use of org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor in project asterixdb by apache.
the class AggregationTest method singleKeyMinMaxStringExtGroupTest.
@Test
public void singleKeyMinMaxStringExtGroupTest() throws Exception {
JobSpecification spec = new JobSpecification();
FileScanOperatorDescriptor csvScanner = new FileScanOperatorDescriptor(spec, splitProvider, tupleParserFactory, desc);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, csvScanner, NC2_ID);
RecordDescriptor outputRec = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer() });
int[] keyFields = new int[] { 0 };
int frameLimits = 5;
int tableSize = 8;
long fileSize = frameLimits * spec.getFrameSize();
ExternalGroupOperatorDescriptor grouper = new ExternalGroupOperatorDescriptor(spec, tableSize, fileSize, keyFields, frameLimits, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, new UTF8StringNormalizedKeyComputerFactory(), new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new IntSumFieldAggregatorFactory(1, false), new MinMaxStringFieldAggregatorFactory(15, true, true) }), new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new IntSumFieldAggregatorFactory(1, false), new MinMaxStringFieldAggregatorFactory(2, true, true) }), outputRec, outputRec, new HashSpillableTableFactory(new IBinaryHashFunctionFamily[] { UTF8StringBinaryHashFunctionFamily.INSTANCE }));
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, grouper, NC2_ID, NC1_ID);
IConnectorDescriptor conn1 = new MToNPartitioningConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(keyFields, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }));
spec.connect(conn1, csvScanner, 0, grouper, 0);
AbstractSingleActivityOperatorDescriptor printer = getPrinter(spec, "singleKeyAvgExtGroupTest");
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC2_ID, NC1_ID);
IConnectorDescriptor conn2 = new OneToOneConnectorDescriptor(spec);
spec.connect(conn2, grouper, 0, printer, 0);
spec.addRoot(printer);
runTest(spec);
}
use of org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor in project asterixdb by apache.
the class AggregationTest method singleKeyMinMaxStringPreClusterGroupTest.
@Test
public void singleKeyMinMaxStringPreClusterGroupTest() throws Exception {
JobSpecification spec = new JobSpecification();
FileScanOperatorDescriptor csvScanner = new FileScanOperatorDescriptor(spec, splitProvider, tupleParserFactory, desc);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, csvScanner, NC2_ID);
RecordDescriptor outputRec = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE, new UTF8StringSerializerDeserializer() });
int[] keyFields = new int[] { 0 };
PreclusteredGroupOperatorDescriptor grouper = new PreclusteredGroupOperatorDescriptor(spec, keyFields, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new IntSumFieldAggregatorFactory(1, true), new MinMaxStringFieldAggregatorFactory(15, true, false) }), outputRec);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, grouper, NC2_ID, NC1_ID);
IConnectorDescriptor conn1 = new MToNPartitioningConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(keyFields, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }));
spec.connect(conn1, csvScanner, 0, grouper, 0);
AbstractSingleActivityOperatorDescriptor printer = getPrinter(spec, "singleKeyAvgInmemGroupTest");
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC2_ID, NC1_ID);
IConnectorDescriptor conn2 = new OneToOneConnectorDescriptor(spec);
spec.connect(conn2, grouper, 0, printer, 0);
spec.addRoot(printer);
runTest(spec);
}
use of org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor in project asterixdb by apache.
the class AggregationTest method multiKeySumPreClusterGroupTest.
@Test
public void multiKeySumPreClusterGroupTest() throws Exception {
JobSpecification spec = new JobSpecification();
FileScanOperatorDescriptor csvScanner = new FileScanOperatorDescriptor(spec, splitProvider, tupleParserFactory, desc);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, csvScanner, NC2_ID);
RecordDescriptor outputRec = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), IntegerSerializerDeserializer.INSTANCE, IntegerSerializerDeserializer.INSTANCE });
int[] keyFields = new int[] { 8, 0 };
PreclusteredGroupOperatorDescriptor grouper = new PreclusteredGroupOperatorDescriptor(spec, keyFields, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY), PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, new MultiFieldsAggregatorFactory(new IFieldAggregateDescriptorFactory[] { new IntSumFieldAggregatorFactory(1, true), new IntSumFieldAggregatorFactory(3, true) }), outputRec);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, grouper, NC2_ID, NC1_ID);
IConnectorDescriptor conn1 = new MToNPartitioningConnectorDescriptor(spec, new FieldHashPartitionComputerFactory(keyFields, new IBinaryHashFunctionFactory[] { PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY), PointableBinaryHashFunctionFactory.of(UTF8StringPointable.FACTORY) }));
spec.connect(conn1, csvScanner, 0, grouper, 0);
AbstractSingleActivityOperatorDescriptor printer = getPrinter(spec, "multiKeySumInmemGroupTest");
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC2_ID, NC1_ID);
IConnectorDescriptor conn2 = new OneToOneConnectorDescriptor(spec);
spec.connect(conn2, grouper, 0, printer, 0);
spec.addRoot(printer);
runTest(spec);
}
use of org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor in project asterixdb by apache.
the class AbstractRTreeOperatorTest method loadPrimaryIndex.
protected void loadPrimaryIndex() throws Exception {
JobSpecification spec = new JobSpecification();
FileSplit[] ordersSplits = new FileSplit[] { new ManagedFileSplit(NC1_ID, "data" + File.separator + "orders-with-locations-part1.txt") };
IFileSplitProvider ordersSplitProvider = new ConstantFileSplitProvider(ordersSplits);
RecordDescriptor ordersDesc = new RecordDescriptor(new ISerializerDeserializer[] { new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), new UTF8StringSerializerDeserializer(), DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE });
FileScanOperatorDescriptor ordScanner = new FileScanOperatorDescriptor(spec, ordersSplitProvider, new DelimitedDataTupleParserFactory(new IValueParserFactory[] { UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, UTF8StringParserFactory.INSTANCE, DoubleParserFactory.INSTANCE, DoubleParserFactory.INSTANCE, DoubleParserFactory.INSTANCE, DoubleParserFactory.INSTANCE }, '|'), ordersDesc);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, ordScanner, NC1_ID);
ExternalSortOperatorDescriptor sorter = new ExternalSortOperatorDescriptor(spec, 1000, new int[] { 0 }, new IBinaryComparatorFactory[] { PointableBinaryComparatorFactory.of(UTF8StringPointable.FACTORY) }, ordersDesc);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, sorter, NC1_ID);
int[] fieldPermutation = { 0, 1, 2, 4, 5, 7, 9, 10, 11, 12 };
TreeIndexBulkLoadOperatorDescriptor primaryBulkLoad = new TreeIndexBulkLoadOperatorDescriptor(spec, primaryRecDesc, fieldPermutation, 0.7f, false, 1000L, true, primaryHelperFactory);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, primaryBulkLoad, NC1_ID);
NullSinkOperatorDescriptor nsOpDesc = new NullSinkOperatorDescriptor(spec);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, nsOpDesc, NC1_ID);
spec.connect(new OneToOneConnectorDescriptor(spec), ordScanner, 0, sorter, 0);
spec.connect(new OneToOneConnectorDescriptor(spec), sorter, 0, primaryBulkLoad, 0);
spec.connect(new OneToOneConnectorDescriptor(spec), primaryBulkLoad, 0, nsOpDesc, 0);
spec.addRoot(nsOpDesc);
runTest(spec);
}
use of org.apache.hyracks.dataflow.std.connectors.OneToOneConnectorDescriptor in project asterixdb by apache.
the class RTreeSecondaryIndexInsertOperatorTest method searchUpdatedSecondaryIndexTest.
@Test
public void searchUpdatedSecondaryIndexTest() throws Exception {
JobSpecification spec = new JobSpecification();
// build tuple
ArrayTupleBuilder tb = new ArrayTupleBuilder(secondaryKeyFieldCount);
DataOutput dos = tb.getDataOutput();
tb.reset();
DoubleSerializerDeserializer.INSTANCE.serialize(61.2894, dos);
tb.addFieldEndOffset();
DoubleSerializerDeserializer.INSTANCE.serialize(-149.624, dos);
tb.addFieldEndOffset();
DoubleSerializerDeserializer.INSTANCE.serialize(61.8894, dos);
tb.addFieldEndOffset();
DoubleSerializerDeserializer.INSTANCE.serialize(-149.024, dos);
tb.addFieldEndOffset();
ISerializerDeserializer[] keyRecDescSers = { DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE, DoubleSerializerDeserializer.INSTANCE };
RecordDescriptor keyRecDesc = new RecordDescriptor(keyRecDescSers);
ConstantTupleSourceOperatorDescriptor keyProviderOp = new ConstantTupleSourceOperatorDescriptor(spec, keyRecDesc, tb.getFieldEndOffsets(), tb.getByteArray(), tb.getSize());
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, keyProviderOp, NC1_ID);
int[] keyFields = { 0, 1, 2, 3 };
RTreeSearchOperatorDescriptor secondarySearchOp = new RTreeSearchOperatorDescriptor(spec, secondaryRecDesc, keyFields, true, true, secondaryHelperFactory, false, false, null, NoOpOperationCallbackFactory.INSTANCE, null, null, false);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, secondarySearchOp, NC1_ID);
// fifth field from the tuples coming from secondary index
int[] primaryLowKeyFields = { 4 };
// fifth field from the tuples coming from secondary index
int[] primaryHighKeyFields = { 4 };
// search primary index
BTreeSearchOperatorDescriptor primarySearchOp = new BTreeSearchOperatorDescriptor(spec, primaryRecDesc, primaryLowKeyFields, primaryHighKeyFields, true, true, primaryHelperFactory, false, false, null, NoOpOperationCallbackFactory.INSTANCE, null, null, false);
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, primarySearchOp, NC1_ID);
IFileSplitProvider outSplits = new ConstantFileSplitProvider(new FileSplit[] { createFile(nc1) });
IOperatorDescriptor printer = new PlainFileWriterOperatorDescriptor(spec, outSplits, ",");
PartitionConstraintHelper.addAbsoluteLocationConstraint(spec, printer, NC1_ID);
spec.connect(new OneToOneConnectorDescriptor(spec), keyProviderOp, 0, secondarySearchOp, 0);
spec.connect(new OneToOneConnectorDescriptor(spec), secondarySearchOp, 0, primarySearchOp, 0);
spec.connect(new OneToOneConnectorDescriptor(spec), primarySearchOp, 0, printer, 0);
spec.addRoot(printer);
runTest(spec);
}
Aggregations