use of org.apache.flink.streaming.runtime.streamrecord.StreamElementSerializer in project flink by apache.
the class AllWindowedStream method reduce.
/**
* Applies the given window function to each window. The window function is called for each
* evaluation of the window for each key individually. The output of the window function is
* interpreted as a regular non-windowed stream.
*
* <p>Arriving data is incrementally aggregated using the given reducer.
*
* @param reduceFunction The reduce function that is used for incremental aggregation.
* @param function The window function.
* @param resultType Type information for the result type of the window function
* @return The data stream that is the result of applying the window function to the window.
*/
@PublicEvolving
public <R> SingleOutputStreamOperator<R> reduce(ReduceFunction<T> reduceFunction, AllWindowFunction<T, R, W> function, TypeInformation<R> resultType) {
if (reduceFunction instanceof RichFunction) {
throw new UnsupportedOperationException("ReduceFunction of reduce can not be a RichFunction.");
}
// clean the closures
function = input.getExecutionEnvironment().clean(function);
reduceFunction = input.getExecutionEnvironment().clean(reduceFunction);
String callLocation = Utils.getCallLocationName();
String udfName = "AllWindowedStream." + callLocation;
String opName = windowAssigner.getClass().getSimpleName();
String opDescription;
KeySelector<T, Byte> keySel = input.getKeySelector();
OneInputStreamOperator<T, R> operator;
if (evictor != null) {
@SuppressWarnings({ "unchecked", "rawtypes" }) TypeSerializer<StreamRecord<T>> streamRecordSerializer = (TypeSerializer<StreamRecord<T>>) new StreamElementSerializer(input.getType().createSerializer(getExecutionEnvironment().getConfig()));
ListStateDescriptor<StreamRecord<T>> stateDesc = new ListStateDescriptor<>("window-contents", streamRecordSerializer);
opDescription = "TriggerWindow(" + windowAssigner + ", " + stateDesc + ", " + trigger + ", " + evictor + ", " + udfName + ")";
operator = new EvictingWindowOperator<>(windowAssigner, windowAssigner.getWindowSerializer(getExecutionEnvironment().getConfig()), keySel, input.getKeyType().createSerializer(getExecutionEnvironment().getConfig()), stateDesc, new InternalIterableAllWindowFunction<>(new ReduceApplyAllWindowFunction<>(reduceFunction, function)), trigger, evictor, allowedLateness, lateDataOutputTag);
} else {
ReducingStateDescriptor<T> stateDesc = new ReducingStateDescriptor<>("window-contents", reduceFunction, input.getType().createSerializer(getExecutionEnvironment().getConfig()));
opDescription = "TriggerWindow(" + windowAssigner + ", " + stateDesc + ", " + trigger + ", " + udfName + ")";
operator = new WindowOperator<>(windowAssigner, windowAssigner.getWindowSerializer(getExecutionEnvironment().getConfig()), keySel, input.getKeyType().createSerializer(getExecutionEnvironment().getConfig()), stateDesc, new InternalSingleValueAllWindowFunction<>(function), trigger, allowedLateness, lateDataOutputTag);
}
return input.transform(opName, resultType, operator).setDescription(opDescription).forceNonParallel();
}
use of org.apache.flink.streaming.runtime.streamrecord.StreamElementSerializer in project flink by apache.
the class EvictingWindowOperatorTest method testTimeEvictorEvictBefore.
/**
* Tests TimeEvictor evictBefore behavior.
*/
@Test
public void testTimeEvictorEvictBefore() throws Exception {
AtomicInteger closeCalled = new AtomicInteger(0);
final int triggerCount = 2;
final int windowSize = 4;
@SuppressWarnings({ "unchecked", "rawtypes" }) TypeSerializer<StreamRecord<Tuple2<String, Integer>>> streamRecordSerializer = (TypeSerializer<StreamRecord<Tuple2<String, Integer>>>) new StreamElementSerializer(STRING_INT_TUPLE.createSerializer(new ExecutionConfig()));
ListStateDescriptor<StreamRecord<Tuple2<String, Integer>>> stateDesc = new ListStateDescriptor<>("window-contents", streamRecordSerializer);
EvictingWindowOperator<String, Tuple2<String, Integer>, Tuple2<String, Integer>, TimeWindow> operator = new EvictingWindowOperator<>(TumblingEventTimeWindows.of(Time.of(windowSize, TimeUnit.SECONDS)), new TimeWindow.Serializer(), new TupleKeySelector(), BasicTypeInfo.STRING_TYPE_INFO.createSerializer(new ExecutionConfig()), stateDesc, new InternalIterableWindowFunction<>(new RichSumReducer<TimeWindow>(closeCalled)), CountTrigger.of(triggerCount), TimeEvictor.of(Time.seconds(2)), 0, null);
OneInputStreamOperatorTestHarness<Tuple2<String, Integer>, Tuple2<String, Integer>> testHarness = new KeyedOneInputStreamOperatorTestHarness<>(operator, new TupleKeySelector(), BasicTypeInfo.STRING_TYPE_INFO);
long initialTime = 0L;
ConcurrentLinkedQueue<Object> expectedOutput = new ConcurrentLinkedQueue<>();
testHarness.open();
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 1000));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 3999));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime + 20));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime + 999));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime + 5999));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 3500));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 2001));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 1001));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key2", 1), 3999));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key1", 2), 3999));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key2", 3), 3999));
TestHarnessUtil.assertOutputEqualsSorted("Output was not correct.", expectedOutput, testHarness.getOutput(), new ResultSortComparator());
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime + 6500));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 1002));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key1", 2), 7999));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key2", 3), 3999));
TestHarnessUtil.assertOutputEqualsSorted("Output was not correct.", expectedOutput, testHarness.getOutput(), new ResultSortComparator());
testHarness.close();
Assert.assertEquals("Close was not called.", 1, closeCalled.get());
}
use of org.apache.flink.streaming.runtime.streamrecord.StreamElementSerializer in project flink by apache.
the class EvictingWindowOperatorTest method testTimeEvictorEvictAfter.
/**
* Tests TimeEvictor evictAfter behavior.
*/
@Test
public void testTimeEvictorEvictAfter() throws Exception {
AtomicInteger closeCalled = new AtomicInteger(0);
final int triggerCount = 2;
final boolean evictAfter = true;
@SuppressWarnings({ "unchecked", "rawtypes" }) TypeSerializer<StreamRecord<Tuple2<String, Integer>>> streamRecordSerializer = (TypeSerializer<StreamRecord<Tuple2<String, Integer>>>) new StreamElementSerializer(STRING_INT_TUPLE.createSerializer(new ExecutionConfig()));
ListStateDescriptor<StreamRecord<Tuple2<String, Integer>>> stateDesc = new ListStateDescriptor<>("window-contents", streamRecordSerializer);
EvictingWindowOperator<String, Tuple2<String, Integer>, Tuple2<String, Integer>, GlobalWindow> operator = new EvictingWindowOperator<>(GlobalWindows.create(), new GlobalWindow.Serializer(), new TupleKeySelector(), BasicTypeInfo.STRING_TYPE_INFO.createSerializer(new ExecutionConfig()), stateDesc, new InternalIterableWindowFunction<>(new RichSumReducer<GlobalWindow>(closeCalled)), CountTrigger.of(triggerCount), TimeEvictor.of(Time.seconds(2), evictAfter), 0, null);
OneInputStreamOperatorTestHarness<Tuple2<String, Integer>, Tuple2<String, Integer>> testHarness = new KeyedOneInputStreamOperatorTestHarness<>(operator, new TupleKeySelector(), BasicTypeInfo.STRING_TYPE_INFO);
long initialTime = 0L;
ConcurrentLinkedQueue<Object> expectedOutput = new ConcurrentLinkedQueue<>();
testHarness.open();
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 1000));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 4000));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime + 20));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime + 999));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 3500));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 2001));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 1001));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key2", 2), Long.MAX_VALUE));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key1", 2), Long.MAX_VALUE));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key2", 3), Long.MAX_VALUE));
TestHarnessUtil.assertOutputEqualsSorted("Output was not correct.", expectedOutput, testHarness.getOutput(), new ResultSortComparator());
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key1", 1), initialTime + 10999));
testHarness.processElement(new StreamRecord<>(new Tuple2<>("key2", 1), initialTime + 1002));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key1", 4), Long.MAX_VALUE));
expectedOutput.add(new StreamRecord<>(new Tuple2<>("key2", 5), Long.MAX_VALUE));
TestHarnessUtil.assertOutputEqualsSorted("Output was not correct.", expectedOutput, testHarness.getOutput(), new ResultSortComparator());
testHarness.close();
Assert.assertEquals("Close was not called.", 1, closeCalled.get());
}
Aggregations