Search in sources :

Example 11 with Constraints

use of com.amazonaws.athena.connector.lambda.domain.predicate.Constraints in project aws-athena-query-federation by awslabs.

the class HiveRecordHandlerTest method buildSplitSql.

@Test
public void buildSplitSql() throws SQLException {
    TableName tableName = new TableName("testSchema", "testTable");
    SchemaBuilder schemaBuilder = SchemaBuilder.newBuilder();
    schemaBuilder.addField(FieldBuilder.newBuilder("testCol1", Types.MinorType.INT.getType()).build());
    schemaBuilder.addField(FieldBuilder.newBuilder("testCol2", Types.MinorType.DATEDAY.getType()).build());
    schemaBuilder.addField(FieldBuilder.newBuilder("testCol3", Types.MinorType.DATEMILLI.getType()).build());
    schemaBuilder.addField(FieldBuilder.newBuilder("testCol4", Types.MinorType.VARBINARY.getType()).build());
    schemaBuilder.addField(FieldBuilder.newBuilder("partition", Types.MinorType.VARCHAR.getType()).build());
    Schema schema = schemaBuilder.build();
    Split split = Mockito.mock(Split.class);
    Mockito.when(split.getProperties()).thenReturn(Collections.singletonMap("partition", "p0"));
    Mockito.when(split.getProperty(Mockito.eq("partition"))).thenReturn("p0");
    Range range1a = Mockito.mock(Range.class, Mockito.RETURNS_DEEP_STUBS);
    Mockito.when(range1a.isSingleValue()).thenReturn(true);
    Mockito.when(range1a.getLow().getValue()).thenReturn(1);
    Range range1b = Mockito.mock(Range.class, Mockito.RETURNS_DEEP_STUBS);
    Mockito.when(range1b.isSingleValue()).thenReturn(true);
    Mockito.when(range1b.getLow().getValue()).thenReturn(2);
    ValueSet valueSet1 = Mockito.mock(SortedRangeSet.class, Mockito.RETURNS_DEEP_STUBS);
    Mockito.when(valueSet1.getRanges().getOrderedRanges()).thenReturn(ImmutableList.of(range1a, range1b));
    final long dateDays = TimeUnit.DAYS.toDays(Date.valueOf("2020-01-05").getTime());
    ValueSet valueSet2 = getSingleValueSet(dateDays);
    Constraints constraints = Mockito.mock(Constraints.class);
    Mockito.when(constraints.getSummary()).thenReturn(new ImmutableMap.Builder<String, ValueSet>().put("testCol2", valueSet2).build());
    PreparedStatement expectedPreparedStatement = Mockito.mock(PreparedStatement.class);
    Mockito.when(this.connection.prepareStatement(Mockito.anyString())).thenReturn(expectedPreparedStatement);
    PreparedStatement preparedStatement = this.hiveRecordHandler.buildSplitSql(this.connection, "testCatalogName", tableName, schema, constraints, split);
    Assert.assertEquals(expectedPreparedStatement, preparedStatement);
}
Also used : TableName(com.amazonaws.athena.connector.lambda.domain.TableName) Constraints(com.amazonaws.athena.connector.lambda.domain.predicate.Constraints) Schema(org.apache.arrow.vector.types.pojo.Schema) SchemaBuilder(com.amazonaws.athena.connector.lambda.data.SchemaBuilder) PreparedStatement(java.sql.PreparedStatement) Split(com.amazonaws.athena.connector.lambda.domain.Split) Range(com.amazonaws.athena.connector.lambda.domain.predicate.Range) ValueSet(com.amazonaws.athena.connector.lambda.domain.predicate.ValueSet) ImmutableMap(com.google.common.collect.ImmutableMap) Test(org.junit.Test)

Example 12 with Constraints

use of com.amazonaws.athena.connector.lambda.domain.predicate.Constraints in project aws-athena-query-federation by awslabs.

the class MySqlMetadataHandlerTest method doGetSplitsContinuation.

@Test
public void doGetSplitsContinuation() throws Exception {
    BlockAllocator blockAllocator = new BlockAllocatorImpl();
    Constraints constraints = Mockito.mock(Constraints.class);
    TableName tableName = new TableName("testSchema", "testTable");
    Schema partitionSchema = this.mySqlMetadataHandler.getPartitionSchema("testCatalogName");
    Set<String> partitionCols = partitionSchema.getFields().stream().map(Field::getName).collect(Collectors.toSet());
    GetTableLayoutRequest getTableLayoutRequest = new GetTableLayoutRequest(this.federatedIdentity, "testQueryId", "testCatalogName", tableName, constraints, partitionSchema, partitionCols);
    PreparedStatement preparedStatement = Mockito.mock(PreparedStatement.class);
    Mockito.when(this.connection.prepareStatement(MySqlMetadataHandler.GET_PARTITIONS_QUERY)).thenReturn(preparedStatement);
    String[] columns = { "partition_name" };
    int[] types = { Types.VARCHAR };
    Object[][] values = { { "p0" }, { "p1" } };
    ResultSet resultSet = mockResultSet(columns, types, values, new AtomicInteger(-1));
    final String expectedQuery = String.format(MySqlMetadataHandler.GET_PARTITIONS_QUERY, tableName.getTableName(), tableName.getSchemaName());
    Mockito.when(preparedStatement.executeQuery()).thenReturn(resultSet);
    Mockito.when(this.connection.getMetaData().getSearchStringEscape()).thenReturn(null);
    GetTableLayoutResponse getTableLayoutResponse = this.mySqlMetadataHandler.doGetTableLayout(blockAllocator, getTableLayoutRequest);
    BlockAllocator splitBlockAllocator = new BlockAllocatorImpl();
    GetSplitsRequest getSplitsRequest = new GetSplitsRequest(this.federatedIdentity, "testQueryId", "testCatalogName", tableName, getTableLayoutResponse.getPartitions(), new ArrayList<>(partitionCols), constraints, "1");
    GetSplitsResponse getSplitsResponse = this.mySqlMetadataHandler.doGetSplits(splitBlockAllocator, getSplitsRequest);
    Set<Map<String, String>> expectedSplits = new HashSet<>();
    expectedSplits.add(Collections.singletonMap("partition_name", "p1"));
    Assert.assertEquals(expectedSplits.size(), getSplitsResponse.getSplits().size());
    Set<Map<String, String>> actualSplits = getSplitsResponse.getSplits().stream().map(Split::getProperties).collect(Collectors.toSet());
    Assert.assertEquals(expectedSplits, actualSplits);
}
Also used : GetSplitsRequest(com.amazonaws.athena.connector.lambda.metadata.GetSplitsRequest) Schema(org.apache.arrow.vector.types.pojo.Schema) PreparedStatement(java.sql.PreparedStatement) TableName(com.amazonaws.athena.connector.lambda.domain.TableName) Constraints(com.amazonaws.athena.connector.lambda.domain.predicate.Constraints) GetTableLayoutResponse(com.amazonaws.athena.connector.lambda.metadata.GetTableLayoutResponse) BlockAllocatorImpl(com.amazonaws.athena.connector.lambda.data.BlockAllocatorImpl) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) GetSplitsResponse(com.amazonaws.athena.connector.lambda.metadata.GetSplitsResponse) BlockAllocator(com.amazonaws.athena.connector.lambda.data.BlockAllocator) GetTableLayoutRequest(com.amazonaws.athena.connector.lambda.metadata.GetTableLayoutRequest) ResultSet(java.sql.ResultSet) Map(java.util.Map) HashSet(java.util.HashSet) Test(org.junit.Test)

Example 13 with Constraints

use of com.amazonaws.athena.connector.lambda.domain.predicate.Constraints in project aws-athena-query-federation by awslabs.

the class MySqlMetadataHandlerTest method doGetSplits.

@Test
public void doGetSplits() throws Exception {
    BlockAllocator blockAllocator = new BlockAllocatorImpl();
    Constraints constraints = Mockito.mock(Constraints.class);
    TableName tableName = new TableName("testSchema", "testTable");
    PreparedStatement preparedStatement = Mockito.mock(PreparedStatement.class);
    Mockito.when(this.connection.prepareStatement(MySqlMetadataHandler.GET_PARTITIONS_QUERY)).thenReturn(preparedStatement);
    String[] columns = { MySqlMetadataHandler.PARTITION_COLUMN_NAME };
    int[] types = { Types.VARCHAR };
    Object[][] values = { { "p0" }, { "p1" } };
    ResultSet resultSet = mockResultSet(columns, types, values, new AtomicInteger(-1));
    Mockito.when(preparedStatement.executeQuery()).thenReturn(resultSet);
    Mockito.when(this.connection.getMetaData().getSearchStringEscape()).thenReturn(null);
    Schema partitionSchema = this.mySqlMetadataHandler.getPartitionSchema("testCatalogName");
    Set<String> partitionCols = partitionSchema.getFields().stream().map(Field::getName).collect(Collectors.toSet());
    GetTableLayoutRequest getTableLayoutRequest = new GetTableLayoutRequest(this.federatedIdentity, "testQueryId", "testCatalogName", tableName, constraints, partitionSchema, partitionCols);
    GetTableLayoutResponse getTableLayoutResponse = this.mySqlMetadataHandler.doGetTableLayout(blockAllocator, getTableLayoutRequest);
    BlockAllocator splitBlockAllocator = new BlockAllocatorImpl();
    GetSplitsRequest getSplitsRequest = new GetSplitsRequest(this.federatedIdentity, "testQueryId", "testCatalogName", tableName, getTableLayoutResponse.getPartitions(), new ArrayList<>(partitionCols), constraints, null);
    GetSplitsResponse getSplitsResponse = this.mySqlMetadataHandler.doGetSplits(splitBlockAllocator, getSplitsRequest);
    Set<Map<String, String>> expectedSplits = new HashSet<>();
    expectedSplits.add(Collections.singletonMap(MySqlMetadataHandler.BLOCK_PARTITION_COLUMN_NAME, "p0"));
    expectedSplits.add(Collections.singletonMap(MySqlMetadataHandler.BLOCK_PARTITION_COLUMN_NAME, "p1"));
    Assert.assertEquals(expectedSplits.size(), getSplitsResponse.getSplits().size());
    Set<Map<String, String>> actualSplits = getSplitsResponse.getSplits().stream().map(Split::getProperties).collect(Collectors.toSet());
    Assert.assertEquals(expectedSplits, actualSplits);
}
Also used : GetSplitsRequest(com.amazonaws.athena.connector.lambda.metadata.GetSplitsRequest) Schema(org.apache.arrow.vector.types.pojo.Schema) PreparedStatement(java.sql.PreparedStatement) TableName(com.amazonaws.athena.connector.lambda.domain.TableName) Constraints(com.amazonaws.athena.connector.lambda.domain.predicate.Constraints) GetTableLayoutResponse(com.amazonaws.athena.connector.lambda.metadata.GetTableLayoutResponse) BlockAllocatorImpl(com.amazonaws.athena.connector.lambda.data.BlockAllocatorImpl) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) GetSplitsResponse(com.amazonaws.athena.connector.lambda.metadata.GetSplitsResponse) BlockAllocator(com.amazonaws.athena.connector.lambda.data.BlockAllocator) GetTableLayoutRequest(com.amazonaws.athena.connector.lambda.metadata.GetTableLayoutRequest) ResultSet(java.sql.ResultSet) Map(java.util.Map) HashSet(java.util.HashSet) Test(org.junit.Test)

Example 14 with Constraints

use of com.amazonaws.athena.connector.lambda.domain.predicate.Constraints in project aws-athena-query-federation by awslabs.

the class MySqlMetadataHandlerTest method doGetTableLayout.

@Test
public void doGetTableLayout() throws Exception {
    BlockAllocator blockAllocator = new BlockAllocatorImpl();
    Constraints constraints = Mockito.mock(Constraints.class);
    TableName tableName = new TableName("testSchema", "testTable");
    Schema partitionSchema = this.mySqlMetadataHandler.getPartitionSchema("testCatalogName");
    Set<String> partitionCols = partitionSchema.getFields().stream().map(Field::getName).collect(Collectors.toSet());
    GetTableLayoutRequest getTableLayoutRequest = new GetTableLayoutRequest(this.federatedIdentity, "testQueryId", "testCatalogName", tableName, constraints, partitionSchema, partitionCols);
    PreparedStatement preparedStatement = Mockito.mock(PreparedStatement.class);
    Mockito.when(this.connection.prepareStatement(MySqlMetadataHandler.GET_PARTITIONS_QUERY)).thenReturn(preparedStatement);
    String[] columns = { "partition_name" };
    int[] types = { Types.VARCHAR };
    Object[][] values = { { "p0" }, { "p1" } };
    ResultSet resultSet = mockResultSet(columns, types, values, new AtomicInteger(-1));
    Mockito.when(preparedStatement.executeQuery()).thenReturn(resultSet);
    Mockito.when(this.connection.getMetaData().getSearchStringEscape()).thenReturn(null);
    GetTableLayoutResponse getTableLayoutResponse = this.mySqlMetadataHandler.doGetTableLayout(blockAllocator, getTableLayoutRequest);
    Assert.assertEquals(values.length, getTableLayoutResponse.getPartitions().getRowCount());
    List<String> expectedValues = new ArrayList<>();
    for (int i = 0; i < getTableLayoutResponse.getPartitions().getRowCount(); i++) {
        expectedValues.add(BlockUtils.rowToString(getTableLayoutResponse.getPartitions(), i));
    }
    Assert.assertEquals(expectedValues, Arrays.asList("[partition_name : p0]", "[partition_name : p1]"));
    SchemaBuilder expectedSchemaBuilder = SchemaBuilder.newBuilder();
    expectedSchemaBuilder.addField(FieldBuilder.newBuilder(MySqlMetadataHandler.BLOCK_PARTITION_COLUMN_NAME, org.apache.arrow.vector.types.Types.MinorType.VARCHAR.getType()).build());
    Schema expectedSchema = expectedSchemaBuilder.build();
    Assert.assertEquals(expectedSchema, getTableLayoutResponse.getPartitions().getSchema());
    Assert.assertEquals(tableName, getTableLayoutResponse.getTableName());
    Mockito.verify(preparedStatement, Mockito.times(1)).setString(1, tableName.getTableName());
    Mockito.verify(preparedStatement, Mockito.times(1)).setString(2, tableName.getSchemaName());
}
Also used : Schema(org.apache.arrow.vector.types.pojo.Schema) ArrayList(java.util.ArrayList) PreparedStatement(java.sql.PreparedStatement) TableName(com.amazonaws.athena.connector.lambda.domain.TableName) Constraints(com.amazonaws.athena.connector.lambda.domain.predicate.Constraints) GetTableLayoutResponse(com.amazonaws.athena.connector.lambda.metadata.GetTableLayoutResponse) BlockAllocatorImpl(com.amazonaws.athena.connector.lambda.data.BlockAllocatorImpl) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) BlockAllocator(com.amazonaws.athena.connector.lambda.data.BlockAllocator) GetTableLayoutRequest(com.amazonaws.athena.connector.lambda.metadata.GetTableLayoutRequest) ResultSet(java.sql.ResultSet) SchemaBuilder(com.amazonaws.athena.connector.lambda.data.SchemaBuilder) Test(org.junit.Test)

Example 15 with Constraints

use of com.amazonaws.athena.connector.lambda.domain.predicate.Constraints in project aws-athena-query-federation by awslabs.

the class HiveMetadataHandlerTest method doGetTableLayoutWithNoPartitions.

@Test
public void doGetTableLayoutWithNoPartitions() throws Exception {
    BlockAllocator blockAllocator = new BlockAllocatorImpl();
    String[] schema = { "data_type", "col_name" };
    Object[][] values = { { "INTEGER", "case_number" }, { "VARCHAR", "case_location" }, { "TIMESTAMP", "case_instance" }, { "DATE", "case_date" } };
    AtomicInteger rowNumber = new AtomicInteger(-1);
    ResultSet resultSet = mockResultSet(schema, values, rowNumber);
    Constraints constraints = Mockito.mock(Constraints.class);
    TableName tempTableName = new TableName("testSchema", "testTable");
    Schema partitionSchema = this.hiveMetadataHandler.getPartitionSchema("testCatalogName");
    Set<String> partitionCols = new HashSet<>(Arrays.asList("partition"));
    GetTableLayoutRequest getTableLayoutRequest = new GetTableLayoutRequest(this.federatedIdentity, "testQueryId", "testCatalogName", tempTableName, constraints, partitionSchema, partitionCols);
    String[] columns2 = { "Partition" };
    int[] types2 = { Types.VARCHAR };
    Object[][] values1 = {};
    Mockito.when(jdbcConnectionFactory.getConnection(Mockito.any(JdbcCredentialProvider.class))).thenReturn(connection);
    String tableName = getTableLayoutRequest.getTableName().getTableName().toUpperCase();
    final String getPartitionDetailsSql = "show partitions " + getTableLayoutRequest.getTableName().getTableName().toUpperCase();
    Statement statement1 = Mockito.mock(Statement.class);
    PreparedStatement preparestatement1 = Mockito.mock(PreparedStatement.class);
    Mockito.when(this.connection.prepareStatement(HiveMetadataHandler.GET_METADATA_QUERY + tableName)).thenReturn(preparestatement1);
    Mockito.when(this.connection.createStatement()).thenReturn(statement1);
    ResultSet resultSet1 = mockResultSet(columns2, types2, values1, new AtomicInteger(-1));
    Mockito.when(preparestatement1.executeQuery()).thenReturn(resultSet);
    Mockito.when(statement1.executeQuery(getPartitionDetailsSql)).thenReturn(resultSet1);
    GetTableLayoutResponse getTableLayoutResponse = this.hiveMetadataHandler.doGetTableLayout(blockAllocator, getTableLayoutRequest);
    List<String> expectedValues = new ArrayList<>();
    for (int i = 0; i < getTableLayoutResponse.getPartitions().getRowCount(); i++) {
        expectedValues.add(BlockUtils.rowToString(getTableLayoutResponse.getPartitions(), i));
    }
    Assert.assertEquals(expectedValues, Arrays.asList("[partition : *]"));
    SchemaBuilder expectedSchemaBuilder = SchemaBuilder.newBuilder();
    expectedSchemaBuilder.addField(FieldBuilder.newBuilder("partition", org.apache.arrow.vector.types.Types.MinorType.VARCHAR.getType()).build());
    Schema expectedSchema = expectedSchemaBuilder.build();
    Assert.assertEquals(expectedSchema, getTableLayoutResponse.getPartitions().getSchema());
    Assert.assertEquals(tempTableName, getTableLayoutResponse.getTableName());
}
Also used : Schema(org.apache.arrow.vector.types.pojo.Schema) TableName(com.amazonaws.athena.connector.lambda.domain.TableName) Constraints(com.amazonaws.athena.connector.lambda.domain.predicate.Constraints) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) JdbcCredentialProvider(com.amazonaws.athena.connectors.jdbc.connection.JdbcCredentialProvider) Test(org.junit.Test)

Aggregations

Constraints (com.amazonaws.athena.connector.lambda.domain.predicate.Constraints)182 Test (org.junit.Test)172 TableName (com.amazonaws.athena.connector.lambda.domain.TableName)148 Schema (org.apache.arrow.vector.types.pojo.Schema)136 ValueSet (com.amazonaws.athena.connector.lambda.domain.predicate.ValueSet)64 HashMap (java.util.HashMap)63 Split (com.amazonaws.athena.connector.lambda.domain.Split)59 ReadRecordsRequest (com.amazonaws.athena.connector.lambda.records.ReadRecordsRequest)55 AtomicInteger (java.util.concurrent.atomic.AtomicInteger)54 BlockAllocator (com.amazonaws.athena.connector.lambda.data.BlockAllocator)50 GetTableLayoutRequest (com.amazonaws.athena.connector.lambda.metadata.GetTableLayoutRequest)47 BlockAllocatorImpl (com.amazonaws.athena.connector.lambda.data.BlockAllocatorImpl)44 ArrayList (java.util.ArrayList)44 PreparedStatement (java.sql.PreparedStatement)42 GetTableLayoutResponse (com.amazonaws.athena.connector.lambda.metadata.GetTableLayoutResponse)41 Block (com.amazonaws.athena.connector.lambda.data.Block)36 SchemaBuilder (com.amazonaws.athena.connector.lambda.data.SchemaBuilder)35 RecordResponse (com.amazonaws.athena.connector.lambda.records.RecordResponse)33 ResultSet (java.sql.ResultSet)32 GetSplitsResponse (com.amazonaws.athena.connector.lambda.metadata.GetSplitsResponse)30