Github user rtudoran commented on a diff in the pull request: https://github.com/apache/flink/pull/3715#discussion_r116396723 --- Diff: flink-libraries/flink-table/src/main/scala/org/apache/flink/table/runtime/join/ProcTimeInnerJoin.scala --- @@ -0,0 +1,323 @@ +/* + * 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.runtime.join + +import java.util +import java.util.{List => JList} + +import org.apache.flink.api.common.functions.RichFilterFunction +import org.apache.flink.api.common.state._ +import org.apache.flink.api.common.typeinfo.{BasicTypeInfo, TypeInformation} +import org.apache.flink.api.java.typeutils.ListTypeInfo +import org.apache.flink.configuration.Configuration +import org.apache.flink.streaming.api.functions.co.CoProcessFunction +import org.apache.flink.table.runtime.types.{CRow, CRowTypeInfo} +import org.apache.flink.types.Row +import org.apache.flink.util.Collector + +/** + * A CoProcessFunction to support stream join stream, currently just support inner-join + * + * @param leftStreamWindowSize the left stream window size + * @param rightStreamWindowSize the right stream window size + * @param element1Type the input type of left stream + * @param element2Type the input type of right stream + * @param filterFunc the function of other non-equi condition include time condition + * + */ +class ProcTimeInnerJoin( + private val leftStreamWindowSize: Long, + private val rightStreamWindowSize: Long, + private val element1Type: TypeInformation[CRow], + private val element2Type: TypeInformation[CRow], + private val filterFunc: RichFilterFunction[Row]) + extends CoProcessFunction[CRow, CRow, CRow] { + + private var outputC: CRow = _ + private var listToRemove: JList[Long] = _ + + /** state to hold left stream element **/ + private var row1MapState: MapState[Long, JList[Row]] = _ + /** state to hold right stream element **/ + private var row2MapState: MapState[Long, JList[Row]] = _ + + /** state to record last timer of left stream, 0 means no timer **/ + private var timerState1: ValueState[Long] = _ --- End diff -- @fhueske @hongyuhong As far as i know there is an automatic deduplication of timers. So even if you would register a timer for every event that comes, as long as it is registered to trigger on the same time - the onTimer method will be executed once. If there is no way to register the triggering on the same time - then i am wondering whether you need a state. In case there would be a crash and the events would be restored - you would anyway had to register a timer again - which would happen also if you would only have a class variable field (e.g. a boolean).
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---