Java Code Examples for org.apache.flink.runtime.operators.testutils.TestData.TupleGenerator.KeyMode#SORTED
The following examples show how to use
org.apache.flink.runtime.operators.testutils.TestData.TupleGenerator.KeyMode#SORTED .
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
| Source File: ReOpenableHashTableTestBase.java From Flink-CEPplus with Apache License 2.0 | 6 votes |
|
/**
* Verify proper operation if the build side is spilled to disk.
*/
@Test
public void testDoubleProbeSpilling() {
int buildSize = 1000;
int probeSize = 1000;
try {
TupleGenerator bgen = new TupleGenerator(SEED1, 0, 1024, KeyMode.SORTED, ValueMode.FIX_LENGTH);
TupleGenerator pgen = new TupleGenerator(SEED2, 0, 1024, KeyMode.SORTED, ValueMode.FIX_LENGTH);
final TupleGeneratorIterator buildInput = new TupleGeneratorIterator(bgen, buildSize);
final TupleGeneratorIterator probeInput = new TupleGeneratorIterator(pgen, probeSize);
doTest(buildInput,probeInput, bgen, pgen);
}
catch (Exception e) {
e.printStackTrace();
Assert.fail("An exception occurred during the test: " + e.getMessage());
}
}
Example 2
| Source File: ReOpenableHashTableTestBase.java From flink with Apache License 2.0 | 6 votes |
|
/**
* Test behavior with overflow buckets (Overflow buckets must be initialized correctly
* if the input is reopened again)
*/
@Test
public void testOverflow() {
int buildSize = 1000;
int probeSize = 1000;
try {
TupleGenerator bgen = new TupleGenerator(SEED1, 200, 1024, KeyMode.RANDOM, ValueMode.FIX_LENGTH);
TupleGenerator pgen = new TupleGenerator(SEED2, 0, 1024, KeyMode.SORTED, ValueMode.FIX_LENGTH);
final TupleGeneratorIterator buildInput = new TupleGeneratorIterator(bgen, buildSize);
final TupleGeneratorIterator probeInput = new TupleGeneratorIterator(pgen, probeSize);
doTest(buildInput,probeInput, bgen, pgen);
}
catch (Exception e) {
e.printStackTrace();
Assert.fail("An exception occurred during the test: " + e.getMessage());
}
}
Example 3
| Source File: ReOpenableHashTableTestBase.java From flink with Apache License 2.0 | 6 votes |
|
/**
* This test case verifies that hybrid hash join is able to handle multiple probe phases
* when the build side fits completely into memory.
*/
@Test
public void testDoubleProbeInMemory() {
int buildSize = 1000;
int probeSize = 1000;
try {
TupleGenerator bgen = new TupleGenerator(SEED1, 0, 28, KeyMode.SORTED, ValueMode.FIX_LENGTH);
TupleGenerator pgen = new TupleGenerator(SEED2, 0, 28, KeyMode.SORTED, ValueMode.FIX_LENGTH);
final TupleGeneratorIterator buildInput = new TupleGeneratorIterator(bgen, buildSize);
final TupleGeneratorIterator probeInput = new TupleGeneratorIterator(pgen, probeSize);
doTest(buildInput,probeInput, bgen, pgen);
}
catch (Exception e) {
e.printStackTrace();
Assert.fail("An exception occurred during the test: " + e.getMessage());
}
}
Example 4
| Source File: ReOpenableHashTableTestBase.java From flink with Apache License 2.0 | 6 votes |
|
/**
* Test behavior with overflow buckets (Overflow buckets must be initialized correctly
* if the input is reopened again)
*/
@Test
public void testOverflow() {
int buildSize = 1000;
int probeSize = 1000;
try {
TupleGenerator bgen = new TupleGenerator(SEED1, 200, 1024, KeyMode.RANDOM, ValueMode.FIX_LENGTH);
TupleGenerator pgen = new TupleGenerator(SEED2, 0, 1024, KeyMode.SORTED, ValueMode.FIX_LENGTH);
final TupleGeneratorIterator buildInput = new TupleGeneratorIterator(bgen, buildSize);
final TupleGeneratorIterator probeInput = new TupleGeneratorIterator(pgen, probeSize);
doTest(buildInput,probeInput, bgen, pgen);
}
catch (Exception e) {
e.printStackTrace();
Assert.fail("An exception occurred during the test: " + e.getMessage());
}
}
Example 5
| Source File: ReOpenableHashTableTestBase.java From Flink-CEPplus with Apache License 2.0 | 6 votes |
|
/**
* This test case verifies that hybrid hash join is able to handle multiple probe phases
* when the build side fits completely into memory.
*/
@Test
public void testDoubleProbeInMemory() {
int buildSize = 1000;
int probeSize = 1000;
try {
TupleGenerator bgen = new TupleGenerator(SEED1, 0, 28, KeyMode.SORTED, ValueMode.FIX_LENGTH);
TupleGenerator pgen = new TupleGenerator(SEED2, 0, 28, KeyMode.SORTED, ValueMode.FIX_LENGTH);
final TupleGeneratorIterator buildInput = new TupleGeneratorIterator(bgen, buildSize);
final TupleGeneratorIterator probeInput = new TupleGeneratorIterator(pgen, probeSize);
doTest(buildInput,probeInput, bgen, pgen);
}
catch (Exception e) {
e.printStackTrace();
Assert.fail("An exception occurred during the test: " + e.getMessage());
}
}
Example 6
| Source File: NormalizedKeySorterTest.java From Flink-CEPplus with Apache License 2.0 | 5 votes |
|
/**
* The compare test creates a sorted stream, writes it to the buffer and
* compares random elements. It expects that earlier elements are lower than later
* ones.
*/
@Test
public void testCompare() throws Exception {
final int numSegments = MEMORY_SIZE / MEMORY_PAGE_SIZE;
final List<MemorySegment> memory = this.memoryManager.allocatePages(new DummyInvokable(), numSegments);
NormalizedKeySorter<Tuple2<Integer, String>> sorter = newSortBuffer(memory);
TestData.TupleGenerator generator = new TestData.TupleGenerator(SEED, KEY_MAX, VALUE_LENGTH, KeyMode.SORTED,
ValueMode.RANDOM_LENGTH);
// write the records
Tuple2<Integer, String> record = new Tuple2<>();
int num = -1;
do {
generator.next(record);
num++;
}
while (sorter.write(record));
// compare random elements
Random rnd = new Random(SEED << 1);
for (int i = 0; i < 2 * num; i++) {
int pos1 = rnd.nextInt(num);
int pos2 = rnd.nextInt(num);
int cmp = sorter.compare(pos1, pos2);
if (pos1 < pos2) {
Assert.assertTrue(cmp <= 0);
}
else {
Assert.assertTrue(cmp >= 0);
}
}
// release the memory occupied by the buffers
sorter.dispose();
this.memoryManager.release(memory);
}
Example 7
| Source File: RandomSortMergeInnerJoinTest.java From flink with Apache License 2.0 | 5 votes |
|
@Test
public void test() throws Exception {
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_FIRST_SIZE);
final TestData.TupleGeneratorIterator input2 = new TestData.TupleGeneratorIterator(generator2, INPUT_SECOND_SIZE);
// collect expected data
final Map<Integer, Collection<Match>> expectedMatchesMap = matchValues(
collectData(input1), collectData(input2));
// reset the generators
generator1.reset();
generator2.reset();
input1.reset();
input2.reset();
StreamOperator operator = getOperator();
match(expectedMatchesMap, transformToBinary(join(operator, input1, input2)));
// assert that each expected match was seen
for (Map.Entry<Integer, Collection<Match>> entry : expectedMatchesMap.entrySet()) {
Assert.assertTrue("Collection for key " + entry.getKey() + " is not empty", entry.getValue().isEmpty());
}
}
Example 8
| Source File: NormalizedKeySorterTest.java From flink with Apache License 2.0 | 5 votes |
|
/**
* The compare test creates a sorted stream, writes it to the buffer and
* compares random elements. It expects that earlier elements are lower than later
* ones.
*/
@Test
public void testCompare() throws Exception {
final int numSegments = MEMORY_SIZE / MEMORY_PAGE_SIZE;
final List<MemorySegment> memory = this.memoryManager.allocatePages(new DummyInvokable(), numSegments);
NormalizedKeySorter<Tuple2<Integer, String>> sorter = newSortBuffer(memory);
TestData.TupleGenerator generator = new TestData.TupleGenerator(SEED, KEY_MAX, VALUE_LENGTH, KeyMode.SORTED,
ValueMode.RANDOM_LENGTH);
// write the records
Tuple2<Integer, String> record = new Tuple2<>();
int num = -1;
do {
generator.next(record);
num++;
}
while (sorter.write(record));
// compare random elements
Random rnd = new Random(SEED << 1);
for (int i = 0; i < 2 * num; i++) {
int pos1 = rnd.nextInt(num);
int pos2 = rnd.nextInt(num);
int cmp = sorter.compare(pos1, pos2);
if (pos1 < pos2) {
Assert.assertTrue(cmp <= 0);
}
else {
Assert.assertTrue(cmp >= 0);
}
}
// release the memory occupied by the buffers
sorter.dispose();
this.memoryManager.release(memory);
}
Example 9
| Source File: NormalizedKeySorterTest.java From flink with Apache License 2.0 | 5 votes |
|
/**
* The compare test creates a sorted stream, writes it to the buffer and
* compares random elements. It expects that earlier elements are lower than later
* ones.
*/
@Test
public void testCompare() throws Exception {
final int numSegments = MEMORY_SIZE / MEMORY_PAGE_SIZE;
final List<MemorySegment> memory = this.memoryManager.allocatePages(new DummyInvokable(), numSegments);
NormalizedKeySorter<Tuple2<Integer, String>> sorter = newSortBuffer(memory);
TestData.TupleGenerator generator = new TestData.TupleGenerator(SEED, KEY_MAX, VALUE_LENGTH, KeyMode.SORTED,
ValueMode.RANDOM_LENGTH);
// write the records
Tuple2<Integer, String> record = new Tuple2<>();
int num = -1;
do {
generator.next(record);
num++;
}
while (sorter.write(record));
// compare random elements
Random rnd = new Random(SEED << 1);
for (int i = 0; i < 2 * num; i++) {
int pos1 = rnd.nextInt(num);
int pos2 = rnd.nextInt(num);
int cmp = sorter.compare(pos1, pos2);
if (pos1 < pos2) {
Assert.assertTrue(cmp <= 0);
}
else {
Assert.assertTrue(cmp >= 0);
}
}
// release the memory occupied by the buffers
sorter.dispose();
this.memoryManager.release(memory);
}
Example 10
| Source File: NonReusingSortMergeInnerJoinIteratorITCase.java From Flink-CEPplus with Apache License 2.0 | 4 votes |
|
@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
| Source File: ReusingSortMergeInnerJoinIteratorITCase.java From Flink-CEPplus with Apache License 2.0 | 4 votes |
|
@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
| Source File: ReusingSortMergeInnerJoinIteratorITCase.java From flink with Apache License 2.0 | 4 votes |
|
@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 13
| Source File: RandomSortMergeInnerJoinTest.java From flink with Apache License 2.0 | 4 votes |
|
@Test
public void testMergeWithHighNumberOfCommonKeys() {
// the size of the left and right inputs
final int input1Size = 200;
final int input2Size = 100;
final int input1Duplicates = 10;
final int input2Duplicates = 4000;
final int duplicateKey = 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, input1Size);
final TestData.TupleGeneratorIterator gen2Iter = new TestData.TupleGeneratorIterator(generator2, input2Size);
final TestData.TupleConstantValueIterator const1Iter = new TestData.TupleConstantValueIterator(duplicateKey, "LEFT String for Duplicate Keys", input1Duplicates);
final TestData.TupleConstantValueIterator const2Iter = new TestData.TupleConstantValueIterator(duplicateKey, "RIGHT String for Duplicate Keys", input2Duplicates);
final List<MutableObjectIterator<Tuple2<Integer, String>>> inList1 = new ArrayList<>();
inList1.add(gen1Iter);
inList1.add(const1Iter);
final List<MutableObjectIterator<Tuple2<Integer, String>>> inList2 = new ArrayList<>();
inList2.add(gen2Iter);
inList2.add(const2Iter);
MutableObjectIterator<Tuple2<Integer, String>> input1 = new MergeIterator<>(inList1, comparator1.duplicate());
MutableObjectIterator<Tuple2<Integer, String>> input2 = new MergeIterator<>(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<>(inList1, comparator1.duplicate());
input2 = new MergeIterator<>(inList2, comparator2.duplicate());
StreamOperator operator = getOperator();
match(expectedMatchesMap, transformToBinary(join(operator, input1, input2)));
// assert that each expected match was seen
for (Map.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 14
| Source File: NonReusingSortMergeInnerJoinIteratorITCase.java From Flink-CEPplus with Apache License 2.0 | 4 votes |
|
@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 15
| Source File: NonReusingSortMergeInnerJoinIteratorITCase.java From flink with Apache License 2.0 | 4 votes |
|
@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 16
| Source File: NonReusingSortMergeInnerJoinIteratorITCase.java From flink with Apache License 2.0 | 4 votes |
|
@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 17
| Source File: ReusingSortMergeInnerJoinIteratorITCase.java From flink with Apache License 2.0 | 4 votes |
|
@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 18
| Source File: ReusingSortMergeInnerJoinIteratorITCase.java From flink with Apache License 2.0 | 4 votes |
|
@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 19
| Source File: ReusingSortMergeInnerJoinIteratorITCase.java From flink with Apache License 2.0 | 4 votes |
|
@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 20
| Source File: RandomSortMergeInnerJoinTest.java From flink with Apache License 2.0 | 4 votes |
|
@Test
public void testMergeWithHighNumberOfCommonKeys() {
// the size of the left and right inputs
final int input1Size = 200;
final int input2Size = 100;
final int input1Duplicates = 10;
final int input2Duplicates = 4000;
final int duplicateKey = 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, input1Size);
final TestData.TupleGeneratorIterator gen2Iter = new TestData.TupleGeneratorIterator(generator2, input2Size);
final TestData.TupleConstantValueIterator const1Iter = new TestData.TupleConstantValueIterator(duplicateKey, "LEFT String for Duplicate Keys", input1Duplicates);
final TestData.TupleConstantValueIterator const2Iter = new TestData.TupleConstantValueIterator(duplicateKey, "RIGHT String for Duplicate Keys", input2Duplicates);
final List<MutableObjectIterator<Tuple2<Integer, String>>> inList1 = new ArrayList<>();
inList1.add(gen1Iter);
inList1.add(const1Iter);
final List<MutableObjectIterator<Tuple2<Integer, String>>> inList2 = new ArrayList<>();
inList2.add(gen2Iter);
inList2.add(const2Iter);
MutableObjectIterator<Tuple2<Integer, String>> input1 = new MergeIterator<>(inList1, comparator1.duplicate());
MutableObjectIterator<Tuple2<Integer, String>> input2 = new MergeIterator<>(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<>(inList1, comparator1.duplicate());
input2 = new MergeIterator<>(inList2, comparator2.duplicate());
StreamOperator operator = getOperator();
match(expectedMatchesMap, transformToBinary(join(operator, input1, input2)));
// assert that each expected match was seen
for (Map.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());
}
}
