[
https://issues.apache.org/jira/browse/FLINK-5804?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15900982#comment-15900982
]
ASF GitHub Bot commented on FLINK-5804:
---------------------------------------
Github user sunjincheng121 commented on a diff in the pull request:
https://github.com/apache/flink/pull/3491#discussion_r104879219
--- Diff:
flink-libraries/flink-table/src/main/scala/org/apache/flink/table/runtime/aggregate/UnboundedNonPartitionedProcessingOverProcessFunction.scala
---
@@ -0,0 +1,107 @@
+/*
+ * 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.aggregate
+
+import org.apache.flink.api.common.state.{ListState, ListStateDescriptor}
+import org.apache.flink.api.java.typeutils.RowTypeInfo
+import org.apache.flink.configuration.Configuration
+import org.apache.flink.runtime.state.{FunctionInitializationContext,
FunctionSnapshotContext}
+import org.apache.flink.streaming.api.checkpoint.CheckpointedFunction
+import org.apache.flink.streaming.api.functions.ProcessFunction
+import org.apache.flink.table.functions.{Accumulator, AggregateFunction}
+import org.apache.flink.types.Row
+import org.apache.flink.util.{Collector, Preconditions}
+
+/**
+ * Process Function used for the aggregate in
+ * [[org.apache.flink.streaming.api.datastream.DataStream]]
+ *
+ * @param aggregates the list of all
[[org.apache.flink.table.functions.AggregateFunction]]
+ * used for this aggregation
+ * @param aggFields the position (in the input Row) of the input value
for each aggregate
+ */
+class UnboundedNonPartitionedProcessingOverProcessFunction(
+ private val aggregates: Array[AggregateFunction[_]],
+ private val aggFields: Array[Int],
+ private val forwardedFieldCount: Int,
+ private val aggregationStateType: RowTypeInfo)
+ extends ProcessFunction[Row, Row] with CheckpointedFunction{
+
+ Preconditions.checkNotNull(aggregates)
+ Preconditions.checkNotNull(aggFields)
+ Preconditions.checkArgument(aggregates.length == aggFields.length)
+
+ private var accumulators: Row = _
+ private var output: Row = _
+ private var state: ListState[Row] = null
+
+ override def open(config: Configuration) {
+ output = new Row(forwardedFieldCount + aggregates.length)
+ }
+
+ override def processElement(
+ input: Row,
+ ctx: ProcessFunction[Row, Row]#Context,
+ out: Collector[Row]): Unit = {
+
+ if (null == accumulators) {
+ val it = state.get().iterator()
+ if (it.hasNext) {
+ accumulators = it.next()
+ } else {
+ accumulators = new Row(aggregates.length)
+ var i = 0
+ while (i < aggregates.length) {
+ accumulators.setField(i, aggregates(i).createAccumulator())
+ i += 1
+ }
+ }
+ }
+
+ var i = 0
+ while (i < forwardedFieldCount) {
+ output.setField(i, input.getField(i))
+ i += 1
+ }
+
+ i = 0
+ while (i < aggregates.length) {
+ val index = forwardedFieldCount + i
+ val accumulator = accumulators.getField(i).asInstanceOf[Accumulator]
+ aggregates(i).accumulate(accumulator, input.getField(aggFields(i)))
+ output.setField(index, aggregates(i).getValue(accumulator))
+ i += 1
+ }
+
+ out.collect(output)
+ }
+
+ override def snapshotState(context: FunctionSnapshotContext): Unit = {
+ state.clear()
+ if (null != accumulators) {
+ state.add(accumulators)
+ }
+ }
+
+ override def initializeState(context: FunctionInitializationContext):
Unit = {
+ val stateSerializer =
+
aggregationStateType.createSerializer(getRuntimeContext.getExecutionConfig)
+ val accumulatorsDescriptor = new ListStateDescriptor[Row]("overState",
stateSerializer)
--- End diff --
When you review FLINK-5803 you had told me that use `TypeInformation`
instead of `TypeSerializer`，I had tried this, but unfortunately reported
exception. So, I use TypeSerializer to create `ListStateDescriptor`. the
exception info：
```
Caused by: java.lang.IllegalStateException: Serializer not yet initialized.
at
org.apache.flink.api.common.state.StateDescriptor.getSerializer(StateDescriptor.java:169)
at
org.apache.flink.api.common.state.ListStateDescriptor.getElementSerializer(ListStateDescriptor.java:93)
at
org.apache.flink.runtime.state.DefaultOperatorStateBackend.getOperatorState(DefaultOperatorStateBackend.java:110)
at
org.apache.flink.runtime.state.DefaultOperatorStateBackend.getOperatorState(DefaultOperatorStateBackend.java:91)
at
org.apache.flink.table.runtime.aggregate.UnboundedNonPartitionedProcessingOverProcessFunction.initializeState(UnboundedNonPartitionedProcessingOverProcessFunction.scala:104)
at
org.apache.flink.streaming.api.functions.util.StreamingFunctionUtils.tryRestoreFunction(StreamingFunctionUtils.java:178)
```
Do you want me fix this, and use `aggregationStateType`, If so, I glad to
try.
> Add [non-partitioned] processing time OVER RANGE BETWEEN UNBOUNDED PRECEDING
> aggregation to SQL
> -----------------------------------------------------------------------------------------------
>
> Key: FLINK-5804
> URL: https://issues.apache.org/jira/browse/FLINK-5804
> Project: Flink
> Issue Type: Sub-task
> Components: Table API & SQL
> Reporter: sunjincheng
> Assignee: sunjincheng
>
> The goal of this issue is to add support for OVER RANGE aggregations on
> processing time streams to the SQL interface.
> Queries similar to the following should be supported:
> {code}
> SELECT
> a,
> SUM(b) OVER (ORDER BY procTime() RANGE BETWEEN UNBOUNDED PRECEDING AND
> CURRENT ROW) AS sumB,
> MIN(b) OVER (ORDER BY procTime() RANGE BETWEEN UNBOUNDED PRECEDING AND
> CURRENT ROW) AS minB
> FROM myStream
> {code}
> The following restrictions should initially apply:
> - All OVER clauses in the same SELECT clause must be exactly the same.
> - Since no PARTITION BY clause is specified, the execution will be single
> threaded.
> - The ORDER BY clause may only have procTime() as parameter. procTime() is a
> parameterless scalar function that just indicates processing time mode.
> - bounded PRECEDING is not supported (see FLINK-5654)
> - FOLLOWING is not supported.
> The restrictions will be resolved in follow up issues. If we find that some
> of the restrictions are trivial to address, we can add the functionality in
> this issue as well.
> This issue includes:
> - Design of the DataStream operator to compute OVER ROW aggregates
> - Translation from Calcite's RelNode representation (LogicalProject with
> RexOver expression).
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
