Search in sources :

Example 11 with TypeManager

use of io.trino.spi.type.TypeManager in project trino by trinodb.

the class IcebergMetadata method getTableProperties.

@Override
public ConnectorTableProperties getTableProperties(ConnectorSession session, ConnectorTableHandle tableHandle) {
    IcebergTableHandle table = (IcebergTableHandle) tableHandle;
    if (table.getSnapshotId().isEmpty()) {
        // TupleDomain.none() as the predicate
        return new ConnectorTableProperties(TupleDomain.none(), Optional.empty(), Optional.empty(), Optional.empty(), ImmutableList.of());
    }
    Table icebergTable = catalog.loadTable(session, table.getSchemaTableName());
    // Extract identity partition fields that are present in all partition specs, for creating the discrete predicates.
    Set<Integer> partitionSourceIds = identityPartitionColumnsInAllSpecs(icebergTable);
    TupleDomain<IcebergColumnHandle> enforcedPredicate = table.getEnforcedPredicate();
    DiscretePredicates discretePredicates = null;
    if (!partitionSourceIds.isEmpty()) {
        // Extract identity partition columns
        Map<Integer, IcebergColumnHandle> columns = getColumns(icebergTable.schema(), typeManager).stream().filter(column -> partitionSourceIds.contains(column.getId())).collect(toImmutableMap(IcebergColumnHandle::getId, Function.identity()));
        Supplier<List<FileScanTask>> lazyFiles = Suppliers.memoize(() -> {
            TableScan tableScan = icebergTable.newScan().useSnapshot(table.getSnapshotId().get()).filter(toIcebergExpression(enforcedPredicate)).includeColumnStats();
            try (CloseableIterable<FileScanTask> iterator = tableScan.planFiles()) {
                return ImmutableList.copyOf(iterator);
            } catch (IOException e) {
                throw new UncheckedIOException(e);
            }
        });
        Iterable<FileScanTask> files = () -> lazyFiles.get().iterator();
        Iterable<TupleDomain<ColumnHandle>> discreteTupleDomain = Iterables.transform(files, fileScan -> {
            // Extract partition values in the data file
            Map<Integer, Optional<String>> partitionColumnValueStrings = getPartitionKeys(fileScan);
            Map<ColumnHandle, NullableValue> partitionValues = partitionSourceIds.stream().filter(partitionColumnValueStrings::containsKey).collect(toImmutableMap(columns::get, columnId -> {
                IcebergColumnHandle column = columns.get(columnId);
                Object prestoValue = deserializePartitionValue(column.getType(), partitionColumnValueStrings.get(columnId).orElse(null), column.getName());
                return NullableValue.of(column.getType(), prestoValue);
            }));
            return TupleDomain.fromFixedValues(partitionValues);
        });
        discretePredicates = new DiscretePredicates(columns.values().stream().map(ColumnHandle.class::cast).collect(toImmutableList()), discreteTupleDomain);
    }
    return new ConnectorTableProperties(// over all tableScan.planFiles() and caching partition values in table handle.
    enforcedPredicate.transformKeys(ColumnHandle.class::cast), // TODO: implement table partitioning
    Optional.empty(), Optional.empty(), Optional.ofNullable(discretePredicates), ImmutableList.of());
}
Also used : IcebergUtil.getPartitionKeys(io.trino.plugin.iceberg.IcebergUtil.getPartitionKeys) TrinoCatalog(io.trino.plugin.iceberg.catalog.TrinoCatalog) FileSystem(org.apache.hadoop.fs.FileSystem) ConnectorTableExecuteHandle(io.trino.spi.connector.ConnectorTableExecuteHandle) HiveApplyProjectionUtil.replaceWithNewVariables(io.trino.plugin.hive.HiveApplyProjectionUtil.replaceWithNewVariables) Collections.singletonList(java.util.Collections.singletonList) NOT_SUPPORTED(io.trino.spi.StandardErrorCode.NOT_SUPPORTED) TableNotFoundException(io.trino.spi.connector.TableNotFoundException) Matcher(java.util.regex.Matcher) ConnectorOutputTableHandle(io.trino.spi.connector.ConnectorOutputTableHandle) ConnectorTableHandle(io.trino.spi.connector.ConnectorTableHandle) Map(java.util.Map) RewriteFiles(org.apache.iceberg.RewriteFiles) ProjectionApplicationResult(io.trino.spi.connector.ProjectionApplicationResult) HdfsEnvironment(io.trino.plugin.hive.HdfsEnvironment) CloseableIterable(org.apache.iceberg.io.CloseableIterable) IcebergUtil.newCreateTableTransaction(io.trino.plugin.iceberg.IcebergUtil.newCreateTableTransaction) Domain(io.trino.spi.predicate.Domain) ImmutableList.toImmutableList(com.google.common.collect.ImmutableList.toImmutableList) IcebergTableExecuteHandle(io.trino.plugin.iceberg.procedure.IcebergTableExecuteHandle) Set(java.util.Set) Schema(org.apache.iceberg.Schema) ColumnIdentity.primitiveColumnIdentity(io.trino.plugin.iceberg.ColumnIdentity.primitiveColumnIdentity) SchemaTableName(io.trino.spi.connector.SchemaTableName) PartitionSpecParser(org.apache.iceberg.PartitionSpecParser) Collectors.joining(java.util.stream.Collectors.joining) Type(org.apache.iceberg.types.Type) UncheckedIOException(java.io.UncheckedIOException) ImmutableMap.toImmutableMap(com.google.common.collect.ImmutableMap.toImmutableMap) TrinoPrincipal(io.trino.spi.security.TrinoPrincipal) CatalogSchemaTableName(io.trino.spi.connector.CatalogSchemaTableName) SchemaTablePrefix(io.trino.spi.connector.SchemaTablePrefix) RemoteIterator(org.apache.hadoop.fs.RemoteIterator) Iterables(com.google.common.collect.Iterables) ConnectorTableLayout(io.trino.spi.connector.ConnectorTableLayout) ConnectorInsertTableHandle(io.trino.spi.connector.ConnectorInsertTableHandle) IcebergUtil.deserializePartitionValue(io.trino.plugin.iceberg.IcebergUtil.deserializePartitionValue) Slice(io.airlift.slice.Slice) NullableValue(io.trino.spi.predicate.NullableValue) ColumnMetadata(io.trino.spi.connector.ColumnMetadata) ConnectorTableMetadata(io.trino.spi.connector.ConnectorTableMetadata) Variable(io.trino.spi.expression.Variable) Supplier(java.util.function.Supplier) OptionalLong(java.util.OptionalLong) MaterializedViewFreshness(io.trino.spi.connector.MaterializedViewFreshness) ColumnHandle(io.trino.spi.connector.ColumnHandle) ImmutableSet.toImmutableSet(com.google.common.collect.ImmutableSet.toImmutableSet) FILE_FORMAT_PROPERTY(io.trino.plugin.iceberg.IcebergTableProperties.FILE_FORMAT_PROPERTY) OPTIMIZE(io.trino.plugin.iceberg.procedure.IcebergTableProcedureId.OPTIMIZE) ConstraintApplicationResult(io.trino.spi.connector.ConstraintApplicationResult) DEPENDS_ON_TABLES(io.trino.plugin.iceberg.catalog.hms.TrinoHiveCatalog.DEPENDS_ON_TABLES) Table(org.apache.iceberg.Table) IOException(java.io.IOException) ConnectorSession(io.trino.spi.connector.ConnectorSession) ICEBERG_INVALID_METADATA(io.trino.plugin.iceberg.IcebergErrorCode.ICEBERG_INVALID_METADATA) ConnectorTableProperties(io.trino.spi.connector.ConnectorTableProperties) DiscretePredicates(io.trino.spi.connector.DiscretePredicates) ConnectorExpression(io.trino.spi.expression.ConnectorExpression) PartitionFields.parsePartitionFields(io.trino.plugin.iceberg.PartitionFields.parsePartitionFields) ArrayDeque(java.util.ArrayDeque) MaterializedViewNotFoundException(io.trino.spi.connector.MaterializedViewNotFoundException) TypeConverter.toTrinoType(io.trino.plugin.iceberg.TypeConverter.toTrinoType) IcebergUtil.getColumns(io.trino.plugin.iceberg.IcebergUtil.getColumns) IcebergSessionProperties.isProjectionPushdownEnabled(io.trino.plugin.iceberg.IcebergSessionProperties.isProjectionPushdownEnabled) IcebergTableProcedureId(io.trino.plugin.iceberg.procedure.IcebergTableProcedureId) IcebergSessionProperties.isStatisticsEnabled(io.trino.plugin.iceberg.IcebergSessionProperties.isStatisticsEnabled) AppendFiles(org.apache.iceberg.AppendFiles) NO_RETRIES(io.trino.spi.connector.RetryMode.NO_RETRIES) ConnectorMaterializedViewDefinition(io.trino.spi.connector.ConnectorMaterializedViewDefinition) TypeConverter.toIcebergType(io.trino.plugin.iceberg.TypeConverter.toIcebergType) PartitionField(org.apache.iceberg.PartitionField) DataFiles(org.apache.iceberg.DataFiles) CatalogSchemaName(io.trino.spi.connector.CatalogSchemaName) LOCATION_PROPERTY(io.trino.plugin.iceberg.IcebergTableProperties.LOCATION_PROPERTY) Path(org.apache.hadoop.fs.Path) DATA(io.trino.plugin.iceberg.TableType.DATA) ConnectorViewDefinition(io.trino.spi.connector.ConnectorViewDefinition) FileScanTask(org.apache.iceberg.FileScanTask) DataFile(org.apache.iceberg.DataFile) Splitter(com.google.common.base.Splitter) IcebergTableProperties.getPartitioning(io.trino.plugin.iceberg.IcebergTableProperties.getPartitioning) IcebergUtil.getFileFormat(io.trino.plugin.iceberg.IcebergUtil.getFileFormat) ImmutableSet(com.google.common.collect.ImmutableSet) ImmutableMap(com.google.common.collect.ImmutableMap) Collection(java.util.Collection) HiveWrittenPartitions(io.trino.plugin.hive.HiveWrittenPartitions) LocatedFileStatus(org.apache.hadoop.fs.LocatedFileStatus) ConcurrentHashMap(java.util.concurrent.ConcurrentHashMap) ComputedStatistics(io.trino.spi.statistics.ComputedStatistics) TrinoException(io.trino.spi.TrinoException) TableScan(org.apache.iceberg.TableScan) ConnectorOutputMetadata(io.trino.spi.connector.ConnectorOutputMetadata) String.format(java.lang.String.format) SchemaParser(org.apache.iceberg.SchemaParser) DataSize(io.airlift.units.DataSize) HdfsContext(io.trino.plugin.hive.HdfsEnvironment.HdfsContext) List(java.util.List) BIGINT(io.trino.spi.type.BigintType.BIGINT) ClassLoaderSafeSystemTable(io.trino.plugin.base.classloader.ClassLoaderSafeSystemTable) IcebergUtil.getTableComment(io.trino.plugin.iceberg.IcebergUtil.getTableComment) Assignment(io.trino.spi.connector.Assignment) HiveApplyProjectionUtil(io.trino.plugin.hive.HiveApplyProjectionUtil) BeginTableExecuteResult(io.trino.spi.connector.BeginTableExecuteResult) PartitionSpec(org.apache.iceberg.PartitionSpec) Function.identity(java.util.function.Function.identity) TableProperties(org.apache.iceberg.TableProperties) Optional(java.util.Optional) ProjectedColumnRepresentation(io.trino.plugin.hive.HiveApplyProjectionUtil.ProjectedColumnRepresentation) ConnectorMetadata(io.trino.spi.connector.ConnectorMetadata) Pattern(java.util.regex.Pattern) SystemTable(io.trino.spi.connector.SystemTable) JsonCodec(io.airlift.json.JsonCodec) Constraint(io.trino.spi.connector.Constraint) IcebergOptimizeHandle(io.trino.plugin.iceberg.procedure.IcebergOptimizeHandle) Logger(io.airlift.log.Logger) PartitionFields.toPartitionFields(io.trino.plugin.iceberg.PartitionFields.toPartitionFields) HashMap(java.util.HashMap) Deque(java.util.Deque) Function(java.util.function.Function) ExpressionConverter.toIcebergExpression(io.trino.plugin.iceberg.ExpressionConverter.toIcebergExpression) PARTITIONING_PROPERTY(io.trino.plugin.iceberg.IcebergTableProperties.PARTITIONING_PROPERTY) HashSet(java.util.HashSet) BiPredicate(java.util.function.BiPredicate) ImmutableList(com.google.common.collect.ImmutableList) Verify.verify(com.google.common.base.Verify.verify) Objects.requireNonNull(java.util.Objects.requireNonNull) Suppliers(com.google.common.base.Suppliers) TableStatistics(io.trino.spi.statistics.TableStatistics) HiveApplyProjectionUtil.extractSupportedProjectedColumns(io.trino.plugin.hive.HiveApplyProjectionUtil.extractSupportedProjectedColumns) VerifyException(com.google.common.base.VerifyException) RetryMode(io.trino.spi.connector.RetryMode) Iterator(java.util.Iterator) TupleDomain(io.trino.spi.predicate.TupleDomain) IcebergUtil.toIcebergSchema(io.trino.plugin.iceberg.IcebergUtil.toIcebergSchema) Transaction(org.apache.iceberg.Transaction) Comparator(java.util.Comparator) TypeManager(io.trino.spi.type.TypeManager) HiveUtil.isStructuralType(io.trino.plugin.hive.util.HiveUtil.isStructuralType) Snapshot(org.apache.iceberg.Snapshot) UncheckedIOException(java.io.UncheckedIOException) Collections.singletonList(java.util.Collections.singletonList) ImmutableList.toImmutableList(com.google.common.collect.ImmutableList.toImmutableList) List(java.util.List) ImmutableList(com.google.common.collect.ImmutableList) TableScan(org.apache.iceberg.TableScan) ColumnHandle(io.trino.spi.connector.ColumnHandle) Table(org.apache.iceberg.Table) ClassLoaderSafeSystemTable(io.trino.plugin.base.classloader.ClassLoaderSafeSystemTable) SystemTable(io.trino.spi.connector.SystemTable) Optional(java.util.Optional) DiscretePredicates(io.trino.spi.connector.DiscretePredicates) NullableValue(io.trino.spi.predicate.NullableValue) UncheckedIOException(java.io.UncheckedIOException) IOException(java.io.IOException) TupleDomain(io.trino.spi.predicate.TupleDomain) FileScanTask(org.apache.iceberg.FileScanTask) ConnectorTableProperties(io.trino.spi.connector.ConnectorTableProperties)

Example 12 with TypeManager

use of io.trino.spi.type.TypeManager in project trino by trinodb.

the class IcebergMetadata method getWriteLayout.

private Optional<ConnectorTableLayout> getWriteLayout(Schema tableSchema, PartitionSpec partitionSpec, boolean forceRepartitioning) {
    if (partitionSpec.isUnpartitioned()) {
        return Optional.empty();
    }
    Map<Integer, IcebergColumnHandle> columnById = getColumns(tableSchema, typeManager).stream().collect(toImmutableMap(IcebergColumnHandle::getId, identity()));
    List<IcebergColumnHandle> partitioningColumns = partitionSpec.fields().stream().sorted(Comparator.comparing(PartitionField::sourceId)).map(field -> requireNonNull(columnById.get(field.sourceId()), () -> "Cannot find source column for partitioning field " + field)).distinct().collect(toImmutableList());
    List<String> partitioningColumnNames = partitioningColumns.stream().map(IcebergColumnHandle::getName).collect(toImmutableList());
    if (!forceRepartitioning && partitionSpec.fields().stream().allMatch(field -> field.transform().isIdentity())) {
        // Do not set partitioningHandle, to let engine determine whether to repartition data or not, on stat-based basis.
        return Optional.of(new ConnectorTableLayout(partitioningColumnNames));
    }
    IcebergPartitioningHandle partitioningHandle = new IcebergPartitioningHandle(toPartitionFields(partitionSpec), partitioningColumns);
    return Optional.of(new ConnectorTableLayout(partitioningHandle, partitioningColumnNames));
}
Also used : IcebergUtil.getPartitionKeys(io.trino.plugin.iceberg.IcebergUtil.getPartitionKeys) TrinoCatalog(io.trino.plugin.iceberg.catalog.TrinoCatalog) FileSystem(org.apache.hadoop.fs.FileSystem) ConnectorTableExecuteHandle(io.trino.spi.connector.ConnectorTableExecuteHandle) HiveApplyProjectionUtil.replaceWithNewVariables(io.trino.plugin.hive.HiveApplyProjectionUtil.replaceWithNewVariables) Collections.singletonList(java.util.Collections.singletonList) NOT_SUPPORTED(io.trino.spi.StandardErrorCode.NOT_SUPPORTED) TableNotFoundException(io.trino.spi.connector.TableNotFoundException) Matcher(java.util.regex.Matcher) ConnectorOutputTableHandle(io.trino.spi.connector.ConnectorOutputTableHandle) ConnectorTableHandle(io.trino.spi.connector.ConnectorTableHandle) Map(java.util.Map) RewriteFiles(org.apache.iceberg.RewriteFiles) ProjectionApplicationResult(io.trino.spi.connector.ProjectionApplicationResult) HdfsEnvironment(io.trino.plugin.hive.HdfsEnvironment) CloseableIterable(org.apache.iceberg.io.CloseableIterable) IcebergUtil.newCreateTableTransaction(io.trino.plugin.iceberg.IcebergUtil.newCreateTableTransaction) Domain(io.trino.spi.predicate.Domain) ImmutableList.toImmutableList(com.google.common.collect.ImmutableList.toImmutableList) IcebergTableExecuteHandle(io.trino.plugin.iceberg.procedure.IcebergTableExecuteHandle) Set(java.util.Set) Schema(org.apache.iceberg.Schema) ColumnIdentity.primitiveColumnIdentity(io.trino.plugin.iceberg.ColumnIdentity.primitiveColumnIdentity) SchemaTableName(io.trino.spi.connector.SchemaTableName) PartitionSpecParser(org.apache.iceberg.PartitionSpecParser) Collectors.joining(java.util.stream.Collectors.joining) Type(org.apache.iceberg.types.Type) UncheckedIOException(java.io.UncheckedIOException) ImmutableMap.toImmutableMap(com.google.common.collect.ImmutableMap.toImmutableMap) TrinoPrincipal(io.trino.spi.security.TrinoPrincipal) CatalogSchemaTableName(io.trino.spi.connector.CatalogSchemaTableName) SchemaTablePrefix(io.trino.spi.connector.SchemaTablePrefix) RemoteIterator(org.apache.hadoop.fs.RemoteIterator) Iterables(com.google.common.collect.Iterables) ConnectorTableLayout(io.trino.spi.connector.ConnectorTableLayout) ConnectorInsertTableHandle(io.trino.spi.connector.ConnectorInsertTableHandle) IcebergUtil.deserializePartitionValue(io.trino.plugin.iceberg.IcebergUtil.deserializePartitionValue) Slice(io.airlift.slice.Slice) NullableValue(io.trino.spi.predicate.NullableValue) ColumnMetadata(io.trino.spi.connector.ColumnMetadata) ConnectorTableMetadata(io.trino.spi.connector.ConnectorTableMetadata) Variable(io.trino.spi.expression.Variable) Supplier(java.util.function.Supplier) OptionalLong(java.util.OptionalLong) MaterializedViewFreshness(io.trino.spi.connector.MaterializedViewFreshness) ColumnHandle(io.trino.spi.connector.ColumnHandle) ImmutableSet.toImmutableSet(com.google.common.collect.ImmutableSet.toImmutableSet) FILE_FORMAT_PROPERTY(io.trino.plugin.iceberg.IcebergTableProperties.FILE_FORMAT_PROPERTY) OPTIMIZE(io.trino.plugin.iceberg.procedure.IcebergTableProcedureId.OPTIMIZE) ConstraintApplicationResult(io.trino.spi.connector.ConstraintApplicationResult) DEPENDS_ON_TABLES(io.trino.plugin.iceberg.catalog.hms.TrinoHiveCatalog.DEPENDS_ON_TABLES) Table(org.apache.iceberg.Table) IOException(java.io.IOException) ConnectorSession(io.trino.spi.connector.ConnectorSession) ICEBERG_INVALID_METADATA(io.trino.plugin.iceberg.IcebergErrorCode.ICEBERG_INVALID_METADATA) ConnectorTableProperties(io.trino.spi.connector.ConnectorTableProperties) DiscretePredicates(io.trino.spi.connector.DiscretePredicates) ConnectorExpression(io.trino.spi.expression.ConnectorExpression) PartitionFields.parsePartitionFields(io.trino.plugin.iceberg.PartitionFields.parsePartitionFields) ArrayDeque(java.util.ArrayDeque) MaterializedViewNotFoundException(io.trino.spi.connector.MaterializedViewNotFoundException) TypeConverter.toTrinoType(io.trino.plugin.iceberg.TypeConverter.toTrinoType) IcebergUtil.getColumns(io.trino.plugin.iceberg.IcebergUtil.getColumns) IcebergSessionProperties.isProjectionPushdownEnabled(io.trino.plugin.iceberg.IcebergSessionProperties.isProjectionPushdownEnabled) IcebergTableProcedureId(io.trino.plugin.iceberg.procedure.IcebergTableProcedureId) IcebergSessionProperties.isStatisticsEnabled(io.trino.plugin.iceberg.IcebergSessionProperties.isStatisticsEnabled) AppendFiles(org.apache.iceberg.AppendFiles) NO_RETRIES(io.trino.spi.connector.RetryMode.NO_RETRIES) ConnectorMaterializedViewDefinition(io.trino.spi.connector.ConnectorMaterializedViewDefinition) TypeConverter.toIcebergType(io.trino.plugin.iceberg.TypeConverter.toIcebergType) PartitionField(org.apache.iceberg.PartitionField) DataFiles(org.apache.iceberg.DataFiles) CatalogSchemaName(io.trino.spi.connector.CatalogSchemaName) LOCATION_PROPERTY(io.trino.plugin.iceberg.IcebergTableProperties.LOCATION_PROPERTY) Path(org.apache.hadoop.fs.Path) DATA(io.trino.plugin.iceberg.TableType.DATA) ConnectorViewDefinition(io.trino.spi.connector.ConnectorViewDefinition) FileScanTask(org.apache.iceberg.FileScanTask) DataFile(org.apache.iceberg.DataFile) Splitter(com.google.common.base.Splitter) IcebergTableProperties.getPartitioning(io.trino.plugin.iceberg.IcebergTableProperties.getPartitioning) IcebergUtil.getFileFormat(io.trino.plugin.iceberg.IcebergUtil.getFileFormat) ImmutableSet(com.google.common.collect.ImmutableSet) ImmutableMap(com.google.common.collect.ImmutableMap) Collection(java.util.Collection) HiveWrittenPartitions(io.trino.plugin.hive.HiveWrittenPartitions) LocatedFileStatus(org.apache.hadoop.fs.LocatedFileStatus) ConcurrentHashMap(java.util.concurrent.ConcurrentHashMap) ComputedStatistics(io.trino.spi.statistics.ComputedStatistics) TrinoException(io.trino.spi.TrinoException) TableScan(org.apache.iceberg.TableScan) ConnectorOutputMetadata(io.trino.spi.connector.ConnectorOutputMetadata) String.format(java.lang.String.format) SchemaParser(org.apache.iceberg.SchemaParser) DataSize(io.airlift.units.DataSize) HdfsContext(io.trino.plugin.hive.HdfsEnvironment.HdfsContext) List(java.util.List) BIGINT(io.trino.spi.type.BigintType.BIGINT) ClassLoaderSafeSystemTable(io.trino.plugin.base.classloader.ClassLoaderSafeSystemTable) IcebergUtil.getTableComment(io.trino.plugin.iceberg.IcebergUtil.getTableComment) Assignment(io.trino.spi.connector.Assignment) HiveApplyProjectionUtil(io.trino.plugin.hive.HiveApplyProjectionUtil) BeginTableExecuteResult(io.trino.spi.connector.BeginTableExecuteResult) PartitionSpec(org.apache.iceberg.PartitionSpec) Function.identity(java.util.function.Function.identity) TableProperties(org.apache.iceberg.TableProperties) Optional(java.util.Optional) ProjectedColumnRepresentation(io.trino.plugin.hive.HiveApplyProjectionUtil.ProjectedColumnRepresentation) ConnectorMetadata(io.trino.spi.connector.ConnectorMetadata) Pattern(java.util.regex.Pattern) SystemTable(io.trino.spi.connector.SystemTable) JsonCodec(io.airlift.json.JsonCodec) Constraint(io.trino.spi.connector.Constraint) IcebergOptimizeHandle(io.trino.plugin.iceberg.procedure.IcebergOptimizeHandle) Logger(io.airlift.log.Logger) PartitionFields.toPartitionFields(io.trino.plugin.iceberg.PartitionFields.toPartitionFields) HashMap(java.util.HashMap) Deque(java.util.Deque) Function(java.util.function.Function) ExpressionConverter.toIcebergExpression(io.trino.plugin.iceberg.ExpressionConverter.toIcebergExpression) PARTITIONING_PROPERTY(io.trino.plugin.iceberg.IcebergTableProperties.PARTITIONING_PROPERTY) HashSet(java.util.HashSet) BiPredicate(java.util.function.BiPredicate) ImmutableList(com.google.common.collect.ImmutableList) Verify.verify(com.google.common.base.Verify.verify) Objects.requireNonNull(java.util.Objects.requireNonNull) Suppliers(com.google.common.base.Suppliers) TableStatistics(io.trino.spi.statistics.TableStatistics) HiveApplyProjectionUtil.extractSupportedProjectedColumns(io.trino.plugin.hive.HiveApplyProjectionUtil.extractSupportedProjectedColumns) VerifyException(com.google.common.base.VerifyException) RetryMode(io.trino.spi.connector.RetryMode) Iterator(java.util.Iterator) TupleDomain(io.trino.spi.predicate.TupleDomain) IcebergUtil.toIcebergSchema(io.trino.plugin.iceberg.IcebergUtil.toIcebergSchema) Transaction(org.apache.iceberg.Transaction) Comparator(java.util.Comparator) TypeManager(io.trino.spi.type.TypeManager) HiveUtil.isStructuralType(io.trino.plugin.hive.util.HiveUtil.isStructuralType) Snapshot(org.apache.iceberg.Snapshot) PartitionField(org.apache.iceberg.PartitionField) ConnectorTableLayout(io.trino.spi.connector.ConnectorTableLayout)

Example 13 with TypeManager

use of io.trino.spi.type.TypeManager in project trino by trinodb.

the class IcebergFileWriterFactory method createOrcWriter.

private IcebergFileWriter createOrcWriter(MetricsConfig metricsConfig, Path outputPath, Schema icebergSchema, JobConf jobConf, ConnectorSession session) {
    try {
        FileSystem fileSystem = hdfsEnvironment.getFileSystem(session.getIdentity(), outputPath, jobConf);
        OrcDataSink orcDataSink = hdfsEnvironment.doAs(session.getIdentity(), () -> new OutputStreamOrcDataSink(fileSystem.create(outputPath)));
        Callable<Void> rollbackAction = () -> {
            hdfsEnvironment.doAs(session.getIdentity(), () -> fileSystem.delete(outputPath, false));
            return null;
        };
        List<Types.NestedField> columnFields = icebergSchema.columns();
        List<String> fileColumnNames = columnFields.stream().map(Types.NestedField::name).collect(toImmutableList());
        List<Type> fileColumnTypes = columnFields.stream().map(Types.NestedField::type).map(type -> toTrinoType(type, typeManager)).collect(toImmutableList());
        Optional<Supplier<OrcDataSource>> validationInputFactory = Optional.empty();
        if (isOrcWriterValidate(session)) {
            validationInputFactory = Optional.of(() -> {
                try {
                    return new HdfsOrcDataSource(new OrcDataSourceId(outputPath.toString()), hdfsEnvironment.doAs(session.getIdentity(), () -> fileSystem.getFileStatus(outputPath).getLen()), new OrcReaderOptions(), hdfsEnvironment.doAs(session.getIdentity(), () -> fileSystem.open(outputPath)), readStats);
                } catch (IOException e) {
                    throw new TrinoException(ICEBERG_WRITE_VALIDATION_FAILED, e);
                }
            });
        }
        return new IcebergOrcFileWriter(metricsConfig, icebergSchema, orcDataSink, rollbackAction, fileColumnNames, fileColumnTypes, toOrcType(icebergSchema), getCompressionCodec(session).getOrcCompressionKind(), orcWriterOptions.withStripeMinSize(getOrcWriterMinStripeSize(session)).withStripeMaxSize(getOrcWriterMaxStripeSize(session)).withStripeMaxRowCount(getOrcWriterMaxStripeRows(session)).withDictionaryMaxMemory(getOrcWriterMaxDictionaryMemory(session)).withMaxStringStatisticsLimit(getOrcStringStatisticsLimit(session)), IntStream.range(0, fileColumnNames.size()).toArray(), ImmutableMap.<String, String>builder().put(PRESTO_VERSION_NAME, nodeVersion.toString()).put(PRESTO_QUERY_ID_NAME, session.getQueryId()).buildOrThrow(), validationInputFactory, getOrcWriterValidateMode(session), orcWriterStats);
    } catch (IOException e) {
        throw new TrinoException(ICEBERG_WRITER_OPEN_ERROR, "Error creating ORC file", e);
    }
}
Also used : OutputStreamOrcDataSink(io.trino.orc.OutputStreamOrcDataSink) Types(org.apache.iceberg.types.Types) TypeConverter.toTrinoType(io.trino.plugin.iceberg.TypeConverter.toTrinoType) FileSystem(org.apache.hadoop.fs.FileSystem) IcebergSessionProperties.getOrcWriterMaxStripeSize(io.trino.plugin.iceberg.IcebergSessionProperties.getOrcWriterMaxStripeSize) IcebergSessionProperties.getParquetWriterPageSize(io.trino.plugin.iceberg.IcebergSessionProperties.getParquetWriterPageSize) OrcDataSink(io.trino.orc.OrcDataSink) OrcWriterStats(io.trino.orc.OrcWriterStats) NOT_SUPPORTED(io.trino.spi.StandardErrorCode.NOT_SUPPORTED) Preconditions.checkArgument(com.google.common.base.Preconditions.checkArgument) IcebergSessionProperties.isOrcWriterValidate(io.trino.plugin.iceberg.IcebergSessionProperties.isOrcWriterValidate) PrimitiveTypeMapBuilder.makeTypeMap(io.trino.plugin.iceberg.util.PrimitiveTypeMapBuilder.makeTypeMap) Path(org.apache.hadoop.fs.Path) PRESTO_QUERY_ID_NAME(io.trino.plugin.hive.HiveMetadata.PRESTO_QUERY_ID_NAME) OrcDataSource(io.trino.orc.OrcDataSource) IcebergSessionProperties.getParquetWriterBatchSize(io.trino.plugin.iceberg.IcebergSessionProperties.getParquetWriterBatchSize) ParquetSchemaUtil.convert(org.apache.iceberg.parquet.ParquetSchemaUtil.convert) FileFormatDataSourceStats(io.trino.plugin.hive.FileFormatDataSourceStats) HdfsEnvironment(io.trino.plugin.hive.HdfsEnvironment) ImmutableMap(com.google.common.collect.ImmutableMap) ImmutableList.toImmutableList(com.google.common.collect.ImmutableList.toImmutableList) TrinoException(io.trino.spi.TrinoException) Schema(org.apache.iceberg.Schema) IcebergSessionProperties.getCompressionCodec(io.trino.plugin.iceberg.IcebergSessionProperties.getCompressionCodec) IcebergSessionProperties.getOrcWriterMaxStripeRows(io.trino.plugin.iceberg.IcebergSessionProperties.getOrcWriterMaxStripeRows) OrcDataSourceId(io.trino.orc.OrcDataSourceId) HdfsContext(io.trino.plugin.hive.HdfsEnvironment.HdfsContext) List(java.util.List) ICEBERG_WRITER_OPEN_ERROR(io.trino.plugin.iceberg.IcebergErrorCode.ICEBERG_WRITER_OPEN_ERROR) Optional(java.util.Optional) IntStream(java.util.stream.IntStream) Type(io.trino.spi.type.Type) Callable(java.util.concurrent.Callable) ICEBERG_WRITE_VALIDATION_FAILED(io.trino.plugin.iceberg.IcebergErrorCode.ICEBERG_WRITE_VALIDATION_FAILED) Supplier(java.util.function.Supplier) Inject(javax.inject.Inject) Managed(org.weakref.jmx.Managed) IcebergSessionProperties.getOrcWriterMaxDictionaryMemory(io.trino.plugin.iceberg.IcebergSessionProperties.getOrcWriterMaxDictionaryMemory) NodeVersion(io.trino.plugin.hive.NodeVersion) OrcReaderOptions(io.trino.orc.OrcReaderOptions) Objects.requireNonNull(java.util.Objects.requireNonNull) TypeConverter.toOrcType(io.trino.plugin.iceberg.TypeConverter.toOrcType) ParquetWriterOptions(io.trino.parquet.writer.ParquetWriterOptions) IcebergSessionProperties.getOrcWriterMinStripeSize(io.trino.plugin.iceberg.IcebergSessionProperties.getOrcWriterMinStripeSize) OrcWriterConfig(io.trino.plugin.hive.orc.OrcWriterConfig) HdfsOrcDataSource(io.trino.plugin.hive.orc.HdfsOrcDataSource) IcebergSessionProperties.getOrcWriterValidateMode(io.trino.plugin.iceberg.IcebergSessionProperties.getOrcWriterValidateMode) MetricsConfig(org.apache.iceberg.MetricsConfig) IcebergSessionProperties.getParquetWriterBlockSize(io.trino.plugin.iceberg.IcebergSessionProperties.getParquetWriterBlockSize) IOException(java.io.IOException) ConnectorSession(io.trino.spi.connector.ConnectorSession) IcebergSessionProperties.getOrcStringStatisticsLimit(io.trino.plugin.iceberg.IcebergSessionProperties.getOrcStringStatisticsLimit) JobConf(org.apache.hadoop.mapred.JobConf) OrcWriterOptions(io.trino.orc.OrcWriterOptions) PRESTO_VERSION_NAME(io.trino.plugin.hive.HiveMetadata.PRESTO_VERSION_NAME) TypeManager(io.trino.spi.type.TypeManager) Types(org.apache.iceberg.types.Types) OrcDataSourceId(io.trino.orc.OrcDataSourceId) OutputStreamOrcDataSink(io.trino.orc.OutputStreamOrcDataSink) OrcDataSink(io.trino.orc.OrcDataSink) HdfsOrcDataSource(io.trino.plugin.hive.orc.HdfsOrcDataSource) IOException(java.io.IOException) TypeConverter.toTrinoType(io.trino.plugin.iceberg.TypeConverter.toTrinoType) Type(io.trino.spi.type.Type) TypeConverter.toOrcType(io.trino.plugin.iceberg.TypeConverter.toOrcType) OrcReaderOptions(io.trino.orc.OrcReaderOptions) OutputStreamOrcDataSink(io.trino.orc.OutputStreamOrcDataSink) FileSystem(org.apache.hadoop.fs.FileSystem) TrinoException(io.trino.spi.TrinoException) Supplier(java.util.function.Supplier)

Example 14 with TypeManager

use of io.trino.spi.type.TypeManager in project trino by trinodb.

the class DeltaLakeSplitManager method getSplits.

private Stream<DeltaLakeSplit> getSplits(ConnectorTransactionHandle transaction, DeltaLakeTableHandle tableHandle, ConnectorSession session, Optional<DataSize> maxScannedFileSize, Set<ColumnHandle> columnsCoveredByDynamicFilter, Constraint constraint) {
    DeltaLakeMetastore metastore = getMetastore(session, transaction);
    String tableLocation = metastore.getTableLocation(tableHandle.getSchemaTableName(), session);
    List<AddFileEntry> validDataFiles = metastore.getValidDataFiles(tableHandle.getSchemaTableName(), session);
    TupleDomain<DeltaLakeColumnHandle> enforcedPartitionConstraint = tableHandle.getEnforcedPartitionConstraint();
    TupleDomain<DeltaLakeColumnHandle> nonPartitionConstraint = tableHandle.getNonPartitionConstraint();
    // Delta Lake handles updates and deletes by copying entire data files, minus updates/deletes. Because of this we can only have one Split/UpdatablePageSource
    // per file.
    boolean splittable = tableHandle.getWriteType().isEmpty();
    AtomicInteger remainingInitialSplits = new AtomicInteger(maxInitialSplits);
    Optional<Instant> filesModifiedAfter = tableHandle.getAnalyzeHandle().flatMap(AnalyzeHandle::getFilesModifiedAfter);
    Optional<Long> maxScannedFileSizeInBytes = maxScannedFileSize.map(DataSize::toBytes);
    Set<String> predicatedColumnNames = Stream.concat(nonPartitionConstraint.getDomains().orElseThrow().keySet().stream(), columnsCoveredByDynamicFilter.stream().map(DeltaLakeColumnHandle.class::cast)).map(// TODO is DeltaLakeColumnHandle.name normalized?
    column -> column.getName().toLowerCase(ENGLISH)).collect(toImmutableSet());
    List<ColumnMetadata> schema = extractSchema(tableHandle.getMetadataEntry(), typeManager);
    List<ColumnMetadata> predicatedColumns = schema.stream().filter(// ColumnMetadata.name is lowercase
    column -> predicatedColumnNames.contains(column.getName())).collect(toImmutableList());
    return validDataFiles.stream().flatMap(addAction -> {
        if (tableHandle.getAnalyzeHandle().isPresent() && !tableHandle.getAnalyzeHandle().get().isInitialAnalyze() && !addAction.isDataChange()) {
            // skip files which do not introduce data change on non-initial ANALYZE
            return Stream.empty();
        }
        if (filesModifiedAfter.isPresent() && addAction.getModificationTime() <= filesModifiedAfter.get().toEpochMilli()) {
            return Stream.empty();
        }
        if (maxScannedFileSizeInBytes.isPresent() && addAction.getSize() > maxScannedFileSizeInBytes.get()) {
            return Stream.empty();
        }
        Map<DeltaLakeColumnHandle, Domain> enforcedDomains = enforcedPartitionConstraint.getDomains().orElseThrow();
        if (!partitionMatchesPredicate(addAction.getCanonicalPartitionValues(), enforcedDomains)) {
            return Stream.empty();
        }
        TupleDomain<DeltaLakeColumnHandle> statisticsPredicate = createStatisticsPredicate(addAction, predicatedColumns, tableHandle.getMetadataEntry().getCanonicalPartitionColumns());
        if (!nonPartitionConstraint.overlaps(statisticsPredicate)) {
            return Stream.empty();
        }
        if (constraint.predicate().isPresent()) {
            Map<String, Optional<String>> partitionValues = addAction.getCanonicalPartitionValues();
            Map<ColumnHandle, NullableValue> deserializedValues = constraint.getPredicateColumns().orElseThrow().stream().filter(column -> column instanceof DeltaLakeColumnHandle).filter(column -> partitionValues.containsKey(((DeltaLakeColumnHandle) column).getName())).collect(toImmutableMap(identity(), column -> {
                DeltaLakeColumnHandle deltaLakeColumn = (DeltaLakeColumnHandle) column;
                return NullableValue.of(deltaLakeColumn.getType(), deserializePartitionValue(deltaLakeColumn, addAction.getCanonicalPartitionValues().get(deltaLakeColumn.getName())));
            }));
            if (!constraint.predicate().get().test(deserializedValues)) {
                return Stream.empty();
            }
        }
        return splitsForFile(session, addAction, tableLocation, addAction.getCanonicalPartitionValues(), statisticsPredicate, splittable, remainingInitialSplits).stream();
    });
}
Also used : ConnectorSplitManager(io.trino.spi.connector.ConnectorSplitManager) Constraint(io.trino.spi.connector.Constraint) DeltaLakeSessionProperties.getMaxInitialSplitSize(io.trino.plugin.deltalake.DeltaLakeSessionProperties.getMaxInitialSplitSize) URLDecoder(java.net.URLDecoder) NullableValue(io.trino.spi.predicate.NullableValue) ColumnMetadata(io.trino.spi.connector.ColumnMetadata) BiFunction(java.util.function.BiFunction) DeltaLakeSchemaSupport.extractSchema(io.trino.plugin.deltalake.transactionlog.DeltaLakeSchemaSupport.extractSchema) AddFileEntry(io.trino.plugin.deltalake.transactionlog.AddFileEntry) FixedSplitSource(io.trino.spi.connector.FixedSplitSource) Inject(javax.inject.Inject) DeltaLakeMetastore(io.trino.plugin.deltalake.metastore.DeltaLakeMetastore) DeltaLakeMetadata.createStatisticsPredicate(io.trino.plugin.deltalake.DeltaLakeMetadata.createStatisticsPredicate) ImmutableList(com.google.common.collect.ImmutableList) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) ConnectorTableHandle(io.trino.spi.connector.ConnectorTableHandle) Map(java.util.Map) Objects.requireNonNull(java.util.Objects.requireNonNull) ColumnHandle(io.trino.spi.connector.ColumnHandle) ImmutableSet.toImmutableSet(com.google.common.collect.ImmutableSet.toImmutableSet) ENGLISH(java.util.Locale.ENGLISH) ExecutorService(java.util.concurrent.ExecutorService) UTF_8(java.nio.charset.StandardCharsets.UTF_8) Domain(io.trino.spi.predicate.Domain) ImmutableList.toImmutableList(com.google.common.collect.ImmutableList.toImmutableList) Set(java.util.Set) ConnectorSplitSource(io.trino.spi.connector.ConnectorSplitSource) ConnectorSession(io.trino.spi.connector.ConnectorSession) TupleDomain(io.trino.spi.predicate.TupleDomain) Instant(java.time.Instant) DeltaLakeSessionProperties.getMaxSplitSize(io.trino.plugin.deltalake.DeltaLakeSessionProperties.getMaxSplitSize) DataSize(io.airlift.units.DataSize) List(java.util.List) ImmutableMap.toImmutableMap(com.google.common.collect.ImmutableMap.toImmutableMap) ClassLoaderSafeConnectorSplitSource(io.trino.plugin.base.classloader.ClassLoaderSafeConnectorSplitSource) Stream(java.util.stream.Stream) DynamicFilter(io.trino.spi.connector.DynamicFilter) Function.identity(java.util.function.Function.identity) Optional(java.util.Optional) TypeManager(io.trino.spi.type.TypeManager) TransactionLogParser.deserializePartitionValue(io.trino.plugin.deltalake.transactionlog.TransactionLogParser.deserializePartitionValue) HiveTransactionHandle(io.trino.plugin.hive.HiveTransactionHandle) ConnectorTransactionHandle(io.trino.spi.connector.ConnectorTransactionHandle) ColumnMetadata(io.trino.spi.connector.ColumnMetadata) DataSize(io.airlift.units.DataSize) ColumnHandle(io.trino.spi.connector.ColumnHandle) Optional(java.util.Optional) Instant(java.time.Instant) NullableValue(io.trino.spi.predicate.NullableValue) DeltaLakeMetastore(io.trino.plugin.deltalake.metastore.DeltaLakeMetastore) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) AddFileEntry(io.trino.plugin.deltalake.transactionlog.AddFileEntry) Domain(io.trino.spi.predicate.Domain) TupleDomain(io.trino.spi.predicate.TupleDomain)

Example 15 with TypeManager

use of io.trino.spi.type.TypeManager in project trino by trinodb.

the class OrcFileWriterFactory method createFileWriter.

@Override
public Optional<FileWriter> createFileWriter(Path path, List<String> inputColumnNames, StorageFormat storageFormat, Properties schema, JobConf configuration, ConnectorSession session, OptionalInt bucketNumber, AcidTransaction transaction, boolean useAcidSchema, WriterKind writerKind) {
    if (!OrcOutputFormat.class.getName().equals(storageFormat.getOutputFormat())) {
        return Optional.empty();
    }
    CompressionKind compression = getCompression(schema, configuration);
    // existing tables and partitions may have columns in a different order than the writer is providing, so build
    // an index to rearrange columns in the proper order
    List<String> fileColumnNames = getColumnNames(schema);
    List<Type> fileColumnTypes = getColumnTypes(schema).stream().map(hiveType -> hiveType.getType(typeManager, getTimestampPrecision(session))).collect(toList());
    int[] fileInputColumnIndexes = fileColumnNames.stream().mapToInt(inputColumnNames::indexOf).toArray();
    if (transaction.isAcidDeleteOperation(writerKind)) {
        // For delete, set the "row" column to -1
        fileInputColumnIndexes[fileInputColumnIndexes.length - 1] = -1;
    }
    try {
        FileSystem fileSystem = hdfsEnvironment.getFileSystem(session.getIdentity(), path, configuration);
        OrcDataSink orcDataSink = createOrcDataSink(fileSystem, path);
        Optional<Supplier<OrcDataSource>> validationInputFactory = Optional.empty();
        if (isOrcOptimizedWriterValidate(session)) {
            validationInputFactory = Optional.of(() -> {
                try {
                    return new HdfsOrcDataSource(new OrcDataSourceId(path.toString()), fileSystem.getFileStatus(path).getLen(), new OrcReaderOptions(), fileSystem.open(path), readStats);
                } catch (IOException e) {
                    throw new TrinoException(HIVE_WRITE_VALIDATION_FAILED, e);
                }
            });
        }
        Callable<Void> rollbackAction = () -> {
            fileSystem.delete(path, false);
            return null;
        };
        if (transaction.isInsert() && useAcidSchema) {
            // Only add the ACID columns if the request is for insert-type operations - - for delete operations,
            // the columns are added by the caller.  This is because the ACID columns for delete operations
            // depend on the rows being deleted, whereas the ACID columns for INSERT are completely determined
            // by bucket and writeId.
            Type rowType = createRowType(fileColumnNames, fileColumnTypes);
            fileColumnNames = ACID_COLUMN_NAMES;
            fileColumnTypes = createAcidColumnPrestoTypes(rowType);
        }
        return Optional.of(new OrcFileWriter(orcDataSink, writerKind, transaction, useAcidSchema, bucketNumber, rollbackAction, fileColumnNames, fileColumnTypes, createRootOrcType(fileColumnNames, fileColumnTypes), compression, getOrcWriterOptions(schema, orcWriterOptions).withStripeMinSize(getOrcOptimizedWriterMinStripeSize(session)).withStripeMaxSize(getOrcOptimizedWriterMaxStripeSize(session)).withStripeMaxRowCount(getOrcOptimizedWriterMaxStripeRows(session)).withDictionaryMaxMemory(getOrcOptimizedWriterMaxDictionaryMemory(session)).withMaxStringStatisticsLimit(getOrcStringStatisticsLimit(session)), fileInputColumnIndexes, ImmutableMap.<String, String>builder().put(PRESTO_VERSION_NAME, nodeVersion.toString()).put(PRESTO_QUERY_ID_NAME, session.getQueryId()).buildOrThrow(), validationInputFactory, getOrcOptimizedWriterValidateMode(session), stats));
    } catch (IOException e) {
        throw new TrinoException(HIVE_WRITER_OPEN_ERROR, "Error creating ORC file", e);
    }
}
Also used : OutputStreamOrcDataSink(io.trino.orc.OutputStreamOrcDataSink) FileSystem(org.apache.hadoop.fs.FileSystem) HiveSessionProperties.getOrcOptimizedWriterMaxDictionaryMemory(io.trino.plugin.hive.HiveSessionProperties.getOrcOptimizedWriterMaxDictionaryMemory) OrcDataSink(io.trino.orc.OrcDataSink) OrcWriterStats(io.trino.orc.OrcWriterStats) HiveUtil.getColumnNames(io.trino.plugin.hive.util.HiveUtil.getColumnNames) OrcConf(org.apache.orc.OrcConf) Path(org.apache.hadoop.fs.Path) PRESTO_QUERY_ID_NAME(io.trino.plugin.hive.HiveMetadata.PRESTO_QUERY_ID_NAME) OrcDataSource(io.trino.orc.OrcDataSource) ENGLISH(java.util.Locale.ENGLISH) StorageFormat(io.trino.plugin.hive.metastore.StorageFormat) AcidTransaction(io.trino.plugin.hive.acid.AcidTransaction) FileFormatDataSourceStats(io.trino.plugin.hive.FileFormatDataSourceStats) HdfsEnvironment(io.trino.plugin.hive.HdfsEnvironment) ImmutableMap(com.google.common.collect.ImmutableMap) AcidSchema.createRowType(io.trino.plugin.hive.acid.AcidSchema.createRowType) HiveSessionProperties.getOrcOptimizedWriterMaxStripeSize(io.trino.plugin.hive.HiveSessionProperties.getOrcOptimizedWriterMaxStripeSize) TrinoException(io.trino.spi.TrinoException) HiveSessionProperties.getOrcOptimizedWriterValidateMode(io.trino.plugin.hive.HiveSessionProperties.getOrcOptimizedWriterValidateMode) WriterKind(io.trino.plugin.hive.WriterKind) HiveSessionProperties.getOrcOptimizedWriterMinStripeSize(io.trino.plugin.hive.HiveSessionProperties.getOrcOptimizedWriterMinStripeSize) OrcDataSourceId(io.trino.orc.OrcDataSourceId) List(java.util.List) AcidSchema.createAcidColumnPrestoTypes(io.trino.plugin.hive.acid.AcidSchema.createAcidColumnPrestoTypes) HIVE_WRITE_VALIDATION_FAILED(io.trino.plugin.hive.HiveErrorCode.HIVE_WRITE_VALIDATION_FAILED) Optional(java.util.Optional) ACID_COLUMN_NAMES(io.trino.plugin.hive.acid.AcidSchema.ACID_COLUMN_NAMES) OrcOutputFormat(org.apache.hadoop.hive.ql.io.orc.OrcOutputFormat) HiveFileWriterFactory(io.trino.plugin.hive.HiveFileWriterFactory) Flatten(org.weakref.jmx.Flatten) Type(io.trino.spi.type.Type) Callable(java.util.concurrent.Callable) OptionalInt(java.util.OptionalInt) Supplier(java.util.function.Supplier) FileWriter(io.trino.plugin.hive.FileWriter) Inject(javax.inject.Inject) Managed(org.weakref.jmx.Managed) NodeVersion(io.trino.plugin.hive.NodeVersion) OrcReaderOptions(io.trino.orc.OrcReaderOptions) Objects.requireNonNull(java.util.Objects.requireNonNull) HiveSessionProperties.isOrcOptimizedWriterValidate(io.trino.plugin.hive.HiveSessionProperties.isOrcOptimizedWriterValidate) HIVE_WRITER_OPEN_ERROR(io.trino.plugin.hive.HiveErrorCode.HIVE_WRITER_OPEN_ERROR) HiveSessionProperties.getOrcStringStatisticsLimit(io.trino.plugin.hive.HiveSessionProperties.getOrcStringStatisticsLimit) HiveUtil.getColumnTypes(io.trino.plugin.hive.util.HiveUtil.getColumnTypes) Properties(java.util.Properties) HIVE_UNSUPPORTED_FORMAT(io.trino.plugin.hive.HiveErrorCode.HIVE_UNSUPPORTED_FORMAT) HiveSessionProperties.getTimestampPrecision(io.trino.plugin.hive.HiveSessionProperties.getTimestampPrecision) IOException(java.io.IOException) ConnectorSession(io.trino.spi.connector.ConnectorSession) HiveUtil.getOrcWriterOptions(io.trino.plugin.hive.util.HiveUtil.getOrcWriterOptions) CompressionKind(io.trino.orc.metadata.CompressionKind) JobConf(org.apache.hadoop.mapred.JobConf) Collectors.toList(java.util.stream.Collectors.toList) OrcType.createRootOrcType(io.trino.orc.metadata.OrcType.createRootOrcType) OrcWriterOptions(io.trino.orc.OrcWriterOptions) PRESTO_VERSION_NAME(io.trino.plugin.hive.HiveMetadata.PRESTO_VERSION_NAME) TypeManager(io.trino.spi.type.TypeManager) HiveSessionProperties.getOrcOptimizedWriterMaxStripeRows(io.trino.plugin.hive.HiveSessionProperties.getOrcOptimizedWriterMaxStripeRows) CompressionKind(io.trino.orc.metadata.CompressionKind) OrcDataSourceId(io.trino.orc.OrcDataSourceId) OutputStreamOrcDataSink(io.trino.orc.OutputStreamOrcDataSink) OrcDataSink(io.trino.orc.OrcDataSink) IOException(java.io.IOException) AcidSchema.createRowType(io.trino.plugin.hive.acid.AcidSchema.createRowType) Type(io.trino.spi.type.Type) OrcType.createRootOrcType(io.trino.orc.metadata.OrcType.createRootOrcType) OrcReaderOptions(io.trino.orc.OrcReaderOptions) FileSystem(org.apache.hadoop.fs.FileSystem) TrinoException(io.trino.spi.TrinoException) Supplier(java.util.function.Supplier)

Aggregations

TypeManager (io.trino.spi.type.TypeManager)35 List (java.util.List)28 Objects.requireNonNull (java.util.Objects.requireNonNull)28 Optional (java.util.Optional)27 ConnectorSession (io.trino.spi.connector.ConnectorSession)25 TrinoException (io.trino.spi.TrinoException)23 IOException (java.io.IOException)22 ImmutableList.toImmutableList (com.google.common.collect.ImmutableList.toImmutableList)21 Path (org.apache.hadoop.fs.Path)21 Map (java.util.Map)20 TupleDomain (io.trino.spi.predicate.TupleDomain)19 ImmutableList (com.google.common.collect.ImmutableList)18 ImmutableMap (com.google.common.collect.ImmutableMap)18 Set (java.util.Set)18 FileSystem (org.apache.hadoop.fs.FileSystem)17 ImmutableMap.toImmutableMap (com.google.common.collect.ImmutableMap.toImmutableMap)15 ImmutableSet (com.google.common.collect.ImmutableSet)14 ImmutableSet.toImmutableSet (com.google.common.collect.ImmutableSet.toImmutableSet)14 HdfsContext (io.trino.plugin.hive.HdfsEnvironment.HdfsContext)14 NOT_SUPPORTED (io.trino.spi.StandardErrorCode.NOT_SUPPORTED)14