Github user fhueske commented on a diff in the pull request:
https://github.com/apache/flink/pull/3715#discussion_r116993635
--- Diff:
flink-libraries/flink-table/src/main/scala/org/apache/flink/table/plan/rules/datastream/DataStreamJoinRule.scala
---
@@ -0,0 +1,69 @@
+/*
+ * 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.datastream
+
+import org.apache.calcite.plan.{RelOptRule, RelOptRuleCall, RelTraitSet}
+import org.apache.calcite.rel.RelNode
+import org.apache.calcite.rel.convert.ConverterRule
+import org.apache.flink.table.plan.nodes.datastream.DataStreamJoin
+import org.apache.flink.table.plan.nodes.FlinkConventions
+import org.apache.flink.table.plan.nodes.logical.FlinkLogicalJoin
+import org.apache.flink.table.plan.schema.RowSchema
+
+class DataStreamJoinRule
+ extends ConverterRule(
+ classOf[FlinkLogicalJoin],
+ FlinkConventions.LOGICAL,
+ FlinkConventions.DATASTREAM,
+ "DataStreamJoinRule") {
+
+ override def matches(call: RelOptRuleCall): Boolean = {
+ val join: FlinkLogicalJoin = call.rel(0).asInstanceOf[FlinkLogicalJoin]
+
+ val joinInfo = join.analyzeCondition
--- End diff --
I think it makes sense to separate different join cases on the level of
plan operators. That would mean we would check here if the join has bounded
time predicates (for example `left.rowtime BETWEEN right.rowtime - INTERVAL '1'
MINUTE AND right.rowtime + INTERVAL '1' MINUTE`).
If this is the case, we would extract the relevant time predicates and
create a stream-stream join RelNode. This would move a bit of code out of the
`DataStreamJoin` into the rule.
IMO, the benefit is that it will be easier to add other joins because we
only need to add a new rule, a new RelNode and runtime code. Hence, we would
not need to touch the existing join strategy.
What do you think about this @hongyuhong?
---
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.
---
