[
https://issues.apache.org/jira/browse/FLINK-10583?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16704805#comment-16704805
]
ASF GitHub Bot commented on FLINK-10583:
----------------------------------------
kl0u commented on a change in pull request #6871:
[FLINK-10583][FLINK-10584][table] Add State Retention to temporal joins.
URL: https://github.com/apache/flink/pull/6871#discussion_r237877929
##########
File path:
flink-libraries/flink-table/src/test/scala/org/apache/flink/table/runtime/harness/TemporalJoinHarnessTest.scala
##########
@@ -530,6 +530,214 @@ class TemporalJoinHarnessTest extends HarnessTestBase {
0)
}
+ // ---------------------- Event time TTL tests ----------------------
+
+ @Test
+ def testEventTimeScenarioWithoutAdvancingProcessingTime(): Unit = {
+ // min=2ms max=4ms
+ val testHarness = createTestHarness(new
OrdersRatesRowtimeTemporalJoinInfo())
+
+ testHarness.open()
+ val expectedOutput = new ConcurrentLinkedQueue[Object]()
+
+ testHarness.setProcessingTime(1L)
+
+ // process without conversion rates
+ testHarness.processElement1(new StreamRecord(CRow(2L, "Euro", 1L)))
+ testHarness.processElement2(new StreamRecord(CRow("Euro", 114L, 0L)))
+
+ expectedOutput.add(new StreamRecord(CRow(2L, "Euro", 1L, "Euro", 114L,
0L)))
+
+ testHarness.processWatermark1(new Watermark(2L))
+ testHarness.processWatermark2(new Watermark(2L))
+
+ verify(expectedOutput, testHarness.getOutput)
+
+ testHarness.processElement1(new StreamRecord(CRow(2L, "Euro", 4L)))
+
+ expectedOutput.add(new StreamRecord(CRow(2L, "Euro", 4L, "Euro", 114L,
0L)))
+
+ testHarness.processWatermark1(new Watermark(5L))
+ testHarness.processWatermark2(new Watermark(5L))
+
+ verify(expectedOutput, testHarness.getOutput)
+
+ testHarness.close()
+ }
+
+ @Test
+ def testEventTimeCleanupShouldSucceed(): Unit = {
Review comment:
The tests have conceptually 3 phases, the build-up where we add some
elements and set the initial cleanup timer, the update of the timer (left or
right depending on the test), and then a final record that would otherwise fire
but due to the cleanup, it does not.
I agree that there is code duplication in the "building phase" of the test,
but even if we merge the two tests, we will still have to go through the
"building phase" for the second one. So I am not sure if we will save a lot of
code.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
> Add support for state retention to the Processing Time versioned joins.
> -----------------------------------------------------------------------
>
> Key: FLINK-10583
> URL: https://issues.apache.org/jira/browse/FLINK-10583
> Project: Flink
> Issue Type: Sub-task
> Components: Table API & SQL
> Affects Versions: 1.7.0
> Reporter: Kostas Kloudas
> Assignee: Kostas Kloudas
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.8.0
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
