use of org.apache.flink.streaming.api.functions.source.SourceFunction in project flink by apache.
the class FoldApplyProcessWindowFunctionTest method testFoldWindowFunctionOutputTypeConfigurable.
/**
* Tests that the FoldWindowFunction gets the output type serializer set by the
* StreamGraphGenerator and checks that the FoldWindowFunction computes the correct result.
*/
@Test
public void testFoldWindowFunctionOutputTypeConfigurable() throws Exception {
StreamExecutionEnvironment env = new DummyStreamExecutionEnvironment();
List<StreamTransformation<?>> transformations = new ArrayList<>();
int initValue = 1;
FoldApplyProcessWindowFunction<Integer, TimeWindow, Integer, Integer, Integer> foldWindowFunction = new FoldApplyProcessWindowFunction<>(initValue, new FoldFunction<Integer, Integer>() {
@Override
public Integer fold(Integer accumulator, Integer value) throws Exception {
return accumulator + value;
}
}, new ProcessWindowFunction<Integer, Integer, Integer, TimeWindow>() {
@Override
public void process(Integer integer, Context context, Iterable<Integer> input, Collector<Integer> out) throws Exception {
for (Integer in : input) {
out.collect(in);
}
}
}, BasicTypeInfo.INT_TYPE_INFO);
AccumulatingProcessingTimeWindowOperator<Integer, Integer, Integer> windowOperator = new AccumulatingProcessingTimeWindowOperator<>(new InternalIterableProcessWindowFunction<>(foldWindowFunction), new KeySelector<Integer, Integer>() {
private static final long serialVersionUID = -7951310554369722809L;
@Override
public Integer getKey(Integer value) throws Exception {
return value;
}
}, IntSerializer.INSTANCE, IntSerializer.INSTANCE, 3000, 3000);
SourceFunction<Integer> sourceFunction = new SourceFunction<Integer>() {
private static final long serialVersionUID = 8297735565464653028L;
@Override
public void run(SourceContext<Integer> ctx) throws Exception {
}
@Override
public void cancel() {
}
};
SourceTransformation<Integer> source = new SourceTransformation<>("", new StreamSource<>(sourceFunction), BasicTypeInfo.INT_TYPE_INFO, 1);
transformations.add(new OneInputTransformation<>(source, "test", windowOperator, BasicTypeInfo.INT_TYPE_INFO, 1));
StreamGraph streamGraph = StreamGraphGenerator.generate(env, transformations, 1);
List<Integer> result = new ArrayList<>();
List<Integer> input = new ArrayList<>();
List<Integer> expected = new ArrayList<>();
input.add(1);
input.add(2);
input.add(3);
for (int value : input) {
initValue += value;
}
expected.add(initValue);
foldWindowFunction.process(0, foldWindowFunction.new Context() {
@Override
public TimeWindow window() {
return new TimeWindow(0, 1);
}
}, input, new ListCollector<>(result));
Assert.assertEquals(expected, result);
}
use of org.apache.flink.streaming.api.functions.source.SourceFunction in project flink by apache.
the class WindowFoldITCase method testFoldWindow.
@Test
public void testFoldWindow() throws Exception {
testResults = new ArrayList<>();
StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);
env.setParallelism(1);
DataStream<Tuple2<String, Integer>> source1 = env.addSource(new SourceFunction<Tuple2<String, Integer>>() {
private static final long serialVersionUID = 1L;
@Override
public void run(SourceContext<Tuple2<String, Integer>> ctx) throws Exception {
ctx.collect(Tuple2.of("a", 0));
ctx.collect(Tuple2.of("a", 1));
ctx.collect(Tuple2.of("a", 2));
ctx.collect(Tuple2.of("b", 3));
ctx.collect(Tuple2.of("b", 4));
ctx.collect(Tuple2.of("b", 5));
ctx.collect(Tuple2.of("a", 6));
ctx.collect(Tuple2.of("a", 7));
ctx.collect(Tuple2.of("a", 8));
// source is finite, so it will have an implicit MAX watermark when it finishes
}
@Override
public void cancel() {
}
}).assignTimestampsAndWatermarks(new Tuple2TimestampExtractor());
source1.keyBy(0).window(TumblingEventTimeWindows.of(Time.of(3, TimeUnit.MILLISECONDS))).fold(Tuple2.of("R:", 0), new FoldFunction<Tuple2<String, Integer>, Tuple2<String, Integer>>() {
@Override
public Tuple2<String, Integer> fold(Tuple2<String, Integer> accumulator, Tuple2<String, Integer> value) throws Exception {
accumulator.f0 += value.f0;
accumulator.f1 += value.f1;
return accumulator;
}
}).addSink(new SinkFunction<Tuple2<String, Integer>>() {
@Override
public void invoke(Tuple2<String, Integer> value) throws Exception {
testResults.add(value.toString());
}
});
env.execute("Fold Window Test");
List<String> expectedResult = Arrays.asList("(R:aaa,3)", "(R:aaa,21)", "(R:bbb,12)");
Collections.sort(expectedResult);
Collections.sort(testResults);
Assert.assertEquals(expectedResult, testResults);
}
use of org.apache.flink.streaming.api.functions.source.SourceFunction in project flink by apache.
the class WindowFoldITCase method testFoldAllWindow.
@Test
public void testFoldAllWindow() throws Exception {
testResults = new ArrayList<>();
StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);
env.setParallelism(1);
DataStream<Tuple2<String, Integer>> source1 = env.addSource(new SourceFunction<Tuple2<String, Integer>>() {
@Override
public void run(SourceContext<Tuple2<String, Integer>> ctx) throws Exception {
ctx.collect(Tuple2.of("a", 0));
ctx.collect(Tuple2.of("a", 1));
ctx.collect(Tuple2.of("a", 2));
ctx.collect(Tuple2.of("b", 3));
ctx.collect(Tuple2.of("a", 3));
ctx.collect(Tuple2.of("b", 4));
ctx.collect(Tuple2.of("a", 4));
ctx.collect(Tuple2.of("b", 5));
ctx.collect(Tuple2.of("a", 5));
// source is finite, so it will have an implicit MAX watermark when it finishes
}
@Override
public void cancel() {
}
}).assignTimestampsAndWatermarks(new Tuple2TimestampExtractor());
source1.windowAll(TumblingEventTimeWindows.of(Time.of(3, TimeUnit.MILLISECONDS))).fold(Tuple2.of("R:", 0), new FoldFunction<Tuple2<String, Integer>, Tuple2<String, Integer>>() {
@Override
public Tuple2<String, Integer> fold(Tuple2<String, Integer> accumulator, Tuple2<String, Integer> value) throws Exception {
accumulator.f0 += value.f0;
accumulator.f1 += value.f1;
return accumulator;
}
}).addSink(new SinkFunction<Tuple2<String, Integer>>() {
@Override
public void invoke(Tuple2<String, Integer> value) throws Exception {
testResults.add(value.toString());
}
});
env.execute("Fold All-Window Test");
List<String> expectedResult = Arrays.asList("(R:aaa,3)", "(R:bababa,24)");
Collections.sort(expectedResult);
Collections.sort(testResults);
Assert.assertEquals(expectedResult, testResults);
}
use of org.apache.flink.streaming.api.functions.source.SourceFunction in project flink by apache.
the class ChangelogStateBackendLoadingTest method getEnvironment.
private StreamExecutionEnvironment getEnvironment() {
StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
SourceFunction<Integer> srcFun = new SourceFunction<Integer>() {
private static final long serialVersionUID = 1L;
@Override
public void run(SourceContext<Integer> ctx) throws Exception {
}
@Override
public void cancel() {
}
};
SingleOutputStreamOperator<Object> operator = env.addSource(srcFun).flatMap(new FlatMapFunction<Integer, Object>() {
private static final long serialVersionUID = 1L;
@Override
public void flatMap(Integer value, Collector<Object> out) throws Exception {
}
});
operator.setParallelism(1);
return env;
}
use of org.apache.flink.streaming.api.functions.source.SourceFunction in project flink by apache.
the class CoGroupJoinITCase method testJoin.
@Test
public void testJoin() throws Exception {
testResults = new ArrayList<>();
StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
env.setParallelism(1);
DataStream<Tuple3<String, String, Integer>> source1 = env.addSource(new SourceFunction<Tuple3<String, String, Integer>>() {
@Override
public void run(SourceContext<Tuple3<String, String, Integer>> ctx) throws Exception {
ctx.collect(Tuple3.of("a", "x", 0));
ctx.collect(Tuple3.of("a", "y", 1));
ctx.collect(Tuple3.of("a", "z", 2));
ctx.collect(Tuple3.of("b", "u", 3));
ctx.collect(Tuple3.of("b", "w", 5));
ctx.collect(Tuple3.of("a", "i", 6));
ctx.collect(Tuple3.of("a", "j", 7));
ctx.collect(Tuple3.of("a", "k", 8));
// source is finite, so it will have an implicit MAX
// watermark when it finishes
}
@Override
public void cancel() {
}
}).assignTimestampsAndWatermarks(new Tuple3TimestampExtractor());
DataStream<Tuple3<String, String, Integer>> source2 = env.addSource(new SourceFunction<Tuple3<String, String, Integer>>() {
@Override
public void run(SourceContext<Tuple3<String, String, Integer>> ctx) throws Exception {
ctx.collect(Tuple3.of("a", "u", 0));
ctx.collect(Tuple3.of("a", "w", 1));
ctx.collect(Tuple3.of("b", "i", 3));
ctx.collect(Tuple3.of("b", "k", 5));
ctx.collect(Tuple3.of("a", "x", 6));
ctx.collect(Tuple3.of("a", "z", 8));
// source is finite, so it will have an implicit MAX
// watermark when it finishes
}
@Override
public void cancel() {
}
}).assignTimestampsAndWatermarks(new Tuple3TimestampExtractor());
source1.join(source2).where(new Tuple3KeyExtractor()).equalTo(new Tuple3KeyExtractor()).window(TumblingEventTimeWindows.of(Time.of(3, TimeUnit.MILLISECONDS))).apply(new JoinFunction<Tuple3<String, String, Integer>, Tuple3<String, String, Integer>, String>() {
@Override
public String join(Tuple3<String, String, Integer> first, Tuple3<String, String, Integer> second) throws Exception {
return first + ":" + second;
}
}).addSink(new SinkFunction<String>() {
@Override
public void invoke(String value) throws Exception {
testResults.add(value);
}
});
env.execute("Join Test");
List<String> expectedResult = Arrays.asList("(a,x,0):(a,u,0)", "(a,x,0):(a,w,1)", "(a,y,1):(a,u,0)", "(a,y,1):(a,w,1)", "(a,z,2):(a,u,0)", "(a,z,2):(a,w,1)", "(b,u,3):(b,i,3)", "(b,u,3):(b,k,5)", "(b,w,5):(b,i,3)", "(b,w,5):(b,k,5)", "(a,i,6):(a,x,6)", "(a,i,6):(a,z,8)", "(a,j,7):(a,x,6)", "(a,j,7):(a,z,8)", "(a,k,8):(a,x,6)", "(a,k,8):(a,z,8)");
Collections.sort(expectedResult);
Collections.sort(testResults);
Assert.assertEquals(expectedResult, testResults);
}
Aggregations