[GitHub] [carbondata] Indhumathi27 commented on a change in pull request #4127: [CARBONDATA-4166] Geo spatial Query Enhancements

[GitBox <gi...@apache.org>]

Indhumathi27 commented on a change in pull request #4127:
URL: https://github.com/apache/carbondata/pull/4127#discussion_r627097420



##########
File path: integration/spark/src/main/scala/org/apache/carbondata/geo/GeoUtilUDFs.scala
##########
@@ -30,6 +32,7 @@ object GeoUtilUDFs {
     sparkSession.udf.register("LatLngToGeoId", new LatLngToGeoIdUDF)
     sparkSession.udf.register("ToUpperLayerGeoId", new ToUpperLayerGeoIdUDF)
     sparkSession.udf.register("ToRangeList", new ToRangeListUDF)
+    sparkSession.udf.register("ToRangeListAsString", new ToRangeListAsStringUDF)

Review comment:
       This udf is not exposed to user. This is for internal purpose only. 

##########
File path: integration/spark/src/main/scala/org/apache/carbondata/geo/InPolygonUDF.scala
##########
@@ -38,6 +43,46 @@ class InPolygonUDF extends (String => Boolean) with Serializable {
   }
 }
 
+@InterfaceAudience.Internal
+class InPolygonJoinRangeListUDF extends ((String, String) => Boolean) with Serializable {
+  override def apply(geoId: String, polygonRanges: String): Boolean = {
+    if (polygonRanges == null || polygonRanges.equalsIgnoreCase("null")) {
+      return false
+    }
+    // parser and get the range list
+    var range: String = polygonRanges
+    val pattern = Pattern.compile(GeoConstants.RANGELIST_REG_EXPRESSION, Pattern.CASE_INSENSITIVE)
+    val matcher = pattern.matcher(polygonRanges)
+    while ( { matcher.find }) {
+      val matchedStr = matcher.group
+      range = matchedStr
+    }
+    val ranges = PolygonRangeListExpression.getRangeListFromString(range)

Review comment:
       NULL check is already handled here in line:49. Handled some refactoring to extract common code to new method.

##########
File path: integration/spark/src/main/scala/org/apache/spark/sql/execution/joins/BroadCastPolygonFilterPushJoin.scala
##########
@@ -0,0 +1,219 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.execution.joins
+
+import scala.collection.JavaConverters._
+import scala.collection.mutable
+import scala.collection.mutable.ArrayBuffer
+
+import org.apache.spark.rdd.RDD
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.catalyst.expressions.{BindReferences, Expression, JoinedRow, Literal, ScalaUDF}
+import org.apache.spark.sql.catalyst.plans.JoinType
+import org.apache.spark.sql.catalyst.plans.physical.BroadcastMode
+import org.apache.spark.sql.execution.{BinaryExecNode, ProjectExec, RowDataSourceScanExec, SparkPlan}
+import org.apache.spark.sql.execution.exchange.{BroadcastExchangeExec, ReusedExchangeExec}
+import org.apache.spark.sql.execution.joins.BroadCastPolygonFilterPushJoin.addPolygonRangeListFilterToPlan
+import org.apache.spark.sql.execution.metric.SQLMetrics
+import org.apache.spark.sql.execution.strategy.CarbonDataSourceScan
+import org.apache.spark.sql.secondaryindex.joins.BroadCastSIFilterPushJoin
+import org.apache.spark.sql.types.TimestampType
+import org.apache.spark.unsafe.types.UTF8String
+
+import org.apache.carbondata.core.constants.CarbonCommonConstants
+import org.apache.carbondata.core.index.{IndexFilter, Segment}
+import org.apache.carbondata.core.metadata.schema.table.CarbonTable
+import org.apache.carbondata.core.util.CarbonProperties
+import org.apache.carbondata.geo.{GeoConstants, GeoUtils}
+import org.apache.carbondata.geo.scan.expression.PolygonRangeListExpression
+import org.apache.carbondata.spark.rdd.CarbonScanRDD
+
+case class BroadCastPolygonFilterPushJoin(

Review comment:
        BroadCastPolygonFilterPushJoin implementation is not same as BroadCastSIFilterPushJoin in terms of filter and method definition. so, better to keep it seperate.

##########
File path: integration/spark/src/main/scala/org/apache/spark/sql/execution/joins/BroadCastPolygonFilterPushJoin.scala
##########
@@ -0,0 +1,219 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.execution.joins
+
+import scala.collection.JavaConverters._
+import scala.collection.mutable
+import scala.collection.mutable.ArrayBuffer
+
+import org.apache.spark.rdd.RDD
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.catalyst.expressions.{BindReferences, Expression, JoinedRow, Literal, ScalaUDF}
+import org.apache.spark.sql.catalyst.plans.JoinType
+import org.apache.spark.sql.catalyst.plans.physical.BroadcastMode
+import org.apache.spark.sql.execution.{BinaryExecNode, ProjectExec, RowDataSourceScanExec, SparkPlan}
+import org.apache.spark.sql.execution.exchange.{BroadcastExchangeExec, ReusedExchangeExec}
+import org.apache.spark.sql.execution.joins.BroadCastPolygonFilterPushJoin.addPolygonRangeListFilterToPlan
+import org.apache.spark.sql.execution.metric.SQLMetrics
+import org.apache.spark.sql.execution.strategy.CarbonDataSourceScan
+import org.apache.spark.sql.secondaryindex.joins.BroadCastSIFilterPushJoin
+import org.apache.spark.sql.types.TimestampType
+import org.apache.spark.unsafe.types.UTF8String
+
+import org.apache.carbondata.core.constants.CarbonCommonConstants
+import org.apache.carbondata.core.index.{IndexFilter, Segment}
+import org.apache.carbondata.core.metadata.schema.table.CarbonTable
+import org.apache.carbondata.core.util.CarbonProperties
+import org.apache.carbondata.geo.{GeoConstants, GeoUtils}
+import org.apache.carbondata.geo.scan.expression.PolygonRangeListExpression
+import org.apache.carbondata.spark.rdd.CarbonScanRDD
+
+case class BroadCastPolygonFilterPushJoin(
+    leftKeys: Seq[Expression],
+    rightKeys: Seq[Expression],
+    joinType: JoinType,
+    buildSide: BuildSide,
+    condition: Option[Expression],
+    left: SparkPlan,
+    right: SparkPlan
+) extends BinaryExecNode with HashJoin {
+
+  override protected lazy val (buildPlan, streamedPlan) = buildSide match {
+    case BuildLeft => (left, right)
+    case BuildRight => (right, left)
+  }
+
+  override lazy val metrics = Map(
+    "numOutputRows" -> SQLMetrics.createMetric(sparkContext, "number of output rows"))
+
+  private lazy val inputCopy: Array[InternalRow] = {
+    getBuildPlan.map(_.copy()).collect().clone()
+  }
+
+  lazy val spatialTableRDD: Option[RDD[InternalRow]] = streamedPlan.collectFirst {
+    case scan: CarbonDataSourceScan => scan.inputRDDs().head
+  }
+
+  @transient private lazy val boundCondition: InternalRow => Boolean = {
+    if (condition.isDefined) {
+      newPredicate(condition.get, streamedPlan.output ++ buildPlan.output).eval _
+    } else {
+      (_: InternalRow) => true
+    }
+  }
+
+  override protected def doExecute(): RDD[InternalRow] = {
+    // get polygon rangeList from polygon table and add as IN_POLYGON_RANGE_LIST filter to the
+    // spatial table
+    addPolygonRangeListFilterToPlan(buildPlan, streamedPlan, inputCopy, condition)
+    // inner join spatial and polygon plan by applying in_polygon join filter
+    streamedPlan.execute().mapPartitionsInternal {
+      streamedIter =>
+        // get polygon table data rows
+        val buildRows = inputCopy
+        val joinedRow = new JoinedRow
+
+        streamedIter.flatMap { streamedRow =>
+          val joinedRows = buildRows.iterator.map(r => joinedRow(streamedRow, r))
+          // apply in_polygon_join filter
+          if (condition.isDefined) {
+            joinedRows.filter(boundCondition)
+          } else {
+            joinedRows
+          }
+        }
+    }
+  }
+
+  protected def getBuildPlan: RDD[InternalRow] = {
+    buildPlan match {
+      case c@CarbonBroadCastExchangeExec(_, _) =>
+        c.asInstanceOf[BroadcastExchangeExec].child.execute()
+      case ReusedExchangeExec(_, c@CarbonBroadCastExchangeExec(_, _)) =>
+        c.asInstanceOf[BroadcastExchangeExec].child.execute()
+      case _ => buildPlan.children.head match {
+        case c@CarbonBroadCastExchangeExec(_, _) =>
+          c.asInstanceOf[BroadcastExchangeExec].child.execute()
+        case ReusedExchangeExec(_, c@CarbonBroadCastExchangeExec(_, _)) =>
+          c.asInstanceOf[BroadcastExchangeExec].child.execute()
+        case _ => buildPlan.execute()
+      }
+    }
+  }
+}
+
+object BroadCastPolygonFilterPushJoin {

Review comment:
       done




-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org