[GitHub] [flink] beyond1920 commented on a change in pull request #8203: [Flink-12208][table-planner-blink] Support translation from StreamExecSort / TemporalSort / SortLimit/ Limit to StreamTransformation.

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

beyond1920 commented on a change in pull request #8203: [Flink-12208][table-planner-blink] Support translation from StreamExecSort / TemporalSort / SortLimit/ Limit to StreamTransformation.
URL: https://github.com/apache/flink/pull/8203#discussion_r276629617
 
 

 ##########
 File path: flink-table/flink-table-planner-blink/src/main/scala/org/apache/flink/table/plan/nodes/physical/stream/StreamExecLimit.scala
 ##########
 @@ -72,4 +89,78 @@ class StreamExecLimit(
       .item("fetch", RelExplainUtil.fetchToString(fetch))
   }
 
+  //~ ExecNode methods -----------------------------------------------------------
+
+  override def getInputNodes: util.List[ExecNode[StreamTableEnvironment, _]] = {
+    List(getInput.asInstanceOf[ExecNode[StreamTableEnvironment, _]])
+  }
+
+  override protected def translateToPlanInternal(
+      tableEnv: StreamTableEnvironment): StreamTransformation[BaseRow] = {
+    if (fetch == null) {
+      throw new TableException(
+        "FETCH is missed, which on streaming table is not supported currently")
+    }
+    val inputRowTypeInfo = FlinkTypeFactory.toInternalRowType(getInput.getRowType).toTypeInfo
+    val generateRetraction = StreamExecRetractionRules.isAccRetract(this)
+    val tableConfig = tableEnv.getConfig
+    val minIdleStateRetentionTime = tableConfig.getMinIdleStateRetentionTime
+    val maxIdleStateRetentionTime = tableConfig.getMaxIdleStateRetentionTime
+
+    // rankStart begin with 1
+    val rankRange = new ConstantRankRange(limitStart + 1, limitEnd)
+    val rankType = RankType.ROW_NUMBER
+    val outputRankNumber = false
+    // TODO Use RankFunction underlying StreamExecLimit currently
 
 Review comment:
   Because SortLimit and Limit takes `limited` elements from `offset`, which is very similar to TopN, and could use AppendRankFunction or UpdateRankFunction under some as optimization in some scenarios. BTW, maybe RankFunction is not a good name, it's only for TopN.

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