Github user rmetzger commented on a diff in the pull request:
https://github.com/apache/incubator-flink/pull/141#discussion_r18435958
--- Diff:
flink-compiler/src/test/java/org/apache/flink/compiler/WorksetIterationsJavaApiCompilerTest.java
---
@@ -78,7 +78,7 @@ public void
testJavaApiWithDeferredSoltionSetUpdateWithMapper() {
// verify joinWithSolutionSet
assertEquals(ShipStrategyType.PARTITION_HASH,
joinWithSolutionSetNode.getInput1().getShipStrategy());
assertEquals(ShipStrategyType.FORWARD,
joinWithSolutionSetNode.getInput2().getShipStrategy());
- assertEquals(new FieldList(0, 1),
joinWithSolutionSetNode.getKeysForInput1());
--- End diff --
The sorting of the Tuple key fields has happened in the "FieldPositionKeys"
class, before the optimizer.
This test is taking its inputs from a program written with the Java API.
Since I've removed the sorting of Tuple keys, the input to the optimizer has
changed with this test case.
I think thats the reason why the key-field order has changed for this test.
The key definition in the input plan is `where(1, 0)`.
---
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.
---
