Examples with OrcWriterOptions io.trino.orc.OrcWriterOptions used on opensource projects

Example 6 with OrcWriterOptions

use of io.trino.orc.OrcWriterOptions 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);
    }
}

Example 7 with OrcWriterOptions

use of io.trino.orc.OrcWriterOptions in project trino by trinodb.

the class TestOrcWriterOptions method testOrcWriterOptionsFromTableProperties.

@Test
public void testOrcWriterOptionsFromTableProperties() {
    Properties tableProperties = new Properties();
    tableProperties.setProperty(ORC_BLOOM_FILTER_COLUMNS_KEY, "column_a, column_b");
    tableProperties.setProperty(ORC_BLOOM_FILTER_FPP_KEY, "0.5");
    OrcWriterOptions orcWriterOptions = getOrcWriterOptions(tableProperties, new OrcWriterOptions());
    assertThat(orcWriterOptions.getBloomFilterFpp()).isEqualTo(0.5);
    assertThat(orcWriterOptions.isBloomFilterColumn("column_a")).isTrue();
    assertThat(orcWriterOptions.isBloomFilterColumn("column_b")).isTrue();
    assertThat(orcWriterOptions.isBloomFilterColumn("unknown_column")).isFalse();
}

Example 8 with OrcWriterOptions

use of io.trino.orc.OrcWriterOptions in project trino by trinodb.

the class TestOrcWriterOptions method testOrcWriterOptionsWithInvalidFPPValue.

@Test
public void testOrcWriterOptionsWithInvalidFPPValue() {
    Properties tableProperties = new Properties();
    tableProperties.setProperty(ORC_BLOOM_FILTER_COLUMNS_KEY, "column_with_bloom_filter");
    tableProperties.setProperty(ORC_BLOOM_FILTER_FPP_KEY, "abc");
    assertThatThrownBy(() -> getOrcWriterOptions(tableProperties, new OrcWriterOptions())).hasMessage("Invalid value for orc_bloom_filter_fpp property: abc");
}