KurtYoung commented on a change in pull request #8051: [FLINK-12018] [table-planner-blink] Add support for generating optimized logical plan for Sort and Rank URL: https://github.com/apache/flink/pull/8051#discussion_r269843899
########## File path: flink-table/flink-table-planner-blink/src/main/scala/org/apache/flink/table/plan/rules/physical/stream/StreamExecFirstLastRowRule.scala ########## @@ -0,0 +1,180 @@ +/* + * 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.rules.physical.stream + +import org.apache.flink.table.calcite.FlinkTypeFactory +import org.apache.flink.table.plan.`trait`.FlinkRelDistribution +import org.apache.flink.table.plan.nodes.FlinkConventions +import org.apache.flink.table.plan.nodes.calcite.ConstantRankRange +import org.apache.flink.table.plan.nodes.logical.{FlinkLogicalRank, FlinkLogicalSort} +import org.apache.flink.table.plan.nodes.physical.stream.StreamExecFirstLastRow +import org.apache.flink.table.plan.util.FlinkRelOptUtil + +import org.apache.calcite.plan.{RelOptRule, RelOptRuleCall} +import org.apache.calcite.rel.`type`.RelDataType +import org.apache.calcite.rel.convert.ConverterRule +import org.apache.calcite.rel.core.Sort +import org.apache.calcite.rel.{RelCollation, RelNode} +import org.apache.calcite.sql.SqlKind + +/** + * Rule that matches [[FlinkLogicalSort]] which is sorted by proc-time attribute and + * fetches only one record started from 0, and converts it to [[StreamExecFirstLastRow]]. + */ +class StreamExecFirstLastRowFromSortRule + extends ConverterRule( + classOf[FlinkLogicalSort], + FlinkConventions.LOGICAL, + FlinkConventions.STREAM_PHYSICAL, + "StreamExecFirstLastRowFromSortRule") { + + override def matches(call: RelOptRuleCall): Boolean = { + val sort: FlinkLogicalSort = call.rel(0) + StreamExecFirstLastRowRule.canConvertToFirstLastRow(sort) + } + + override def convert(rel: RelNode): RelNode = { + val sort = rel.asInstanceOf[FlinkLogicalSort] + val input = sort.getInput + val requiredDistribution = FlinkRelDistribution.SINGLETON + val requiredTraitSet = sort.getInput.getTraitSet + .replace(FlinkConventions.STREAM_PHYSICAL) + .replace(requiredDistribution) + + val convInput: RelNode = RelOptRule.convert(input, requiredTraitSet) + + val fieldCollation = sort.collation.getFieldCollations.get(0) + val fieldType = input.getRowType.getFieldList.get(fieldCollation.getFieldIndex).getType + val isRowtime = FlinkTypeFactory.isRowtimeIndicatorType(fieldType) + val providedTraitSet = rel.getTraitSet.replace(FlinkConventions.STREAM_PHYSICAL) + // order by timeIndicator desc ==> lastRow, otherwise is firstRow + val isLastRow = fieldCollation.direction.isDescending + new StreamExecFirstLastRow( + rel.getCluster, + providedTraitSet, + convInput, + Array(), + isRowtime, + isLastRow) + } +} + +/** + * Rule that matches [[FlinkLogicalRank]] which is sorted by proc-time attribute and + * limits 1 and its RankFunction is ROW_NUMBER, and converts it to [[StreamExecFirstLastRow]]. + */ +class StreamExecFirstLastRowFromRankRule Review comment: Could you give some examples to show what kind of queries will be captured? And why we are translating these kinds of query to the form you wanted ---------------------------------------------------------------- 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: [email protected] With regards, Apache Git Services
