[FLINK-6783] Wrongly extracted TypeInformations for WindowedStream::aggregate

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 1.3.0, 1.3.1
• Fix Version/s: 1.3.1, 1.4.0
• Component/s: API / DataStream
• Labels: flink-rel-1.3.1-blockers
• Assignee: Dawid Wysakowicz
• Reporter: Dawid Wysakowicz
• Created: 31/May/17 12:37
• Updated: 01/Oct/19 15:39
• Resolved: 08/Jun/17 12:09

Description

The following test fails because of wrongly acquired output type for AggregateFunction:

@Test
public void testAggregateWithWindowFunctionDifferentResultTypes() throws Exception {
	StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();

	DataStream<Tuple2<String, Integer>> source = env.fromElements(Tuple2.of("hello", 1), Tuple2.of("hello", 2));

	DataStream<Tuple3<String, String, Integer>> window = source
		.keyBy(new TupleKeySelector())
		.window(TumblingEventTimeWindows.of(Time.of(1, TimeUnit.SECONDS)))
		.aggregate(new AggregateFunction<Tuple2<String, Integer>, Tuple2<String, Integer>, String>() {
			@Override
			public Tuple2<String, Integer> createAccumulator() {
				return Tuple2.of("", 0);
			}

			@Override
			public void add(
				Tuple2<String, Integer> value, Tuple2<String, Integer> accumulator) {

			}

			@Override
			public String getResult(Tuple2<String, Integer> accumulator) {
				return accumulator.f0;
			}

			@Override
			public Tuple2<String, Integer> merge(
				Tuple2<String, Integer> a, Tuple2<String, Integer> b) {
				return Tuple2.of("", 0);
			}
		}, new WindowFunction<String, Tuple3<String, String, Integer>, String, TimeWindow>() {
			@Override
			public void apply(
				String s,
				TimeWindow window,
				Iterable<String> input,
				Collector<Tuple3<String, String, Integer>> out) throws Exception {
				out.collect(Tuple3.of("", "", 0));
			}
		});

	OneInputTransformation<Tuple2<String, Integer>, Tuple3<String, String, Integer>> transform =
		(OneInputTransformation<Tuple2<String, Integer>, Tuple3<String, String, Integer>>) window.getTransformation();

	OneInputStreamOperator<Tuple2<String, Integer>, Tuple3<String, String, Integer>> operator = transform.getOperator();

	Assert.assertTrue(operator instanceof WindowOperator);
	WindowOperator<String, Tuple2<String, Integer>, ?, ?, ?> winOperator =
		(WindowOperator<String, Tuple2<String, Integer>, ?, ?, ?>) operator;

	Assert.assertTrue(winOperator.getTrigger() instanceof EventTimeTrigger);
	Assert.assertTrue(winOperator.getWindowAssigner() instanceof TumblingEventTimeWindows);
	Assert.assertTrue(winOperator.getStateDescriptor() instanceof AggregatingStateDescriptor);

	processElementAndEnsureOutput(
		operator, winOperator.getKeySelector(), BasicTypeInfo.STRING_TYPE_INFO, new Tuple2<>("hello", 1));
}

The test results in

org.apache.flink.api.common.functions.InvalidTypesException: Input mismatch: Tuple type expected.

	at org.apache.flink.api.java.typeutils.TypeExtractor.validateInputType(TypeExtractor.java:1157)
	at org.apache.flink.api.java.typeutils.TypeExtractor.getUnaryOperatorReturnType(TypeExtractor.java:451)
	at org.apache.flink.streaming.api.datastream.WindowedStream.aggregate(WindowedStream.java:855)
	at org.apache.flink.streaming.runtime.operators.windowing.WindowTranslationTest.testAggregateWithWindowFunctionDifferentResultTypes(WindowTranslationTest.java:702)
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.lang.reflect.Method.invoke(Method.java:498)
	at org.junit.runners.model.FrameworkMethod$1.runReflectiveCall(FrameworkMethod.java:50)
	at org.junit.internal.runners.model.ReflectiveCallable.run(ReflectiveCallable.java:12)
	at org.junit.runners.model.FrameworkMethod.invokeExplosively(FrameworkMethod.java:47)
	at org.junit.internal.runners.statements.InvokeMethod.evaluate(InvokeMethod.java:17)
	at org.junit.runners.ParentRunner.runLeaf(ParentRunner.java:325)
	at org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:78)
	at org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:57)
	at org.junit.runners.ParentRunner$3.run(ParentRunner.java:290)
	at org.junit.runners.ParentRunner$1.schedule(ParentRunner.java:71)
	at org.junit.runners.ParentRunner.runChildren(ParentRunner.java:288)
	at org.junit.runners.ParentRunner.access$000(ParentRunner.java:58)
	at org.junit.runners.ParentRunner$2.evaluate(ParentRunner.java:268)
	at org.junit.runners.ParentRunner.run(ParentRunner.java:363)
	at org.junit.runner.JUnitCore.run(JUnitCore.java:137)
	at com.intellij.junit4.JUnit4IdeaTestRunner.startRunnerWithArgs(JUnit4IdeaTestRunner.java:68)
	at com.intellij.rt.execution.junit.IdeaTestRunner$Repeater.startRunnerWithArgs(IdeaTestRunner.java:51)
	at com.intellij.rt.execution.junit.JUnitStarter.prepareStreamsAndStart(JUnitStarter.java:242)
	at com.intellij.rt.execution.junit.JUnitStarter.main(JUnitStarter.java:70)
Caused by: org.apache.flink.api.common.functions.InvalidTypesException: Tuple type expected.
	at org.apache.flink.api.java.typeutils.TypeExtractor.validateInfo(TypeExtractor.java:1204)
	at org.apache.flink.api.java.typeutils.TypeExtractor.validateInputType(TypeExtractor.java:1154)
	... 25 more

I tracked down the issue and the reason is wrongly handled outputTypeArgumentIndex in TypeExtractor::getUnaryOperatorReturnType.

My proposition is to remove/deprecate version of TypeExtractor::getUnaryOperatorReturnType that accepts hasIterable and hasCollector as parameters and move all invocations to explicitly passing index of output type (after fixing outputTypeArgumentIndex handling in line TypeExtractor:455.

Issue Links

• links to: GitHub Pull Request #4039
• links to: GitHub Pull Request #4089