591
-- set datafusion.execution.parquet.binary_as_string = true

SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS k, AVG(length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM hits WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
-- DataFusion length(...) counts characters; use octet_length(...) for ClickBench byte-length semantics.
SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS k, AVG(octet_length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM hits WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
3 changes: 2 additions & 1 deletion benchmarks/sql_benchmarks/clickbench/benchmarks/q27.benchmark
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@ SELECT COUNT(*) > 0 from hits;
true

run
SELECT "CounterID", AVG(length("URL")) AS l, COUNT(*) AS c FROM hits WHERE "URL" <> '' GROUP BY "CounterID" HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
-- DataFusion length(...) counts characters; use octet_length(...) for ClickBench byte-length semantics.

@alamb alamb Jul 12, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah this duplication is unfortunate but I think @Omega359 is working to consolidate them all

SELECT "CounterID", AVG(octet_length("URL")) AS l, COUNT(*) AS c FROM hits WHERE "URL" <> '' GROUP BY "CounterID" HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;

result sql_benchmarks/clickbench/results/${CLICKBENCH_TYPE:-single}/q27.csv
3 changes: 2 additions & 1 deletion benchmarks/sql_benchmarks/clickbench/benchmarks/q28.benchmark
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@ SELECT COUNT(*) > 0 from hits;
true

run
SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS k, AVG(length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM hits WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
-- DataFusion length(...) counts characters; use octet_length(...) for ClickBench byte-length semantics.
SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS k, AVG(octet_length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM hits WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;

result sql_benchmarks/clickbench/results/${CLICKBENCH_TYPE:-single}/q28.csv
32 changes: 16 additions & 16 deletions datafusion/sqllogictest/test_files/clickbench.slt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -728,58 +728,58 @@ SELECT "SearchPhrase" FROM hits WHERE "SearchPhrase" <> '' ORDER BY "EventTime",

## Q27
query TT
EXPLAIN SELECT "CounterID", AVG(length("URL")) AS l, COUNT(*) AS c FROM hits WHERE "URL" <> '' GROUP BY "CounterID" HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
EXPLAIN SELECT "CounterID", AVG(octet_length("URL")) AS l, COUNT(*) AS c FROM hits WHERE "URL" <> '' GROUP BY "CounterID" HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
----
logical_plan
01)Sort: l DESC NULLS FIRST, fetch=25
02)--Projection: hits.CounterID, avg(length(hits.URL)) AS l, count(Int64(1)) AS count(*) AS c
02)--Projection: hits.CounterID, avg(octet_length(hits.URL)) AS l, count(Int64(1)) AS count(*) AS c
03)----Filter: count(Int64(1)) > Int64(100000)
04)------Aggregate: groupBy=[[hits.CounterID]], aggr=[[avg(CAST(character_length(hits.URL) AS length(hits.URL) AS Float64)), count(Int64(1))]]
04)------Aggregate: groupBy=[[hits.CounterID]], aggr=[[avg(CAST(octet_length(hits.URL) AS Float64)), count(Int64(1))]]
05)--------SubqueryAlias: hits
06)----------Filter: hits_raw.URL != Utf8View("")
07)------------TableScan: hits_raw projection=[CounterID, URL], partial_filters=[hits_raw.URL != Utf8View("")]
physical_plan
01)SortPreservingMergeExec: [l@1 DESC], fetch=25
02)--ProjectionExec: expr=[CounterID@0 as CounterID, avg(length(hits.URL))@1 as l, count(Int64(1))@2 as c]
03)----SortExec: TopK(fetch=25), expr=[avg(length(hits.URL))@1 DESC], preserve_partitioning=[true]
02)--ProjectionExec: expr=[CounterID@0 as CounterID, avg(octet_length(hits.URL))@1 as l, count(Int64(1))@2 as c]
03)----SortExec: TopK(fetch=25), expr=[avg(octet_length(hits.URL))@1 DESC], preserve_partitioning=[true]
04)------FilterExec: count(Int64(1))@2 > 100000
05)--------AggregateExec: mode=FinalPartitioned, gby=[CounterID@0 as CounterID], aggr=[avg(length(hits.URL)), count(Int64(1))]
05)--------AggregateExec: mode=FinalPartitioned, gby=[CounterID@0 as CounterID], aggr=[avg(octet_length(hits.URL)), count(Int64(1))]
06)----------RepartitionExec: partitioning=Hash([CounterID@0], 4), input_partitions=4
07)------------AggregateExec: mode=Partial, gby=[CounterID@0 as CounterID], aggr=[avg(length(hits.URL)), count(Int64(1))]
07)------------AggregateExec: mode=Partial, gby=[CounterID@0 as CounterID], aggr=[avg(octet_length(hits.URL)), count(Int64(1))]
08)--------------FilterExec: URL@1 !=
09)----------------RepartitionExec: partitioning=RoundRobinBatch(4), input_partitions=1
10)------------------DataSourceExec: file_groups={1 group: [[WORKSPACE_ROOT/datafusion/core/tests/data/clickbench_hits_10.parquet]]}, projection=[CounterID, URL], file_type=parquet, predicate=URL@13 != , pruning_predicate=URL_null_count@2 != row_count@3 AND (URL_min@0 != OR != URL_max@1), required_guarantees=[URL not in ()]

query IRI
SELECT "CounterID", AVG(length("URL")) AS l, COUNT(*) AS c FROM hits WHERE "URL" <> '' GROUP BY "CounterID" HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
SELECT "CounterID", AVG(octet_length("URL")) AS l, COUNT(*) AS c FROM hits WHERE "URL" <> '' GROUP BY "CounterID" HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
----

## Q28
query TT
EXPLAIN SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS k, AVG(length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM hits WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
EXPLAIN SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS k, AVG(octet_length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM hits WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
----
logical_plan
01)Sort: l DESC NULLS FIRST, fetch=25
02)--Projection: regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1")) AS k, avg(length(hits.Referer)) AS l, count(Int64(1)) AS count(*) AS c, min(hits.Referer)
02)--Projection: regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1")) AS k, avg(octet_length(hits.Referer)) AS l, count(Int64(1)) AS count(*) AS c, min(hits.Referer)
03)----Filter: count(Int64(1)) > Int64(100000)
04)------Aggregate: groupBy=[[regexp_replace(hits.Referer, Utf8View("^https?://(?:www\.)?([^/]+)/.*$"), Utf8View("\1")) AS regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))]], aggr=[[avg(CAST(character_length(hits.Referer) AS length(hits.Referer) AS Float64)), count(Int64(1)), min(hits.Referer)]]
04)------Aggregate: groupBy=[[regexp_replace(hits.Referer, Utf8View("^https?://(?:www\.)?([^/]+)/.*$"), Utf8View("\1")) AS regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))]], aggr=[[avg(CAST(octet_length(hits.Referer) AS Float64)), count(Int64(1)), min(hits.Referer)]]
05)--------SubqueryAlias: hits
06)----------Filter: hits_raw.Referer != Utf8View("")
07)------------TableScan: hits_raw projection=[Referer], partial_filters=[hits_raw.Referer != Utf8View("")]
physical_plan
01)SortPreservingMergeExec: [l@1 DESC], fetch=25
02)--ProjectionExec: expr=[regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))@0 as k, avg(length(hits.Referer))@1 as l, count(Int64(1))@2 as c, min(hits.Referer)@3 as min(hits.Referer)]
03)----SortExec: TopK(fetch=25), expr=[avg(length(hits.Referer))@1 DESC], preserve_partitioning=[true]
02)--ProjectionExec: expr=[regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))@0 as k, avg(octet_length(hits.Referer))@1 as l, count(Int64(1))@2 as c, min(hits.Referer)@3 as min(hits.Referer)]
03)----SortExec: TopK(fetch=25), expr=[avg(octet_length(hits.Referer))@1 DESC], preserve_partitioning=[true]
04)------FilterExec: count(Int64(1))@2 > 100000
05)--------AggregateExec: mode=FinalPartitioned, gby=[regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))@0 as regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))], aggr=[avg(length(hits.Referer)), count(Int64(1)), min(hits.Referer)]
05)--------AggregateExec: mode=FinalPartitioned, gby=[regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))@0 as regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))], aggr=[avg(octet_length(hits.Referer)), count(Int64(1)), min(hits.Referer)]
06)----------RepartitionExec: partitioning=Hash([regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))@0], 4), input_partitions=4
07)------------AggregateExec: mode=Partial, gby=[regexp_replace(Referer@0, ^https?://(?:www\.)?([^/]+)/.*$, \1) as regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))], aggr=[avg(length(hits.Referer)), count(Int64(1)), min(hits.Referer)]
07)------------AggregateExec: mode=Partial, gby=[regexp_replace(Referer@0, ^https?://(?:www\.)?([^/]+)/.*$, \1) as regexp_replace(hits.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))], aggr=[avg(octet_length(hits.Referer)), count(Int64(1)), min(hits.Referer)]
08)--------------FilterExec: Referer@0 !=
09)----------------RepartitionExec: partitioning=RoundRobinBatch(4), input_partitions=1
10)------------------DataSourceExec: file_groups={1 group: [[WORKSPACE_ROOT/datafusion/core/tests/data/clickbench_hits_10.parquet]]}, projection=[Referer], file_type=parquet, predicate=Referer@14 != , pruning_predicate=Referer_null_count@2 != row_count@3 AND (Referer_min@0 != OR != Referer_max@1), required_guarantees=[Referer not in ()]

query TRIT
SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS k, AVG(length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM hits WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS k, AVG(octet_length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM hits WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER BY l DESC LIMIT 25;
----

## Q29
Expand Down
Loading