Github user fhueske commented on a diff in the pull request:
https://github.com/apache/flink/pull/3386#discussion_r106625858
--- Diff:
flink-libraries/flink-table/src/test/scala/org/apache/flink/table/api/scala/stream/sql/SqlITCase.scala
---
@@ -317,4 +320,119 @@ class SqlITCase extends StreamingWithStateTestBase {
result.addSink(new StreamITCase.StringSink)
env.execute()
}
+
+ /** test sliding event-time unbounded window with partition by **/
+ @Test
+ def testUnboundedEventTimeRowWindowWithPartition(): Unit = {
+ val env = StreamExecutionEnvironment.getExecutionEnvironment
+ val tEnv = TableEnvironment.getTableEnvironment(env)
+ env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime)
+ env.setStateBackend(getStateBackend)
+ StreamITCase.testResults = mutable.MutableList()
+ env.setParallelism(1)
+
+ val sqlQuery = "SELECT a, b, c, " +
+ "SUM(a) over (" +
+ "partition by a order by rowtime() range between unbounded preceding
and current row), " +
+ "count(a) over (" +
+ "partition by a order by rowtime() range between unbounded preceding
and current row), " +
+ "avg(a) over (" +
--- End diff --
Also, most groups have just a single record. The max is two records. With
that we cannot really check if the sorting works correctly. Can you make less
groups (less distinct `a` values) and add more rows for some groups with
out-of-order timestamps?
---
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 [email protected] or file a JIRA ticket
with INFRA.
---
