Search in sources :

Example 6 with Filter

use of org.apache.spark.sql.sources.Filter in project Gaffer by gchq.

the class FilterToOperationConverterTest method testSpecifyPropertyFilters.

@Test
public void testSpecifyPropertyFilters() throws OperationException {
    final Schema schema = getSchema();
    final SQLContext sqlContext = getSqlContext("testSpecifyPropertyFilters");
    final Filter[] filters = new Filter[1];
    // GreaterThan
    filters[0] = new GreaterThan("property1", 5);
    FiltersToOperationConverter converter = new FiltersToOperationConverter(sqlContext, getViewFromSchema(schema), schema, filters);
    AbstractGetRDD<?> operation = converter.getOperation();
    assertTrue(operation instanceof GetRDDOfAllElements);
    View opView = operation.getView();
    List<ConsumerFunctionContext<String, FilterFunction>> entityPostAggFilters = opView.getEntity(ENTITY_GROUP).getPostAggregationFilterFunctions();
    assertEquals(1, entityPostAggFilters.size());
    assertEquals(new ArrayList<>(Collections.singleton("property1")), entityPostAggFilters.get(0).getSelection());
    assertEquals(new IsMoreThan(5, false), entityPostAggFilters.get(0).getFunction());
    for (final String edgeGroup : EDGE_GROUPS) {
        final List<ConsumerFunctionContext<String, FilterFunction>> edgePostAggFilters = opView.getEdge(edgeGroup).getPostAggregationFilterFunctions();
        assertEquals(1, edgePostAggFilters.size());
        assertEquals(new ArrayList<>(Collections.singleton("property1")), edgePostAggFilters.get(0).getSelection());
        assertEquals(new IsMoreThan(5, false), edgePostAggFilters.get(0).getFunction());
    }
    // LessThan
    filters[0] = new LessThan("property4", 8L);
    converter = new FiltersToOperationConverter(sqlContext, getViewFromSchema(schema), schema, filters);
    operation = converter.getOperation();
    assertTrue(operation instanceof GetRDDOfAllElements);
    // Only groups ENTITY_GROUP and EDGE_GROUP should be in the view as only they have property4
    opView = operation.getView();
    entityPostAggFilters = opView.getEntity(ENTITY_GROUP).getPostAggregationFilterFunctions();
    assertEquals(1, entityPostAggFilters.size());
    assertEquals(new ArrayList<>(Collections.singleton("property4")), entityPostAggFilters.get(0).getSelection());
    assertEquals(new IsLessThan(8L, false), entityPostAggFilters.get(0).getFunction());
    List<ConsumerFunctionContext<String, FilterFunction>> edgePostAggFilters = opView.getEdge(EDGE_GROUP).getPostAggregationFilterFunctions();
    assertEquals(1, edgePostAggFilters.size());
    assertEquals(new ArrayList<>(Collections.singleton("property4")), edgePostAggFilters.get(0).getSelection());
    assertEquals(new IsLessThan(8L, false), edgePostAggFilters.get(0).getFunction());
    // And
    final Filter left = new GreaterThan("property1", 5);
    final Filter right = new GreaterThan("property4", 8L);
    filters[0] = new And(left, right);
    converter = new FiltersToOperationConverter(sqlContext, getViewFromSchema(schema), schema, filters);
    operation = converter.getOperation();
    assertTrue(operation instanceof GetRDDOfAllElements);
    // Only groups ENTITY_GROUP and EDGE_GROUP should be in the view as only they have property1 and property4
    opView = operation.getView();
    entityPostAggFilters = opView.getEntity(ENTITY_GROUP).getPostAggregationFilterFunctions();
    assertEquals(2, entityPostAggFilters.size());
    final ArrayList<String> expectedProperties = new ArrayList<>();
    expectedProperties.add("property1");
    expectedProperties.add("property4");
    assertEquals(1, entityPostAggFilters.get(0).getSelection().size());
    assertEquals(expectedProperties.get(0), entityPostAggFilters.get(0).getSelection().get(0));
    assertEquals(1, entityPostAggFilters.get(1).getSelection().size());
    assertEquals(expectedProperties.get(1), entityPostAggFilters.get(1).getSelection().get(0));
    final ArrayList<FilterFunction> expectedFunctions = new ArrayList<>();
    expectedFunctions.add(new IsMoreThan(5, false));
    expectedFunctions.add(new IsMoreThan(8L, false));
    assertEquals(expectedFunctions.get(0), entityPostAggFilters.get(0).getFunction());
    assertEquals(expectedFunctions.get(1), entityPostAggFilters.get(1).getFunction());
    edgePostAggFilters = opView.getEdge(EDGE_GROUP).getPostAggregationFilterFunctions();
    assertEquals(2, edgePostAggFilters.size());
    assertEquals(1, edgePostAggFilters.get(0).getSelection().size());
    assertEquals(expectedProperties.get(0), edgePostAggFilters.get(0).getSelection().get(0));
    assertEquals(1, edgePostAggFilters.get(1).getSelection().size());
    assertEquals(expectedProperties.get(1), edgePostAggFilters.get(1).getSelection().get(0));
    sqlContext.sparkContext().stop();
}
Also used : FilterFunction(uk.gov.gchq.gaffer.function.FilterFunction) Schema(uk.gov.gchq.gaffer.store.schema.Schema) ArrayList(java.util.ArrayList) View(uk.gov.gchq.gaffer.data.elementdefinition.view.View) LessThan(org.apache.spark.sql.sources.LessThan) IsLessThan(uk.gov.gchq.gaffer.function.filter.IsLessThan) ConsumerFunctionContext(uk.gov.gchq.gaffer.function.context.ConsumerFunctionContext) IsLessThan(uk.gov.gchq.gaffer.function.filter.IsLessThan) Filter(org.apache.spark.sql.sources.Filter) GreaterThan(org.apache.spark.sql.sources.GreaterThan) GetRDDOfAllElements(uk.gov.gchq.gaffer.spark.operation.scalardd.GetRDDOfAllElements) And(org.apache.spark.sql.sources.And) IsMoreThan(uk.gov.gchq.gaffer.function.filter.IsMoreThan) SQLContext(org.apache.spark.sql.SQLContext) Test(org.junit.Test)

Example 7 with Filter

use of org.apache.spark.sql.sources.Filter in project Gaffer by gchq.

the class FilterToOperationConverterTest method testSingleGroupNotInSchema.

@Test
public void testSingleGroupNotInSchema() throws OperationException {
    final Schema schema = getSchema();
    final SQLContext sqlContext = getSqlContext("testSingleGroupNotInSchema");
    final Filter[] filters = new Filter[1];
    filters[0] = new EqualTo(SchemaToStructTypeConverter.GROUP, "random");
    final FiltersToOperationConverter converter = new FiltersToOperationConverter(sqlContext, getViewFromSchema(schema), schema, filters);
    final AbstractGetRDD<?> operation = converter.getOperation();
    assertNull(operation);
    sqlContext.sparkContext().stop();
}
Also used : Filter(org.apache.spark.sql.sources.Filter) Schema(uk.gov.gchq.gaffer.store.schema.Schema) SQLContext(org.apache.spark.sql.SQLContext) EqualTo(org.apache.spark.sql.sources.EqualTo) Test(org.junit.Test)

Example 8 with Filter

use of org.apache.spark.sql.sources.Filter in project Gaffer by gchq.

the class FilterToOperationConverterTest method testSpecifySourceOrDestinationAndPropertyFilter.

@Test
public void testSpecifySourceOrDestinationAndPropertyFilter() throws OperationException {
    final Schema schema = getSchema();
    final SQLContext sqlContext = getSqlContext("testSpecifyVertexAndPropertyFilter");
    // Specify src and a filter on property1
    Filter[] filters = new Filter[2];
    filters[0] = new GreaterThan("property1", 5);
    filters[1] = new EqualTo(SchemaToStructTypeConverter.SRC_COL_NAME, "0");
    FiltersToOperationConverter converter = new FiltersToOperationConverter(sqlContext, getViewFromSchema(schema), schema, filters);
    AbstractGetRDD<?> operation = converter.getOperation();
    assertTrue(operation instanceof GetRDDOfElements);
    assertEquals(0, operation.getView().getEntityGroups().size());
    assertEquals(2, operation.getView().getEdgeGroups().size());
    final Set<EntitySeed> seeds = new HashSet<>();
    for (final Object seed : ((GetRDDOfElements) operation).getSeeds()) {
        seeds.add((EntitySeed) seed);
    }
    assertEquals(Collections.singleton(new EntitySeed("0")), seeds);
    View opView = operation.getView();
    for (final String edgeGroup : EDGE_GROUPS) {
        final List<ConsumerFunctionContext<String, FilterFunction>> edgePostAggFilters = opView.getEdge(edgeGroup).getPostAggregationFilterFunctions();
        assertEquals(1, edgePostAggFilters.size());
        assertEquals(new ArrayList<>(Collections.singleton("property1")), edgePostAggFilters.get(0).getSelection());
        assertEquals(new IsMoreThan(5, false), edgePostAggFilters.get(0).getFunction());
    }
    // Specify src and filters on property1 and property4
    filters = new Filter[3];
    filters[0] = new GreaterThan("property1", 5);
    filters[1] = new EqualTo(SchemaToStructTypeConverter.SRC_COL_NAME, "0");
    filters[2] = new LessThan("property4", 8);
    converter = new FiltersToOperationConverter(sqlContext, getViewFromSchema(schema), schema, filters);
    operation = converter.getOperation();
    assertTrue(operation instanceof GetRDDOfElements);
    assertEquals(0, operation.getView().getEntityGroups().size());
    assertEquals(1, operation.getView().getEdgeGroups().size());
    seeds.clear();
    for (final Object seed : ((GetRDDOfElements) operation).getSeeds()) {
        seeds.add((EntitySeed) seed);
    }
    assertEquals(Collections.singleton(new EntitySeed("0")), seeds);
    opView = operation.getView();
    final List<ConsumerFunctionContext<String, FilterFunction>> entityPostAggFilters = opView.getEdge(EDGE_GROUP).getPostAggregationFilterFunctions();
    assertEquals(2, entityPostAggFilters.size());
    final List<String> expectedProperties = new ArrayList<>();
    expectedProperties.add("property1");
    expectedProperties.add("property4");
    assertEquals(1, entityPostAggFilters.get(0).getSelection().size());
    assertEquals(expectedProperties.get(0), entityPostAggFilters.get(0).getSelection().get(0));
    assertEquals(new IsMoreThan(5, false), entityPostAggFilters.get(0).getFunction());
    assertEquals(1, entityPostAggFilters.get(1).getSelection().size());
    assertEquals(expectedProperties.get(1), entityPostAggFilters.get(1).getSelection().get(0));
    assertEquals(new IsLessThan(8, false), entityPostAggFilters.get(1).getFunction());
    sqlContext.sparkContext().stop();
}
Also used : Schema(uk.gov.gchq.gaffer.store.schema.Schema) ArrayList(java.util.ArrayList) View(uk.gov.gchq.gaffer.data.elementdefinition.view.View) EqualTo(org.apache.spark.sql.sources.EqualTo) LessThan(org.apache.spark.sql.sources.LessThan) IsLessThan(uk.gov.gchq.gaffer.function.filter.IsLessThan) ConsumerFunctionContext(uk.gov.gchq.gaffer.function.context.ConsumerFunctionContext) IsLessThan(uk.gov.gchq.gaffer.function.filter.IsLessThan) Filter(org.apache.spark.sql.sources.Filter) GreaterThan(org.apache.spark.sql.sources.GreaterThan) EntitySeed(uk.gov.gchq.gaffer.operation.data.EntitySeed) IsMoreThan(uk.gov.gchq.gaffer.function.filter.IsMoreThan) SQLContext(org.apache.spark.sql.SQLContext) GetRDDOfElements(uk.gov.gchq.gaffer.spark.operation.scalardd.GetRDDOfElements) HashSet(java.util.HashSet) Test(org.junit.Test)

Example 9 with Filter

use of org.apache.spark.sql.sources.Filter in project Gaffer by gchq.

the class FilterToOperationConverterTest method testSpecifyVertex.

@Test
public void testSpecifyVertex() throws OperationException {
    final Schema schema = getSchema();
    final SQLContext sqlContext = getSqlContext("testSpecifyVertex");
    final Filter[] filters = new Filter[1];
    filters[0] = new EqualTo(SchemaToStructTypeConverter.VERTEX_COL_NAME, "0");
    final FiltersToOperationConverter converter = new FiltersToOperationConverter(sqlContext, getViewFromSchema(schema), schema, filters);
    final AbstractGetRDD<?> operation = converter.getOperation();
    assertTrue(operation instanceof GetRDDOfElements);
    assertEquals(Collections.singleton(ENTITY_GROUP), operation.getView().getEntityGroups());
    assertEquals(0, operation.getView().getEdgeGroups().size());
    final Set<EntitySeed> seeds = new HashSet<>();
    for (final Object seed : ((GetRDDOfElements) operation).getSeeds()) {
        seeds.add((EntitySeed) seed);
    }
    assertEquals(Collections.singleton(new EntitySeed("0")), seeds);
    sqlContext.sparkContext().stop();
}
Also used : Filter(org.apache.spark.sql.sources.Filter) Schema(uk.gov.gchq.gaffer.store.schema.Schema) EntitySeed(uk.gov.gchq.gaffer.operation.data.EntitySeed) SQLContext(org.apache.spark.sql.SQLContext) EqualTo(org.apache.spark.sql.sources.EqualTo) GetRDDOfElements(uk.gov.gchq.gaffer.spark.operation.scalardd.GetRDDOfElements) HashSet(java.util.HashSet) Test(org.junit.Test)

Example 10 with Filter

use of org.apache.spark.sql.sources.Filter in project Gaffer by gchq.

the class FilterToOperationConverterTest method testSpecifyMultiplePropertyFilters.

@Test
public void testSpecifyMultiplePropertyFilters() throws OperationException {
    final Schema schema = getSchema();
    final SQLContext sqlContext = getSqlContext("testSpecifyMultiplePropertyFilters");
    final Filter[] filters = new Filter[2];
    filters[0] = new GreaterThan("property1", 5);
    filters[1] = new LessThan("property4", 8L);
    FiltersToOperationConverter converter = new FiltersToOperationConverter(sqlContext, getViewFromSchema(schema), schema, filters);
    AbstractGetRDD<?> operation = converter.getOperation();
    assertTrue(operation instanceof GetRDDOfAllElements);
    // Only groups ENTITY_GROUP and EDGE_GROUP should be in the view as only they have property1 and property4
    View opView = operation.getView();
    List<ConsumerFunctionContext<String, FilterFunction>> entityPostAggFilters = opView.getEntity(ENTITY_GROUP).getPostAggregationFilterFunctions();
    assertEquals(2, entityPostAggFilters.size());
    final ArrayList<String> expectedProperties = new ArrayList<>();
    expectedProperties.add("property1");
    expectedProperties.add("property4");
    assertEquals(1, entityPostAggFilters.get(0).getSelection().size());
    assertEquals(expectedProperties.get(0), entityPostAggFilters.get(0).getSelection().get(0));
    assertEquals(1, entityPostAggFilters.get(1).getSelection().size());
    assertEquals(expectedProperties.get(1), entityPostAggFilters.get(1).getSelection().get(0));
    final ArrayList<FilterFunction> expectedFunctions = new ArrayList<>();
    expectedFunctions.add(new IsMoreThan(5, false));
    expectedFunctions.add(new IsLessThan(8L, false));
    assertEquals(expectedFunctions.get(0), entityPostAggFilters.get(0).getFunction());
    assertEquals(expectedFunctions.get(1), entityPostAggFilters.get(1).getFunction());
    final List<ConsumerFunctionContext<String, FilterFunction>> edgePostAggFilters = opView.getEdge(EDGE_GROUP).getPostAggregationFilterFunctions();
    assertEquals(2, edgePostAggFilters.size());
    assertEquals(1, edgePostAggFilters.get(0).getSelection().size());
    assertEquals(expectedProperties.get(0), edgePostAggFilters.get(0).getSelection().get(0));
    assertEquals(1, edgePostAggFilters.get(1).getSelection().size());
    assertEquals(expectedProperties.get(1), edgePostAggFilters.get(1).getSelection().get(0));
    sqlContext.sparkContext().stop();
}
Also used : FilterFunction(uk.gov.gchq.gaffer.function.FilterFunction) Schema(uk.gov.gchq.gaffer.store.schema.Schema) ArrayList(java.util.ArrayList) View(uk.gov.gchq.gaffer.data.elementdefinition.view.View) LessThan(org.apache.spark.sql.sources.LessThan) IsLessThan(uk.gov.gchq.gaffer.function.filter.IsLessThan) ConsumerFunctionContext(uk.gov.gchq.gaffer.function.context.ConsumerFunctionContext) IsLessThan(uk.gov.gchq.gaffer.function.filter.IsLessThan) Filter(org.apache.spark.sql.sources.Filter) GreaterThan(org.apache.spark.sql.sources.GreaterThan) GetRDDOfAllElements(uk.gov.gchq.gaffer.spark.operation.scalardd.GetRDDOfAllElements) IsMoreThan(uk.gov.gchq.gaffer.function.filter.IsMoreThan) SQLContext(org.apache.spark.sql.SQLContext) Test(org.junit.Test)

Aggregations

Filter (org.apache.spark.sql.sources.Filter)15 Test (org.junit.Test)12 Schema (uk.gov.gchq.gaffer.store.schema.Schema)12 SQLContext (org.apache.spark.sql.SQLContext)11 EqualTo (org.apache.spark.sql.sources.EqualTo)10 HashSet (java.util.HashSet)7 View (uk.gov.gchq.gaffer.data.elementdefinition.view.View)7 ArrayList (java.util.ArrayList)6 EntitySeed (uk.gov.gchq.gaffer.operation.data.EntitySeed)6 GetRDDOfElements (uk.gov.gchq.gaffer.spark.operation.scalardd.GetRDDOfElements)6 GreaterThan (org.apache.spark.sql.sources.GreaterThan)5 ConsumerFunctionContext (uk.gov.gchq.gaffer.function.context.ConsumerFunctionContext)5 GetRDDOfAllElements (uk.gov.gchq.gaffer.spark.operation.scalardd.GetRDDOfAllElements)5 LessThan (org.apache.spark.sql.sources.LessThan)4 FilterFunction (uk.gov.gchq.gaffer.function.FilterFunction)4 IsLessThan (uk.gov.gchq.gaffer.function.filter.IsLessThan)4 IsMoreThan (uk.gov.gchq.gaffer.function.filter.IsMoreThan)4 Set (java.util.Set)2 HashMap (java.util.HashMap)1 List (java.util.List)1