From 768b3e90f261c7aea58bdb98dc698b90deeeae34 Mon Sep 17 00:00:00 2001 From: Kazantsev Maksim Date: Sun, 14 Dec 2025 16:24:01 +0400 Subject: [PATCH 1/3] impl map_from_entries --- native/core/src/execution/jni_api.rs | 2 + .../apache/comet/serde/QueryPlanSerde.scala | 3 +- .../scala/org/apache/comet/serde/maps.scala | 29 +++++++++++- .../comet/CometMapExpressionSuite.scala | 45 +++++++++++++++++++ 4 files changed, 77 insertions(+), 2 deletions(-) diff --git a/native/core/src/execution/jni_api.rs b/native/core/src/execution/jni_api.rs index a24d993059..4f53cea3e6 100644 --- a/native/core/src/execution/jni_api.rs +++ b/native/core/src/execution/jni_api.rs @@ -46,6 +46,7 @@ use datafusion_spark::function::datetime::date_add::SparkDateAdd; use datafusion_spark::function::datetime::date_sub::SparkDateSub; use datafusion_spark::function::hash::sha1::SparkSha1; use datafusion_spark::function::hash::sha2::SparkSha2; +use datafusion_spark::function::map::map_from_entries::MapFromEntries; use datafusion_spark::function::math::expm1::SparkExpm1; use datafusion_spark::function::string::char::CharFunc; use datafusion_spark::function::string::concat::SparkConcat; @@ -337,6 +338,7 @@ fn register_datafusion_spark_function(session_ctx: &SessionContext) { session_ctx.register_udf(ScalarUDF::new_from_impl(SparkSha1::default())); session_ctx.register_udf(ScalarUDF::new_from_impl(SparkConcat::default())); session_ctx.register_udf(ScalarUDF::new_from_impl(SparkBitwiseNot::default())); + session_ctx.register_udf(ScalarUDF::new_from_impl(MapFromEntries::default())); } /// Prepares arrow arrays for output. diff --git a/spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala b/spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala index 54df2f1688..a99cf3824b 100644 --- a/spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala +++ b/spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala @@ -125,7 +125,8 @@ object QueryPlanSerde extends Logging with CometExprShim { classOf[MapKeys] -> CometMapKeys, classOf[MapEntries] -> CometMapEntries, classOf[MapValues] -> CometMapValues, - classOf[MapFromArrays] -> CometMapFromArrays) + classOf[MapFromArrays] -> CometMapFromArrays, + classOf[MapFromEntries] -> CometMapFromEntries) private val structExpressions: Map[Class[_ <: Expression], CometExpressionSerde[_]] = Map( classOf[CreateNamedStruct] -> CometCreateNamedStruct, diff --git a/spark/src/main/scala/org/apache/comet/serde/maps.scala b/spark/src/main/scala/org/apache/comet/serde/maps.scala index 2e217f6af0..498aa3594c 100644 --- a/spark/src/main/scala/org/apache/comet/serde/maps.scala +++ b/spark/src/main/scala/org/apache/comet/serde/maps.scala @@ -19,9 +19,12 @@ package org.apache.comet.serde +import scala.annotation.tailrec + import org.apache.spark.sql.catalyst.expressions._ -import org.apache.spark.sql.types.{ArrayType, MapType} +import org.apache.spark.sql.types.{ArrayType, BinaryType, DataType, MapType, StructType} +import org.apache.comet.serde.CometArrayReverse.containsBinary import org.apache.comet.serde.QueryPlanSerde.{exprToProtoInternal, optExprWithInfo, scalarFunctionExprToProto, scalarFunctionExprToProtoWithReturnType} object CometMapKeys extends CometExpressionSerde[MapKeys] { @@ -89,3 +92,27 @@ object CometMapFromArrays extends CometExpressionSerde[MapFromArrays] { optExprWithInfo(mapFromArraysExpr, expr, expr.children: _*) } } + +object CometMapFromEntries extends CometScalarFunction[MapFromEntries]("map_from_entries") { + val keyUnsupportedReason = "Using BinaryType as Map keys is not allowed in map_from_entries" + val valueUnsupportedReason = "Using BinaryType as Map values is not allowed in map_from_entries" + + private def containsBinary(dataType: DataType): Boolean = { + dataType match { + case BinaryType => true + case StructType(fields) => fields.exists(field => containsBinary(field.dataType)) + case ArrayType(elementType, _) => containsBinary(elementType) + case _ => false + } + } + + override def getSupportLevel(expr: MapFromEntries): SupportLevel = { + if (containsBinary(expr.dataType.keyType)) { + return Incompatible(Some(keyUnsupportedReason)) + } + if (containsBinary(expr.dataType.valueType)) { + return Incompatible(Some(valueUnsupportedReason)) + } + Compatible(None) + } +} diff --git a/spark/src/test/scala/org/apache/comet/CometMapExpressionSuite.scala b/spark/src/test/scala/org/apache/comet/CometMapExpressionSuite.scala index 88c13391a6..01b9744ed6 100644 --- a/spark/src/test/scala/org/apache/comet/CometMapExpressionSuite.scala +++ b/spark/src/test/scala/org/apache/comet/CometMapExpressionSuite.scala @@ -25,7 +25,9 @@ import org.apache.hadoop.fs.Path import org.apache.spark.sql.CometTestBase import org.apache.spark.sql.functions._ import org.apache.spark.sql.internal.SQLConf +import org.apache.spark.sql.types.BinaryType +import org.apache.comet.serde.CometMapFromEntries import org.apache.comet.testing.{DataGenOptions, ParquetGenerator, SchemaGenOptions} class CometMapExpressionSuite extends CometTestBase { @@ -125,4 +127,47 @@ class CometMapExpressionSuite extends CometTestBase { } } + test("map_from_entries") { + withTempDir { dir => + val path = new Path(dir.toURI.toString, "test.parquet") + val filename = path.toString + val random = new Random(42) + withSQLConf(CometConf.COMET_ENABLED.key -> "false") { + val schemaGenOptions = + SchemaGenOptions( + generateArray = true, + generateStruct = true, + primitiveTypes = SchemaGenOptions.defaultPrimitiveTypes.filterNot(_ == BinaryType)) + val dataGenOptions = DataGenOptions(allowNull = false, generateNegativeZero = false) + ParquetGenerator.makeParquetFile( + random, + spark, + filename, + 100, + schemaGenOptions, + dataGenOptions) + } + val df = spark.read.parquet(filename) + df.createOrReplaceTempView("t1") + for (field <- df.schema.fieldNames) { + checkSparkAnswerAndOperator( + spark.sql(s"SELECT map_from_entries(array(struct($field as a, $field as b))) FROM t1")) + } + } + } + + test("map_from_entries - fallback for binary type") { + val table = "t2" + withTable(table) { + sql( + s"create table $table using parquet as select cast(array() as array) as c1 from range(10)") + checkSparkAnswerAndFallbackReason( + sql(s"select map_from_entries(array(struct(c1, 0))) from $table"), + CometMapFromEntries.keyUnsupportedReason) + checkSparkAnswerAndFallbackReason( + sql(s"select map_from_entries(array(struct(0, c1))) from $table"), + CometMapFromEntries.valueUnsupportedReason) + } + } + } From c68c3428676b5d991e7ba9e13464bf2ce1ec84e8 Mon Sep 17 00:00:00 2001 From: Kazantsev Maksim Date: Tue, 16 Dec 2025 16:10:43 +0400 Subject: [PATCH 2/3] Revert "impl map_from_entries" This reverts commit 768b3e90f261c7aea58bdb98dc698b90deeeae34. --- native/core/src/execution/jni_api.rs | 2 - .../apache/comet/serde/QueryPlanSerde.scala | 3 +- .../scala/org/apache/comet/serde/maps.scala | 29 +----------- .../comet/CometMapExpressionSuite.scala | 45 ------------------- 4 files changed, 2 insertions(+), 77 deletions(-) diff --git a/native/core/src/execution/jni_api.rs b/native/core/src/execution/jni_api.rs index 4f53cea3e6..a24d993059 100644 --- a/native/core/src/execution/jni_api.rs +++ b/native/core/src/execution/jni_api.rs @@ -46,7 +46,6 @@ use datafusion_spark::function::datetime::date_add::SparkDateAdd; use datafusion_spark::function::datetime::date_sub::SparkDateSub; use datafusion_spark::function::hash::sha1::SparkSha1; use datafusion_spark::function::hash::sha2::SparkSha2; -use datafusion_spark::function::map::map_from_entries::MapFromEntries; use datafusion_spark::function::math::expm1::SparkExpm1; use datafusion_spark::function::string::char::CharFunc; use datafusion_spark::function::string::concat::SparkConcat; @@ -338,7 +337,6 @@ fn register_datafusion_spark_function(session_ctx: &SessionContext) { session_ctx.register_udf(ScalarUDF::new_from_impl(SparkSha1::default())); session_ctx.register_udf(ScalarUDF::new_from_impl(SparkConcat::default())); session_ctx.register_udf(ScalarUDF::new_from_impl(SparkBitwiseNot::default())); - session_ctx.register_udf(ScalarUDF::new_from_impl(MapFromEntries::default())); } /// Prepares arrow arrays for output. diff --git a/spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala b/spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala index a99cf3824b..54df2f1688 100644 --- a/spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala +++ b/spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala @@ -125,8 +125,7 @@ object QueryPlanSerde extends Logging with CometExprShim { classOf[MapKeys] -> CometMapKeys, classOf[MapEntries] -> CometMapEntries, classOf[MapValues] -> CometMapValues, - classOf[MapFromArrays] -> CometMapFromArrays, - classOf[MapFromEntries] -> CometMapFromEntries) + classOf[MapFromArrays] -> CometMapFromArrays) private val structExpressions: Map[Class[_ <: Expression], CometExpressionSerde[_]] = Map( classOf[CreateNamedStruct] -> CometCreateNamedStruct, diff --git a/spark/src/main/scala/org/apache/comet/serde/maps.scala b/spark/src/main/scala/org/apache/comet/serde/maps.scala index 498aa3594c..2e217f6af0 100644 --- a/spark/src/main/scala/org/apache/comet/serde/maps.scala +++ b/spark/src/main/scala/org/apache/comet/serde/maps.scala @@ -19,12 +19,9 @@ package org.apache.comet.serde -import scala.annotation.tailrec - import org.apache.spark.sql.catalyst.expressions._ -import org.apache.spark.sql.types.{ArrayType, BinaryType, DataType, MapType, StructType} +import org.apache.spark.sql.types.{ArrayType, MapType} -import org.apache.comet.serde.CometArrayReverse.containsBinary import org.apache.comet.serde.QueryPlanSerde.{exprToProtoInternal, optExprWithInfo, scalarFunctionExprToProto, scalarFunctionExprToProtoWithReturnType} object CometMapKeys extends CometExpressionSerde[MapKeys] { @@ -92,27 +89,3 @@ object CometMapFromArrays extends CometExpressionSerde[MapFromArrays] { optExprWithInfo(mapFromArraysExpr, expr, expr.children: _*) } } - -object CometMapFromEntries extends CometScalarFunction[MapFromEntries]("map_from_entries") { - val keyUnsupportedReason = "Using BinaryType as Map keys is not allowed in map_from_entries" - val valueUnsupportedReason = "Using BinaryType as Map values is not allowed in map_from_entries" - - private def containsBinary(dataType: DataType): Boolean = { - dataType match { - case BinaryType => true - case StructType(fields) => fields.exists(field => containsBinary(field.dataType)) - case ArrayType(elementType, _) => containsBinary(elementType) - case _ => false - } - } - - override def getSupportLevel(expr: MapFromEntries): SupportLevel = { - if (containsBinary(expr.dataType.keyType)) { - return Incompatible(Some(keyUnsupportedReason)) - } - if (containsBinary(expr.dataType.valueType)) { - return Incompatible(Some(valueUnsupportedReason)) - } - Compatible(None) - } -} diff --git a/spark/src/test/scala/org/apache/comet/CometMapExpressionSuite.scala b/spark/src/test/scala/org/apache/comet/CometMapExpressionSuite.scala index 01b9744ed6..88c13391a6 100644 --- a/spark/src/test/scala/org/apache/comet/CometMapExpressionSuite.scala +++ b/spark/src/test/scala/org/apache/comet/CometMapExpressionSuite.scala @@ -25,9 +25,7 @@ import org.apache.hadoop.fs.Path import org.apache.spark.sql.CometTestBase import org.apache.spark.sql.functions._ import org.apache.spark.sql.internal.SQLConf -import org.apache.spark.sql.types.BinaryType -import org.apache.comet.serde.CometMapFromEntries import org.apache.comet.testing.{DataGenOptions, ParquetGenerator, SchemaGenOptions} class CometMapExpressionSuite extends CometTestBase { @@ -127,47 +125,4 @@ class CometMapExpressionSuite extends CometTestBase { } } - test("map_from_entries") { - withTempDir { dir => - val path = new Path(dir.toURI.toString, "test.parquet") - val filename = path.toString - val random = new Random(42) - withSQLConf(CometConf.COMET_ENABLED.key -> "false") { - val schemaGenOptions = - SchemaGenOptions( - generateArray = true, - generateStruct = true, - primitiveTypes = SchemaGenOptions.defaultPrimitiveTypes.filterNot(_ == BinaryType)) - val dataGenOptions = DataGenOptions(allowNull = false, generateNegativeZero = false) - ParquetGenerator.makeParquetFile( - random, - spark, - filename, - 100, - schemaGenOptions, - dataGenOptions) - } - val df = spark.read.parquet(filename) - df.createOrReplaceTempView("t1") - for (field <- df.schema.fieldNames) { - checkSparkAnswerAndOperator( - spark.sql(s"SELECT map_from_entries(array(struct($field as a, $field as b))) FROM t1")) - } - } - } - - test("map_from_entries - fallback for binary type") { - val table = "t2" - withTable(table) { - sql( - s"create table $table using parquet as select cast(array() as array) as c1 from range(10)") - checkSparkAnswerAndFallbackReason( - sql(s"select map_from_entries(array(struct(c1, 0))) from $table"), - CometMapFromEntries.keyUnsupportedReason) - checkSparkAnswerAndFallbackReason( - sql(s"select map_from_entries(array(struct(0, c1))) from $table"), - CometMapFromEntries.valueUnsupportedReason) - } - } - } From 113c8d0a24d2640303c345390051af39a66729a6 Mon Sep 17 00:00:00 2001 From: Kazantsev Maksim Date: Sat, 4 Jul 2026 18:16:05 +0400 Subject: [PATCH 3/3] migrate to native soundex --- native/core/src/execution/jni_api.rs | 2 + .../org/apache/comet/serde/strings.scala | 2 +- .../sql-tests/expressions/string/soundex.sql | 40 ++++++++++++++++++- 3 files changed, 42 insertions(+), 2 deletions(-) diff --git a/native/core/src/execution/jni_api.rs b/native/core/src/execution/jni_api.rs index 6dc00e9cf6..d345c33f51 100644 --- a/native/core/src/execution/jni_api.rs +++ b/native/core/src/execution/jni_api.rs @@ -618,6 +618,7 @@ fn register_datafusion_spark_function(session_ctx: &SessionContext) { session_ctx.register_udf(ScalarUDF::new_from_impl(SparkSec::default())); session_ctx.register_udf(ScalarUDF::new_from_impl(SparkRint::default())); session_ctx.register_udf(ScalarUDF::new_from_impl(SparkBitShift::right_unsigned())); + session_ctx.register_udf(ScalarUDF::new_from_impl(SparkSoundex::default())); } /// Prepares arrow arrays for output. @@ -1191,6 +1192,7 @@ pub extern "system" fn Java_org_apache_comet_Native_getRustThreadId( use crate::execution::columnar_to_row::ColumnarToRowContext; use arrow::ffi::{from_ffi, FFI_ArrowArray, FFI_ArrowSchema}; use datafusion_spark::function::math::bin::SparkBin; +use datafusion_spark::function::string::soundex::SparkSoundex; /// Initialize a native columnar to row converter. /// diff --git a/spark/src/main/scala/org/apache/comet/serde/strings.scala b/spark/src/main/scala/org/apache/comet/serde/strings.scala index 3186818c9c..00b339a195 100644 --- a/spark/src/main/scala/org/apache/comet/serde/strings.scala +++ b/spark/src/main/scala/org/apache/comet/serde/strings.scala @@ -730,7 +730,7 @@ object CometFormatString extends CometCodegenDispatch[FormatString] object CometOverlay extends CometCodegenDispatch[Overlay] -object CometSoundEx extends CometCodegenDispatch[SoundEx] +object CometSoundEx extends CometScalarFunction[SoundEx]("soundex") object CometStringLocate extends CometCodegenDispatch[StringLocate] diff --git a/spark/src/test/resources/sql-tests/expressions/string/soundex.sql b/spark/src/test/resources/sql-tests/expressions/string/soundex.sql index b1892de0ee..3d94980277 100644 --- a/spark/src/test/resources/sql-tests/expressions/string/soundex.sql +++ b/spark/src/test/resources/sql-tests/expressions/string/soundex.sql @@ -21,7 +21,41 @@ statement CREATE TABLE test_soundex(s string) USING parquet statement -INSERT INTO test_soundex VALUES ('Miller'), ('Robert'), ('Rupert'), (''), (NULL) +INSERT INTO test_soundex VALUES + ('Miller'), + (NULL), + (''), + ('Apache Spark'), + ('123'), + ('a123'), + ('Datafusion'), + ('Ashcroft'), + ('B1B'), + ('B B'), + ('BAB'), + ('#hello'), + (' hello'), + ('\thello'), + ('😀hello'), + ('1abc'), + ('A'), + ('BFPV'), + ('Robert'), + ('Rupert'), + ('robert'), + ('rObErT'), + ('Müller'), + ('Abcdefghijklmnop'), + ('Lloyd'), + ('BWB'), + ('BHB'), + ('Tymczak'), + ('Aeiou'), + ('1Robert'), + ('Smith-Jones'), + ('#'), + ('\nhello'), + (' '); query SELECT s, soundex(s) FROM test_soundex @@ -29,3 +63,7 @@ SELECT s, soundex(s) FROM test_soundex -- literal arguments query SELECT soundex('Miller'), soundex('Tymczak') + +-- additional test for soundex with concat +query +SELECT concat(soundex(' '), 'Spark') as concat_spark;