[GitHub] [flink] KurtYoung commented on a change in pull request #8056: [FLINK-12013][table-planner-blink] Support calc and correlate in blink

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

KurtYoung commented on a change in pull request #8056: [FLINK-12013][table-planner-blink] Support calc and correlate in blink
URL: https://github.com/apache/flink/pull/8056#discussion_r269857820
 
 

 ##########
 File path: flink-table/flink-table-planner-blink/src/main/scala/org/apache/flink/table/plan/util/CorrelateUtil.scala
 ##########
 @@ -0,0 +1,173 @@
+/*
+ * 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.flink.table.plan.util
+
+import org.apache.flink.table.calcite.{FlinkTypeFactory, FlinkTypeSystem}
+import org.apache.flink.table.functions.utils.TableSqlFunction
+import org.apache.flink.table.plan.nodes.logical.FlinkLogicalTableFunctionScan
+
+import org.apache.calcite.rel.`type`.{RelDataType, RelDataTypeField, RelDataTypeFieldImpl}
+import org.apache.calcite.rex.{RexBuilder, RexCall, RexInputRef, RexNode, RexProgram, RexProgramBuilder}
+
+import scala.collection.JavaConversions._
+import scala.collection.JavaConverters._
+import scala.collection.mutable.ListBuffer
+
+/**
+  * An utility class for optimizing and generating ExecCorrelate operators.
+  */
+object CorrelateUtil {
+
+  def projectable(downsideCalc: RexProgram, correlateProgram: Option[RexProgram]): Boolean = {
+    val refs = downsideCalc.getReferenceCounts
+    // if the correlate output any unused column to the calc
+    val calcInputFieldCnt = downsideCalc.getInputRowType.getFieldCount
+    val projectable = refs.zipWithIndex.exists {
+      case (refCnt: Int, index: Int) if index < calcInputFieldCnt =>
+        refCnt == 0
+      case _ => false
+    }
+    // If correlate.projectProgram.nonEmpty, that means this rule has matched before,
+    // We should merge the top calc with the correlate program if we want to crop
+    // the top unused projections and push used one into the correlate.
+    // This is not supported yet, so we do a short-cut and return early.
+    // TODO: add case for pattern that we need a RexProgram merge.
+    projectable && correlateProgram.isEmpty
+  }
+
+  def getProjectableFieldSet(
+      refs: Seq[Int],
+      calcProgram: RexProgram,
+      leftInputFieldCnt: Int): Set[Int] = {
+    // calculate left/right projectable field(s)' index, only crop input refs.
+    calcProgram.getExprList.zipWithIndex.filter { case (expr, index: Int) =>
+      refs(index) == 0 && expr.isInstanceOf[RexInputRef]
+    }.map(_._1.asInstanceOf[RexInputRef].getIndex).toSet
+  }
+
+  def projectCorrelateOutputType(
+      originalType: RelDataType,
+      projectableFieldSet: Set[Int]): (RelDataType, ListBuffer[Int]) = {
+    val selects =  new ListBuffer[Int]
+    // generate new output type that removed unused column(s) for Correlate
+    val typeFactory = new FlinkTypeFactory(new FlinkTypeSystem)
+    val typeBuilder = typeFactory.builder
+    val reserveFieldTypes = originalType.getFieldList.zipWithIndex.filter {
+      // filter unused fields
+      f => !projectableFieldSet.contains(f._2)
+    }.zipWithIndex.map {
+      // create relField using new indexes
+      case ((f: RelDataTypeField, srcIdx: Int), newIdx: Int) =>
+        selects += srcIdx
+        new RelDataTypeFieldImpl(
+          f.getName,
+          newIdx, // shift to new index
+          f.getType)
+    }
+    if (reserveFieldTypes.size == 0) {
+      // downside operator only cares records number, so we must output at least one column.
+      // typical case: 'select count(*)' be pushed down here (count(0), count(1) ... as well)
+      // we choose the last column to output(columns from left input more likely to be bigger).
+      val reservedFieldIdx = originalType.getFieldCount - 1
+      selects += reservedFieldIdx
+      typeBuilder.add(
+        originalType.getFieldNames.get(reservedFieldIdx),
+        originalType.getFieldList.get(reservedFieldIdx).getType)
+    } else {
+      typeBuilder.addAll(reserveFieldTypes)
+    }
+    (typeBuilder.build(), selects)
+  }
+
+  def createProjectProgram(
+      originalType: RelDataType,
+      rexBuilder: RexBuilder,
+      selects: Seq[Int]): RexProgram = {
+    val rexProgBuilder = new RexProgramBuilder(originalType, rexBuilder)
+    val fieldNames = originalType.getFieldNames
+    selects.zipWithIndex.map {
+      case (ordinal, newIdx) =>
+        rexProgBuilder.addProject(newIdx, ordinal, fieldNames.get(ordinal))
+    }
+    rexProgBuilder.getProgram
+  }
+
+  def shiftProjectsAndCondition(
+      refs: Seq[Int],
+      calcProgram: RexProgram,
+      projectableFieldSet: Set[Int],
+      newInputType: RelDataType): (List[RexNode], RexNode) = {
+    val reservedFieldsMapping = calcProgram.getExprList.zipWithIndex.filter {
+      case (_, index) => refs(index) > 0
+    }.zipWithIndex.map {
+      case (((expr, srcIndex), newIndex)) => srcIndex -> newIndex
+    }.toMap
+
+    val shiftProjects = calcProgram.getProjectList.map {
+      ref => (calcProgram.expandLocalRef(ref), ref.getIndex)
+    }.filter {
+      case (_, idx) => !projectableFieldSet.contains(idx)
+    }.map {
+      case (rex, _) =>
+        FlinkRexUtil.adjustInputRefs(rex, reservedFieldsMapping, newInputType)
+    }.toList
+
+    val shiftCondition = if (null != calcProgram.getCondition) {
+      FlinkRexUtil.adjustInputRefs(
+        calcProgram.expandLocalRef(calcProgram.getCondition),
+        reservedFieldsMapping,
+        newInputType)
+    } else {
+      null
+    }
+    (shiftProjects, shiftCondition)
+  }
+
+  def selectToString(rowType: RelDataType): String = rowType.getFieldNames.asScala.mkString(",")
+
+  def correlateOpName(
 
 Review comment:
   Merge name related utilities to RelExplainUtil
   
   

----------------------------------------------------------------
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


With regards,
Apache Git Services