This is an automated email from the ASF dual-hosted git repository.
Jackie-Jiang pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/pinot.git
The following commit(s) were added to refs/heads/master by this push:
new 5e914c979b3 Support optional parameters argument for tuple sketch
aggregations in the multi-stage engine (#19312)
5e914c979b3 is described below
commit 5e914c979b355c0e8c9fd4cd6f401f5f899594b5
Author: David Cromberge <[email protected]>
AuthorDate: Thu Aug 27 19:23:30 2026 +0200
Support optional parameters argument for tuple sketch aggregations in the
multi-stage engine (#19312)
---
.../IntegerTupleSketchAggregationFunction.java | 6 ++---
.../integration/tests/custom/TupleSketchTest.java | 15 +++++++++++
.../pinot/segment/spi/AggregationFunctionType.java | 18 ++++++++++---
.../segment/spi/AggregationFunctionTypeTest.java | 30 ++++++++++++++++++++++
4 files changed, 62 insertions(+), 7 deletions(-)
diff --git
a/pinot-core/src/main/java/org/apache/pinot/core/query/aggregation/function/IntegerTupleSketchAggregationFunction.java
b/pinot-core/src/main/java/org/apache/pinot/core/query/aggregation/function/IntegerTupleSketchAggregationFunction.java
index a56ce257f71..c7366e8227b 100644
---
a/pinot-core/src/main/java/org/apache/pinot/core/query/aggregation/function/IntegerTupleSketchAggregationFunction.java
+++
b/pinot-core/src/main/java/org/apache/pinot/core/query/aggregation/function/IntegerTupleSketchAggregationFunction.java
@@ -59,16 +59,16 @@ import org.apache.pinot.spi.utils.CommonConstants;
/// Usage examples:
///
/// - Simple union (1 or 2 arguments): main expression to aggregate on,
followed by an optional Tuple sketch size
-/// argument. The second argument is the sketch lgK – the given log_base2 of
k, and defaults to 16.
+/// argument. The second argument is the nominal entries, and defaults to
16384.
/// The "raw" equivalents return serialised sketches in base64-encoded
strings.
///
/// DISTINCT_COUNT_TUPLE_SKETCH(col)
///
-/// DISTINCT_COUNT_TUPLE_SKETCH(col, 12)
+/// DISTINCT_COUNT_TUPLE_SKETCH(col, 16384)
///
/// DISTINCT_COUNT_RAW_INTEGER_SUM_TUPLE_SKETCH(col)
///
-/// DISTINCT_COUNT_RAW_INTEGER_SUM_TUPLE_SKETCH(col, 12)
+/// DISTINCT_COUNT_RAW_INTEGER_SUM_TUPLE_SKETCH(col, 16384)
/// - Extracting a cardinality estimate from a CPC sketch:
///
/// GET_INT_TUPLE_SKETCH_ESTIMATE(sketch_bytes)
diff --git
a/pinot-integration-tests/src/test/java/org/apache/pinot/integration/tests/custom/TupleSketchTest.java
b/pinot-integration-tests/src/test/java/org/apache/pinot/integration/tests/custom/TupleSketchTest.java
index 15789c501bf..000c7b26d96 100644
---
a/pinot-integration-tests/src/test/java/org/apache/pinot/integration/tests/custom/TupleSketchTest.java
+++
b/pinot-integration-tests/src/test/java/org/apache/pinot/integration/tests/custom/TupleSketchTest.java
@@ -71,6 +71,21 @@ public class TupleSketchTest extends
CustomDataQueryClusterIntegrationTest {
assertEquals(Double.valueOf(deserializedSketch.getEstimate()).longValue(),
distinctCount);
assertTrue(jsonNode.get("resultTable").get("rows").get(0).get(2).asLong()
> 0);
assertTrue(jsonNode.get("resultTable").get("rows").get(0).get(3).asLong()
> 0);
+
+ String parameterizedQuery =
+ String.format(
+ "SELECT DISTINCT_COUNT_TUPLE_SKETCH(%s,
'nominalEntries=16384;accumulatorThreshold=10'), "
+ + "DISTINCT_COUNT_RAW_INTEGER_SUM_TUPLE_SKETCH(%s,
'nominalEntries=16384;accumulatorThreshold=10'), "
+ + "SUM_VALUES_INTEGER_SUM_TUPLE_SKETCH(%s,
'nominalEntries=16384;accumulatorThreshold=10'), "
+ + "AVG_VALUE_INTEGER_SUM_TUPLE_SKETCH(%s,
'nominalEntries=16384;accumulatorThreshold=10') FROM %s",
+ MET_TUPLE_SKETCH_BYTES, MET_TUPLE_SKETCH_BYTES,
MET_TUPLE_SKETCH_BYTES, MET_TUPLE_SKETCH_BYTES,
+ getTableName());
+ jsonNode = postQuery(parameterizedQuery);
+ assertTrue(jsonNode.get("resultTable").get("rows").get(0).get(0).asLong()
> 0);
+ assertTrue(ObjectSerDeUtils.DATA_SKETCH_INT_TUPLE_SER_DE.deserialize(
+
Base64.getDecoder().decode(jsonNode.get("resultTable").get("rows").get(0).get(1).asText())).getEstimate()
> 0);
+ assertTrue(jsonNode.get("resultTable").get("rows").get(0).get(2).asLong()
> 0);
+ assertTrue(jsonNode.get("resultTable").get("rows").get(0).get(3).asLong()
> 0);
}
@Test(dataProvider = "useV2QueryEngine")
diff --git
a/pinot-segment-spi/src/main/java/org/apache/pinot/segment/spi/AggregationFunctionType.java
b/pinot-segment-spi/src/main/java/org/apache/pinot/segment/spi/AggregationFunctionType.java
index a6e9900759e..080138b3a32 100644
---
a/pinot-segment-spi/src/main/java/org/apache/pinot/segment/spi/AggregationFunctionType.java
+++
b/pinot-segment-spi/src/main/java/org/apache/pinot/segment/spi/AggregationFunctionType.java
@@ -107,12 +107,13 @@ public enum AggregationFunctionType {
DISTINCTCOUNTTHETASKETCH("distinctCountThetaSketch", ReturnTypes.BIGINT,
OperandTypes.ONE_OR_MORE, SqlTypeName.OTHER),
DISTINCTCOUNTRAWTHETASKETCH("distinctCountRawThetaSketch",
ReturnTypes.VARCHAR, OperandTypes.ONE_OR_MORE,
SqlTypeName.OTHER),
- DISTINCTCOUNTTUPLESKETCH("distinctCountTupleSketch", ReturnTypes.BIGINT,
OperandTypes.BINARY, SqlTypeName.OTHER),
+ DISTINCTCOUNTTUPLESKETCH("distinctCountTupleSketch", ReturnTypes.BIGINT,
TupleSketchOperands.CHECKER,
+ SqlTypeName.OTHER),
DISTINCTCOUNTRAWINTEGERSUMTUPLESKETCH("distinctCountRawIntegerSumTupleSketch",
ReturnTypes.VARCHAR,
- OperandTypes.BINARY, SqlTypeName.OTHER),
- SUMVALUESINTEGERSUMTUPLESKETCH("sumValuesIntegerSumTupleSketch",
ReturnTypes.BIGINT, OperandTypes.BINARY,
+ TupleSketchOperands.CHECKER, SqlTypeName.OTHER),
+ SUMVALUESINTEGERSUMTUPLESKETCH("sumValuesIntegerSumTupleSketch",
ReturnTypes.BIGINT, TupleSketchOperands.CHECKER,
SqlTypeName.OTHER),
- AVGVALUEINTEGERSUMTUPLESKETCH("avgValueIntegerSumTupleSketch",
ReturnTypes.BIGINT, OperandTypes.BINARY,
+ AVGVALUEINTEGERSUMTUPLESKETCH("avgValueIntegerSumTupleSketch",
ReturnTypes.BIGINT, TupleSketchOperands.CHECKER,
SqlTypeName.OTHER),
DISTINCTCOUNTCPCSKETCH("distinctCountCPCSketch", ReturnTypes.BIGINT,
OperandTypes.family(List.of(SqlTypeFamily.ANY, SqlTypeFamily.ANY), i ->
i == 1), SqlTypeName.OTHER),
@@ -241,6 +242,15 @@ public enum AggregationFunctionType {
SqlTypeName.OTHER),
TIMESERIESAGGREGATE("timeSeriesAggregate", SqlTypeName.OTHER,
SqlTypeName.OTHER);
+ /// Nested because an enum constant's arguments cannot reference a static
field of its own enum.
+ private static final class TupleSketchOperands {
+ static final SqlOperandTypeChecker CHECKER = OperandTypes.or(
+ OperandTypes.family(SqlTypeFamily.BINARY),
+ OperandTypes.family(SqlTypeFamily.BINARY, SqlTypeFamily.CHARACTER),
+ OperandTypes.family(SqlTypeFamily.BINARY, SqlTypeFamily.INTEGER)
+ );
+ }
+
private static final Set<String> NAMES = Arrays.stream(values())
.flatMap(func -> Stream.of(func.name(), func.getName(),
func.getName().toLowerCase()))
.collect(Collectors.toSet());
diff --git
a/pinot-segment-spi/src/test/java/org/apache/pinot/segment/spi/AggregationFunctionTypeTest.java
b/pinot-segment-spi/src/test/java/org/apache/pinot/segment/spi/AggregationFunctionTypeTest.java
index 09de660eebd..183458395cc 100644
---
a/pinot-segment-spi/src/test/java/org/apache/pinot/segment/spi/AggregationFunctionTypeTest.java
+++
b/pinot-segment-spi/src/test/java/org/apache/pinot/segment/spi/AggregationFunctionTypeTest.java
@@ -18,8 +18,12 @@
*/
package org.apache.pinot.segment.spi;
+import org.apache.calcite.sql.SqlOperandCountRange;
+import org.apache.calcite.sql.fun.SqlStdOperatorTable;
+import org.apache.calcite.sql.type.SqlOperandTypeChecker;
import org.testng.annotations.Test;
+import static org.testng.Assert.assertEquals;
import static org.testng.Assert.assertFalse;
import static org.testng.Assert.assertTrue;
@@ -38,4 +42,30 @@ public class AggregationFunctionTypeTest {
assertTrue(AggregationFunctionType.isAggregationFunction("percentileest90"));
assertFalse(AggregationFunctionType.isAggregationFunction("toEpochSeconds"));
}
+
+ @Test
+ public void testTupleSketchOperandTypes() {
+ AggregationFunctionType[] tupleSketchFunctions = {
+ AggregationFunctionType.DISTINCTCOUNTTUPLESKETCH,
+ AggregationFunctionType.DISTINCTCOUNTRAWINTEGERSUMTUPLESKETCH,
+ AggregationFunctionType.SUMVALUESINTEGERSUMTUPLESKETCH,
+ AggregationFunctionType.AVGVALUEINTEGERSUMTUPLESKETCH
+ };
+
+ for (AggregationFunctionType functionType : tupleSketchFunctions) {
+ SqlOperandTypeChecker operandTypeChecker =
functionType.getOperandTypeChecker();
+
+ SqlOperandCountRange countRange =
operandTypeChecker.getOperandCountRange();
+ assertEquals(countRange.getMin(), 1, functionType.name());
+ assertEquals(countRange.getMax(), 2, functionType.name());
+ assertFalse(countRange.isValidCount(3), functionType.name());
+
+ String signatures =
+ operandTypeChecker.getAllowedSignatures(SqlStdOperatorTable.COUNT,
functionType.name());
+ assertTrue(signatures.contains("BINARY"), functionType.name() + ": " +
signatures);
+ assertTrue(signatures.contains("CHARACTER"), functionType.name() + ": "
+ signatures);
+ assertTrue(signatures.contains("INTEGER"), functionType.name() + ": " +
signatures);
+ assertFalse(signatures.contains("ANY"), functionType.name() + ": " +
signatures);
+ }
+ }
}
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
