Merge pull request #18 from apache/rename_packages

Rename packages
apache · Jan 29, 2025 · 255b048 · 255b048
2 parents 6779cde + 050710f
commit 255b048
Show file tree

Hide file tree

Showing 18 changed files with 64 additions and 39 deletions.
diff --git a/...istrar/DatasketchesFunctionRegistry.scala → ...common/DatasketchesFunctionRegistry.scala b/...istrar/DatasketchesFunctionRegistry.scala → ...common/DatasketchesFunctionRegistry.scala
@@ -15,7 +15,7 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.registrar
+package org.apache.spark.sql.datasketches.common
 
 import org.apache.spark.sql.SparkSession
 import org.apache.spark.sql.catalyst.FunctionIdentifier

diff --git a/...trar/DatasketchesScalaFunctionsBase.scala → ...mmon/DatasketchesScalaFunctionsBase.scala b/...trar/DatasketchesScalaFunctionsBase.scala → ...mmon/DatasketchesScalaFunctionsBase.scala
@@ -15,10 +15,11 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql
+package org.apache.spark.sql.datasketches.common
 
 import org.apache.spark.sql.catalyst.expressions.Expression
 import org.apache.spark.sql.catalyst.expressions.aggregate.AggregateFunction
+import org.apache.spark.sql.Column
 
 // this interfact provides a few helper methods defines and maps all the variants of each function invocation, analagous
 // to the functions object in core Spark's org.apache.spark.sql.functions

diff --git a/...k/sql/registrar/KllFunctionRegistry.scala → ...atasketches/kll/KllFunctionRegistry.scala b/...k/sql/registrar/KllFunctionRegistry.scala → ...atasketches/kll/KllFunctionRegistry.scala
@@ -15,13 +15,15 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.registrar
+package org.apache.spark.sql.datasketches.kll
 
 import org.apache.spark.sql.catalyst.analysis.FunctionRegistry.FunctionBuilder
 import org.apache.spark.sql.catalyst.expressions.{ExpressionInfo}
 
-import org.apache.spark.sql.aggregate.{KllDoublesSketchAggBuild, KllDoublesSketchAggMerge}
-import org.apache.spark.sql.expressions.{KllDoublesSketchGetMin, KllDoublesSketchGetMax, KllDoublesSketchGetPmf, KllDoublesSketchGetCdf}
+import org.apache.spark.sql.datasketches.common.DatasketchesFunctionRegistry
+import org.apache.spark.sql.datasketches.kll.aggregate.{KllDoublesSketchAggBuild, KllDoublesSketchAggMerge}
+import org.apache.spark.sql.datasketches.kll.expressions.{KllDoublesSketchGetMin, KllDoublesSketchGetMax, KllDoublesSketchGetPmf, KllDoublesSketchGetCdf}
+import org.apache.spark.sql.datasketches.common.DatasketchesFunctionRegistry
 
 object KllFunctionRegistry extends DatasketchesFunctionRegistry {
   override val expressions: Map[String, (ExpressionInfo, FunctionBuilder)] = Map(

diff --git a/.../aggregate/KllDoublesSketchAggBuild.scala → .../aggregate/KllDoublesSketchAggBuild.scala b/.../aggregate/KllDoublesSketchAggBuild.scala → .../aggregate/KllDoublesSketchAggBuild.scala
@@ -15,14 +15,16 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.aggregate
+package org.apache.spark.sql.datasketches.kll.aggregate
 
 import org.apache.datasketches.kll.{KllSketch, KllDoublesSketch}
+import org.apache.spark.sql.datasketches.kll.types.KllDoublesSketchType
+
 import org.apache.spark.sql.catalyst.InternalRow
 import org.apache.spark.sql.catalyst.expressions.{ExpectsInputTypes, Expression, ExpressionDescription, Literal}
 import org.apache.spark.sql.catalyst.expressions.aggregate.TypedImperativeAggregate
 import org.apache.spark.sql.catalyst.trees.BinaryLike
-import org.apache.spark.sql.types.{AbstractDataType, DataType, IntegerType, LongType, NumericType, FloatType, DoubleType, KllDoublesSketchType}
+import org.apache.spark.sql.types.{AbstractDataType, DataType, IntegerType, LongType, NumericType, FloatType, DoubleType}
 import org.apache.spark.sql.catalyst.analysis.TypeCheckResult
 
 /**

diff --git a/.../aggregate/KllDoublesSketchAggMerge.scala → .../aggregate/KllDoublesSketchAggMerge.scala b/.../aggregate/KllDoublesSketchAggMerge.scala → .../aggregate/KllDoublesSketchAggMerge.scala
@@ -15,16 +15,17 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.aggregate
+package org.apache.spark.sql.datasketches.kll.aggregate
 
 import org.apache.datasketches.memory.Memory
 import org.apache.datasketches.kll.{KllSketch, KllDoublesSketch}
+import org.apache.spark.sql.datasketches.kll.types.KllDoublesSketchType
 
 import org.apache.spark.sql.catalyst.InternalRow
 import org.apache.spark.sql.catalyst.expressions.{ExpectsInputTypes, Expression, ExpressionDescription, Literal}
 import org.apache.spark.sql.catalyst.expressions.aggregate.TypedImperativeAggregate
 import org.apache.spark.sql.catalyst.trees.BinaryLike
-import org.apache.spark.sql.types.{AbstractDataType, DataType, IntegerType, KllDoublesSketchType}
+import org.apache.spark.sql.types.{AbstractDataType, DataType, IntegerType}
 import org.apache.spark.sql.catalyst.analysis.TypeCheckResult
 
 /**

diff --git a/...essions/KllDoublesSketchExpressions.scala → ...essions/KllDoublesSketchExpressions.scala b/...essions/KllDoublesSketchExpressions.scala → ...essions/KllDoublesSketchExpressions.scala
@@ -15,12 +15,12 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.expressions
+package org.apache.spark.sql.datasketches.kll.expressions
 
 import org.apache.datasketches.memory.Memory
 import org.apache.datasketches.kll.KllDoublesSketch
 import org.apache.datasketches.quantilescommon.QuantileSearchCriteria
-import org.apache.spark.sql.types.KllDoublesSketchType
+import org.apache.spark.sql.datasketches.kll.types.KllDoublesSketchType
 
 import org.apache.spark.sql.types.{AbstractDataType, ArrayType, BooleanType, DataType, DoubleType}
 import org.apache.spark.sql.catalyst.expressions.{Expression, ExpressionDescription, ExpectsInputTypes, ImplicitCastInputTypes}

diff --git a/...egistrar/functions_datasketches_kll.scala → ...park/sql/datasketches/kll/functions.scala b/...egistrar/functions_datasketches_kll.scala → ...park/sql/datasketches/kll/functions.scala
@@ -15,16 +15,18 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql
+package org.apache.spark.sql.datasketches.kll
 
+import org.apache.spark.sql.Column
 import org.apache.spark.sql.catalyst.expressions.Literal
 import org.apache.spark.sql.functions.lit
 import org.apache.spark.sql.types.{ArrayType, BooleanType, DoubleType}
 
-import org.apache.spark.sql.aggregate.{KllDoublesSketchAggMerge, KllDoublesSketchAggBuild}
-import org.apache.spark.sql.expressions.{KllDoublesSketchGetMin, KllDoublesSketchGetMax, KllDoublesSketchGetPmfCdf}
+import org.apache.spark.sql.datasketches.common.DatasketchesScalaFunctionBase
+import org.apache.spark.sql.datasketches.kll.aggregate.{KllDoublesSketchAggMerge, KllDoublesSketchAggBuild}
+import org.apache.spark.sql.datasketches.kll.expressions.{KllDoublesSketchGetMin, KllDoublesSketchGetMax, KllDoublesSketchGetPmfCdf}
 
-object functions_datasketches_kll extends DatasketchesScalaFunctionBase {
+object functions extends DatasketchesScalaFunctionBase {
 
   // build sketch
   def kll_sketch_double_agg_build(expr: Column, k: Column): Column = withAggregateFunction {

diff --git a/.../sql/kll/types/KllDoublesSketchType.scala → ...ches/kll/types/KllDoublesSketchType.scala b/.../sql/kll/types/KllDoublesSketchType.scala → ...ches/kll/types/KllDoublesSketchType.scala
@@ -15,11 +15,12 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.types
+package org.apache.spark.sql.datasketches.kll.types
 
 import org.apache.spark.sql.functions.udf
 import org.apache.datasketches.kll.KllDoublesSketch
 import org.apache.datasketches.memory.Memory
+import org.apache.spark.sql.types.{DataType, DataTypes, UDTRegistration, UserDefinedType}
 
 class KllDoublesSketchType extends UserDefinedType[KllDoublesSketch] with Serializable {
   override def sqlType: DataType = DataTypes.BinaryType

diff --git a/...sql/registrar/ThetaFunctionRegistry.scala → ...ketches/theta/ThetaFunctionRegistry.scala b/...sql/registrar/ThetaFunctionRegistry.scala → ...ketches/theta/ThetaFunctionRegistry.scala
@@ -15,13 +15,15 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.registrar
+package org.apache.spark.sql.datasketches.theta
 
 import org.apache.spark.sql.catalyst.analysis.FunctionRegistry.FunctionBuilder
 import org.apache.spark.sql.catalyst.expressions.{ExpressionInfo}
 
-import org.apache.spark.sql.aggregate.{ThetaSketchAggBuild, ThetaSketchAggUnion}
-import org.apache.spark.sql.expressions.ThetaSketchGetEstimate
+import org.apache.spark.sql.datasketches.common.DatasketchesFunctionRegistry
+import org.apache.spark.sql.datasketches.theta.aggregate.{ThetaSketchAggBuild, ThetaSketchAggUnion}
+import org.apache.spark.sql.datasketches.theta.expressions.ThetaSketchGetEstimate
+import org.apache.spark.sql.datasketches.common.DatasketchesFunctionRegistry
 
 object ThetaFunctionRegistry extends DatasketchesFunctionRegistry {
   override val expressions: Map[String, (ExpressionInfo, FunctionBuilder)] = Map(

diff --git a/...theta/aggregate/ThetaSketchAggBuild.scala → ...theta/aggregate/ThetaSketchAggBuild.scala b/...theta/aggregate/ThetaSketchAggBuild.scala → ...theta/aggregate/ThetaSketchAggBuild.scala
@@ -15,14 +15,16 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.aggregate
+package org.apache.spark.sql.datasketches.theta.aggregate
 
 import org.apache.datasketches.theta.{UpdateSketch, SetOperation}
+import org.apache.spark.sql.datasketches.theta.types.{ThetaSketchType, ThetaSketchWrapper}
+
 import org.apache.spark.sql.catalyst.InternalRow
 import org.apache.spark.sql.catalyst.expressions.{ExpectsInputTypes, Expression, ExpressionDescription, Literal}
 import org.apache.spark.sql.catalyst.expressions.aggregate.TypedImperativeAggregate
 import org.apache.spark.sql.catalyst.trees.BinaryLike
-import org.apache.spark.sql.types.{AbstractDataType, DataType, IntegerType, LongType, NumericType, FloatType, DoubleType, ThetaSketchWrapper, ThetaSketchType}
+import org.apache.spark.sql.types.{AbstractDataType, DataType, IntegerType, LongType, NumericType, FloatType, DoubleType}
 
 /**
  * The ThetaSketchBuild function creates a Theta sketch from a column of values

diff --git a/...theta/aggregate/ThetaSketchAggUnion.scala → ...theta/aggregate/ThetaSketchAggUnion.scala b/...theta/aggregate/ThetaSketchAggUnion.scala → ...theta/aggregate/ThetaSketchAggUnion.scala
@@ -15,16 +15,17 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.aggregate
+package org.apache.spark.sql.datasketches.theta.aggregate
 
 import org.apache.datasketches.memory.Memory
 import org.apache.datasketches.theta.{Sketch, SetOperation}
+import org.apache.spark.sql.datasketches.theta.types.{ThetaSketchType, ThetaSketchWrapper}
 
 import org.apache.spark.sql.catalyst.InternalRow
 import org.apache.spark.sql.catalyst.expressions.{ExpectsInputTypes, Expression, ExpressionDescription, Literal}
 import org.apache.spark.sql.catalyst.expressions.aggregate.TypedImperativeAggregate
 import org.apache.spark.sql.catalyst.trees.BinaryLike
-import org.apache.spark.sql.types.{AbstractDataType, DataType, IntegerType, ThetaSketchWrapper, ThetaSketchType}
+import org.apache.spark.sql.types.{AbstractDataType, DataType, IntegerType}
 
 /**
  * Theta Union operation.

diff --git a/.../theta/expressions/ThetaExpressions.scala → .../theta/expressions/ThetaExpressions.scala b/.../theta/expressions/ThetaExpressions.scala → .../theta/expressions/ThetaExpressions.scala
@@ -15,16 +15,17 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.expressions
+package org.apache.spark.sql.datasketches.theta.expressions
 
 import org.apache.datasketches.memory.Memory
 import org.apache.datasketches.theta.Sketch
+import org.apache.spark.sql.datasketches.theta.types.ThetaSketchType
 
 import org.apache.spark.sql.catalyst.expressions.{Expression, ExpectsInputTypes, UnaryExpression}
 import org.apache.spark.sql.catalyst.expressions.NullIntolerant
 import org.apache.spark.sql.catalyst.expressions.ExpressionDescription
 import org.apache.spark.sql.catalyst.expressions.codegen.{CodeBlock, CodegenContext, ExprCode}
-import org.apache.spark.sql.types.{AbstractDataType, DataType, DoubleType, ThetaSketchType}
+import org.apache.spark.sql.types.{AbstractDataType, DataType, DoubleType}
 
 @ExpressionDescription(
   usage = """

diff --git a/...istrar/functions_datasketches_theta.scala → ...rk/sql/datasketches/theta/functions.scala b/...istrar/functions_datasketches_theta.scala → ...rk/sql/datasketches/theta/functions.scala
@@ -15,14 +15,17 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql
+package org.apache.spark.sql.datasketches.theta
 
+import org.apache.spark.sql.Column
 import org.apache.spark.sql.functions.lit
 
-import org.apache.spark.sql.aggregate.{ThetaSketchAggBuild, ThetaSketchAggUnion}
-import org.apache.spark.sql.expressions.ThetaSketchGetEstimate
+import org.apache.spark.sql.datasketches.common.DatasketchesScalaFunctionBase
+import org.apache.spark.sql.datasketches.theta.aggregate.{ThetaSketchAggBuild, ThetaSketchAggUnion}
+import org.apache.spark.sql.datasketches.theta.expressions.ThetaSketchGetEstimate
+import org.apache.spark.sql.datasketches.common.DatasketchesScalaFunctionBase
 
-object functions_datasketches_theta extends DatasketchesScalaFunctionBase {
+object functions extends DatasketchesScalaFunctionBase {
   def theta_sketch_agg_build(column: Column, lgk: Int): Column = withAggregateFunction {
     new ThetaSketchAggBuild(column.expr, lgk)
   }

diff --git a/...ark/sql/theta/types/ThetaSketchType.scala → ...ketches/theta/types/ThetaSketchType.scala b/...ark/sql/theta/types/ThetaSketchType.scala → ...ketches/theta/types/ThetaSketchType.scala
@@ -15,7 +15,9 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.types
+package org.apache.spark.sql.datasketches.theta.types
+
+import org.apache.spark.sql.types.{DataType, DataTypes, UserDefinedType}
 
 class ThetaSketchType extends UserDefinedType[ThetaSketchWrapper] {
   override def sqlType: DataType = DataTypes.BinaryType

diff --git a/.../sql/theta/types/ThetaSketchWrapper.scala → ...ches/theta/types/ThetaSketchWrapper.scala b/.../sql/theta/types/ThetaSketchWrapper.scala → ...ches/theta/types/ThetaSketchWrapper.scala
@@ -15,10 +15,11 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql.types
+package org.apache.spark.sql.datasketches.theta.types
 
 import org.apache.datasketches.theta.{UpdateSketch, CompactSketch, Union}
 import org.apache.datasketches.memory.Memory
+import org.apache.spark.sql.types.SQLUserDefinedType
 
 @SQLUserDefinedType(udt = classOf[ThetaSketchType])
 class ThetaSketchWrapper(var updateSketch: Option[UpdateSketch] = None, var compactSketch: Option[CompactSketch] = None, var union: Option[Union] = None) {

diff --git a/...pache/spark/sql/SparkSessionManager.scala → ...ql/datasketches/SparkSessionManager.scala b/...pache/spark/sql/SparkSessionManager.scala → ...ql/datasketches/SparkSessionManager.scala
@@ -15,11 +15,12 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql
+package org.apache.spark.sql.datasketches.common
 
 import org.apache.log4j.{Level, Logger}
 import org.scalatest.BeforeAndAfterAll
 import org.scalatest.funsuite.AnyFunSuite
+import org.apache.spark.sql.SparkSession
 
 /**
   * This class provides a common base for tests. It can perhaps

diff --git a/.../scala/org/apache/spark/sql/KllTest.scala → .../spark/sql/datasketches/kll/KllTest.scala b/.../scala/org/apache/spark/sql/KllTest.scala → .../spark/sql/datasketches/kll/KllTest.scala
@@ -15,16 +15,17 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql
+package org.apache.spark.sql.datasketches.kll
 
 import scala.util.Random
+import org.apache.spark.sql.Row
 import org.apache.spark.sql.functions._
 import org.apache.spark.sql.types.{StructType, StructField, IntegerType, BinaryType}
 
-import org.apache.spark.sql.functions_datasketches_kll._
 import org.apache.datasketches.kll.KllDoublesSketch
-import org.apache.spark.sql.types.KllDoublesSketchType
-import org.apache.spark.sql.registrar.KllFunctionRegistry
+import org.apache.spark.sql.datasketches.kll.functions._
+import org.apache.spark.sql.datasketches.kll.types.KllDoublesSketchType
+import org.apache.spark.sql.datasketches.common.SparkSessionManager
 
 class KllTest extends SparkSessionManager {
   import spark.implicits._

diff --git a/...cala/org/apache/spark/sql/ThetaTest.scala → ...rk/sql/datasketches/theta/ThetaTest.scala b/...cala/org/apache/spark/sql/ThetaTest.scala → ...rk/sql/datasketches/theta/ThetaTest.scala
@@ -15,10 +15,12 @@
  * limitations under the License.
  */
 
-package org.apache.spark.sql
+package org.apache.spark.sql.theta
 
-import org.apache.spark.sql.functions_datasketches_theta._
-import org.apache.spark.sql.registrar.ThetaFunctionRegistry
+import org.apache.spark.sql.Row
+import org.apache.spark.sql.datasketches.common.SparkSessionManager
+import org.apache.spark.sql.datasketches.theta.functions._
+import org.apache.spark.sql.datasketches.theta.ThetaFunctionRegistry
 
 class ThetaTest extends SparkSessionManager {
   import spark.implicits._