Java Code Examples for org.apache.flink.runtime.operators.testutils.MatchRemovingJoiner

The following examples show how to use org.apache.flink.runtime.operators.testutils.MatchRemovingJoiner. These examples are extracted from open source projects. You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example 1
@Test
public void testMerge() {
	try {

		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator input1 = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator input2 = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
				collectData(input1),
				collectData(input2));

		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction =
				new MatchRemovingJoiner(expectedMatchesMap);

		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		// reset the generators
		generator1.reset();
		generator2.reset();
		input1.reset();
		input2.reset();

		// compare with iterator values
		NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(joinFunction, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			Assert.assertTrue("Collection for key " + entry.getKey() + " is not empty", entry.getValue().isEmpty());
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 2
@Test
public void testMergeWithHighNumberOfCommonKeys()
{
	// the size of the left and right inputs
	final int INPUT_1_SIZE = 200;
	final int INPUT_2_SIZE = 100;
	
	final int INPUT_1_DUPLICATES = 10;
	final int INPUT_2_DUPLICATES = 4000;
	final int DUPLICATE_KEY = 13;
	
	try {
		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator gen1Iter = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator gen2Iter = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		final TestData.TupleConstantValueIterator const1Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "LEFT String for Duplicate Keys", INPUT_1_DUPLICATES);
		final TestData.TupleConstantValueIterator const2Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "RIGHT String for Duplicate Keys", INPUT_2_DUPLICATES);

		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList1 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);

		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList2 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		MutableObjectIterator<Tuple2<Integer, String>> input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		MutableObjectIterator<Tuple2<Integer, String>> input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));
		
		// re-create the whole thing for actual processing
		
		// reset the generators and iterators
		generator1.reset();
		generator2.reset();
		const1Iter.reset();
		const2Iter.reset();
		gen1Iter.reset();
		gen2Iter.reset();
		
		inList1.clear();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		inList2.clear();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction = new MatchRemovingJoiner(expectedMatchesMap);
		
		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		
		// we create this sort-merge iterator with little memory for the block-nested-loops fall-back to make sure it
		// needs to spill for the duplicate keys
		NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(joinFunction, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			if (!entry.getValue().isEmpty()) {
				Assert.fail("Collection for key " + entry.getKey() + " is not empty");
			}
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 3
@Test
public void testMerge() {
	try {

		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator input1 = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator input2 = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));

		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction =
				new MatchRemovingJoiner(expectedMatchesMap);

		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		// reset the generators
		generator1.reset();
		generator2.reset();
		input1.reset();
		input2.reset();

		// compare with iterator values
		ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();

		while (iterator.callWithNextKey(joinFunction, collector));

		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			Assert.assertTrue("Collection for key " + entry.getKey() + " is not empty", entry.getValue().isEmpty());
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 4
@Test
public void testMergeWithHighNumberOfCommonKeys()
{
	// the size of the left and right inputs
	final int INPUT_1_SIZE = 200;
	final int INPUT_2_SIZE = 100;

	final int INPUT_1_DUPLICATES = 10;
	final int INPUT_2_DUPLICATES = 4000;
	final int DUPLICATE_KEY = 13;

	try {
		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		
		final TestData.TupleGeneratorIterator gen1Iter = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator gen2Iter = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);
		
		final TestData.TupleConstantValueIterator const1Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "LEFT String for Duplicate Keys", INPUT_1_DUPLICATES);
		final TestData.TupleConstantValueIterator const2Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "RIGHT String for Duplicate Keys", INPUT_2_DUPLICATES);
		
		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList1 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList2 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);
		
		MutableObjectIterator<Tuple2<Integer, String>> input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		MutableObjectIterator<Tuple2<Integer, String>> input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));
		
		// re-create the whole thing for actual processing
		
		// reset the generators and iterators
		generator1.reset();
		generator2.reset();
		const1Iter.reset();
		const2Iter.reset();
		gen1Iter.reset();
		gen2Iter.reset();
		
		inList1.clear();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		inList2.clear();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> matcher = new MatchRemovingJoiner(expectedMatchesMap);
		
		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		
		// we create this sort-merge iterator with little memory for the block-nested-loops fall-back to make sure it
		// needs to spill for the duplicate keys
		ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(matcher, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			if (!entry.getValue().isEmpty()) {
				Assert.fail("Collection for key " + entry.getKey() + " is not empty");
			}
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 5
@Test
public void testMerge() {
	try {

		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator input1 = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator input2 = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
				collectData(input1),
				collectData(input2));

		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction =
				new MatchRemovingJoiner(expectedMatchesMap);

		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		// reset the generators
		generator1.reset();
		generator2.reset();
		input1.reset();
		input2.reset();

		// compare with iterator values
		NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(joinFunction, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			Assert.assertTrue("Collection for key " + entry.getKey() + " is not empty", entry.getValue().isEmpty());
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 6
@Test
public void testMergeWithHighNumberOfCommonKeys()
{
	// the size of the left and right inputs
	final int INPUT_1_SIZE = 200;
	final int INPUT_2_SIZE = 100;
	
	final int INPUT_1_DUPLICATES = 10;
	final int INPUT_2_DUPLICATES = 4000;
	final int DUPLICATE_KEY = 13;
	
	try {
		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator gen1Iter = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator gen2Iter = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		final TestData.TupleConstantValueIterator const1Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "LEFT String for Duplicate Keys", INPUT_1_DUPLICATES);
		final TestData.TupleConstantValueIterator const2Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "RIGHT String for Duplicate Keys", INPUT_2_DUPLICATES);

		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList1 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);

		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList2 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		MutableObjectIterator<Tuple2<Integer, String>> input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		MutableObjectIterator<Tuple2<Integer, String>> input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));
		
		// re-create the whole thing for actual processing
		
		// reset the generators and iterators
		generator1.reset();
		generator2.reset();
		const1Iter.reset();
		const2Iter.reset();
		gen1Iter.reset();
		gen2Iter.reset();
		
		inList1.clear();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		inList2.clear();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction = new MatchRemovingJoiner(expectedMatchesMap);
		
		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		
		// we create this sort-merge iterator with little memory for the block-nested-loops fall-back to make sure it
		// needs to spill for the duplicate keys
		NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(joinFunction, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			if (!entry.getValue().isEmpty()) {
				Assert.fail("Collection for key " + entry.getKey() + " is not empty");
			}
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 7
@Test
public void testMerge() {
	try {

		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator input1 = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator input2 = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));

		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction =
				new MatchRemovingJoiner(expectedMatchesMap);

		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		// reset the generators
		generator1.reset();
		generator2.reset();
		input1.reset();
		input2.reset();

		// compare with iterator values
		ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();

		while (iterator.callWithNextKey(joinFunction, collector));

		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			Assert.assertTrue("Collection for key " + entry.getKey() + " is not empty", entry.getValue().isEmpty());
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 8
@Test
public void testMergeWithHighNumberOfCommonKeys()
{
	// the size of the left and right inputs
	final int INPUT_1_SIZE = 200;
	final int INPUT_2_SIZE = 100;

	final int INPUT_1_DUPLICATES = 10;
	final int INPUT_2_DUPLICATES = 4000;
	final int DUPLICATE_KEY = 13;

	try {
		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		
		final TestData.TupleGeneratorIterator gen1Iter = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator gen2Iter = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);
		
		final TestData.TupleConstantValueIterator const1Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "LEFT String for Duplicate Keys", INPUT_1_DUPLICATES);
		final TestData.TupleConstantValueIterator const2Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "RIGHT String for Duplicate Keys", INPUT_2_DUPLICATES);
		
		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList1 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList2 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);
		
		MutableObjectIterator<Tuple2<Integer, String>> input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		MutableObjectIterator<Tuple2<Integer, String>> input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));
		
		// re-create the whole thing for actual processing
		
		// reset the generators and iterators
		generator1.reset();
		generator2.reset();
		const1Iter.reset();
		const2Iter.reset();
		gen1Iter.reset();
		gen2Iter.reset();
		
		inList1.clear();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		inList2.clear();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> matcher = new MatchRemovingJoiner(expectedMatchesMap);
		
		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		
		// we create this sort-merge iterator with little memory for the block-nested-loops fall-back to make sure it
		// needs to spill for the duplicate keys
		ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(matcher, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			if (!entry.getValue().isEmpty()) {
				Assert.fail("Collection for key " + entry.getKey() + " is not empty");
			}
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 9
@Test
public void testMerge() {
	try {

		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator input1 = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator input2 = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
				collectData(input1),
				collectData(input2));

		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction =
				new MatchRemovingJoiner(expectedMatchesMap);

		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		// reset the generators
		generator1.reset();
		generator2.reset();
		input1.reset();
		input2.reset();

		// compare with iterator values
		NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(joinFunction, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			Assert.assertTrue("Collection for key " + entry.getKey() + " is not empty", entry.getValue().isEmpty());
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 10
@Test
public void testMergeWithHighNumberOfCommonKeys()
{
	// the size of the left and right inputs
	final int INPUT_1_SIZE = 200;
	final int INPUT_2_SIZE = 100;
	
	final int INPUT_1_DUPLICATES = 10;
	final int INPUT_2_DUPLICATES = 4000;
	final int DUPLICATE_KEY = 13;
	
	try {
		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator gen1Iter = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator gen2Iter = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		final TestData.TupleConstantValueIterator const1Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "LEFT String for Duplicate Keys", INPUT_1_DUPLICATES);
		final TestData.TupleConstantValueIterator const2Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "RIGHT String for Duplicate Keys", INPUT_2_DUPLICATES);

		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList1 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);

		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList2 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		MutableObjectIterator<Tuple2<Integer, String>> input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		MutableObjectIterator<Tuple2<Integer, String>> input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));
		
		// re-create the whole thing for actual processing
		
		// reset the generators and iterators
		generator1.reset();
		generator2.reset();
		const1Iter.reset();
		const2Iter.reset();
		gen1Iter.reset();
		gen2Iter.reset();
		
		inList1.clear();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		inList2.clear();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction = new MatchRemovingJoiner(expectedMatchesMap);
		
		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		
		// we create this sort-merge iterator with little memory for the block-nested-loops fall-back to make sure it
		// needs to spill for the duplicate keys
		NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new NonReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(joinFunction, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			if (!entry.getValue().isEmpty()) {
				Assert.fail("Collection for key " + entry.getKey() + " is not empty");
			}
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 11
@Test
public void testMerge() {
	try {

		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);

		final TestData.TupleGeneratorIterator input1 = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator input2 = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);

		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));

		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> joinFunction =
				new MatchRemovingJoiner(expectedMatchesMap);

		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		// reset the generators
		generator1.reset();
		generator2.reset();
		input1.reset();
		input2.reset();

		// compare with iterator values
		ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();

		while (iterator.callWithNextKey(joinFunction, collector));

		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			Assert.assertTrue("Collection for key " + entry.getKey() + " is not empty", entry.getValue().isEmpty());
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}
 
Example 12
@Test
public void testMergeWithHighNumberOfCommonKeys()
{
	// the size of the left and right inputs
	final int INPUT_1_SIZE = 200;
	final int INPUT_2_SIZE = 100;

	final int INPUT_1_DUPLICATES = 10;
	final int INPUT_2_DUPLICATES = 4000;
	final int DUPLICATE_KEY = 13;

	try {
		final TupleGenerator generator1 = new TupleGenerator(SEED1, 500, 4096, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		final TupleGenerator generator2 = new TupleGenerator(SEED2, 500, 2048, KeyMode.SORTED, ValueMode.RANDOM_LENGTH);
		
		final TestData.TupleGeneratorIterator gen1Iter = new TestData.TupleGeneratorIterator(generator1, INPUT_1_SIZE);
		final TestData.TupleGeneratorIterator gen2Iter = new TestData.TupleGeneratorIterator(generator2, INPUT_2_SIZE);
		
		final TestData.TupleConstantValueIterator const1Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "LEFT String for Duplicate Keys", INPUT_1_DUPLICATES);
		final TestData.TupleConstantValueIterator const2Iter = new TestData.TupleConstantValueIterator(DUPLICATE_KEY, "RIGHT String for Duplicate Keys", INPUT_2_DUPLICATES);
		
		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList1 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		final List<MutableObjectIterator<Tuple2<Integer, String>>> inList2 = new ArrayList<MutableObjectIterator<Tuple2<Integer, String>>>();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);
		
		MutableObjectIterator<Tuple2<Integer, String>> input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		MutableObjectIterator<Tuple2<Integer, String>> input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		// collect expected data
		final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
			collectData(input1),
			collectData(input2));
		
		// re-create the whole thing for actual processing
		
		// reset the generators and iterators
		generator1.reset();
		generator2.reset();
		const1Iter.reset();
		const2Iter.reset();
		gen1Iter.reset();
		gen2Iter.reset();
		
		inList1.clear();
		inList1.add(gen1Iter);
		inList1.add(const1Iter);
		
		inList2.clear();
		inList2.add(gen2Iter);
		inList2.add(const2Iter);

		input1 = new MergeIterator<Tuple2<Integer, String>>(inList1, comparator1.duplicate());
		input2 = new MergeIterator<Tuple2<Integer, String>>(inList2, comparator2.duplicate());
		
		final FlatJoinFunction<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> matcher = new MatchRemovingJoiner(expectedMatchesMap);
		
		final Collector<Tuple2<Integer, String>> collector = new DiscardingOutputCollector<Tuple2<Integer, String>>();

		
		// we create this sort-merge iterator with little memory for the block-nested-loops fall-back to make sure it
		// needs to spill for the duplicate keys
		ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>> iterator =
			new ReusingMergeInnerJoinIterator<Tuple2<Integer, String>, Tuple2<Integer, String>, Tuple2<Integer, String>>(
				input1, input2, this.serializer1, this.comparator1, this.serializer2, this.comparator2,
				this.pairComparator, this.memoryManager, this.ioManager, PAGES_FOR_BNLJN, this.parentTask);

		iterator.open();
		
		while (iterator.callWithNextKey(matcher, collector));
		
		iterator.close();

		// assert that each expected match was seen
		for (Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
			if (!entry.getValue().isEmpty()) {
				Assert.fail("Collection for key " + entry.getKey() + " is not empty");
			}
		}
	}
	catch (Exception e) {
		e.printStackTrace();
		Assert.fail("An exception occurred during the test: " + e.getMessage());
	}
}