Examples with PARQUET com.facebook.presto.hive.HiveStorageFormat.PARQUET used on opensource projects

Example 1 with PARQUET

use of com.facebook.presto.hive.HiveStorageFormat.PARQUET in project presto by prestodb.

the class HiveSplitManager method getPartitionMetadata.

private Iterable<HivePartitionMetadata> getPartitionMetadata(SemiTransactionalHiveMetastore metastore, Table table, SchemaTableName tableName, List<HivePartition> hivePartitions, Optional<HiveBucketHandle> hiveBucketHandle, ConnectorSession session, WarningCollector warningCollector, Optional<Set<HiveColumnHandle>> requestedColumns, Map<String, HiveColumnHandle> predicateColumns, Optional<Map<Subfield, Domain>> domains) {
    if (hivePartitions.isEmpty()) {
        return ImmutableList.of();
    }
    Optional<Set<HiveColumnHandle>> allRequestedColumns = mergeRequestedAndPredicateColumns(requestedColumns, ImmutableSet.copyOf(predicateColumns.values()));
    if (hivePartitions.size() == 1) {
        HivePartition firstPartition = getOnlyElement(hivePartitions);
        if (firstPartition.getPartitionId().equals(UNPARTITIONED_ID)) {
            return ImmutableList.of(new HivePartitionMetadata(firstPartition, Optional.empty(), TableToPartitionMapping.empty(), encryptionInformationProvider.getReadEncryptionInformation(session, table, allRequestedColumns), ImmutableSet.of()));
        }
    }
    StorageFormat storageFormat = table.getStorage().getStorageFormat();
    Optional<HiveStorageFormat> hiveStorageFormat = getHiveStorageFormat(storageFormat);
    Optional<HiveStorageFormat> resolvedHiveStorageFormat;
    if (isUseParquetColumnNames(session)) {
        // Use Hive Storage Format as Parquet if table is of HUDI format
        resolvedHiveStorageFormat = (!hiveStorageFormat.isPresent() && isHudiFormat(storageFormat)) ? Optional.of(PARQUET) : hiveStorageFormat;
    } else {
        resolvedHiveStorageFormat = hiveStorageFormat;
    }
    Iterable<List<HivePartition>> partitionNameBatches = partitionExponentially(hivePartitions, minPartitionBatchSize, maxPartitionBatchSize);
    Iterable<List<HivePartitionMetadata>> partitionBatches = transform(partitionNameBatches, partitionBatch -> {
        Map<String, PartitionSplitInfo> partitionSplitInfo = getPartitionSplitInfo(session, metastore, tableName, partitionBatch, predicateColumns, domains);
        if (partitionBatch.size() != partitionSplitInfo.size()) {
            throw new PrestoException(GENERIC_INTERNAL_ERROR, format("Expected %s partitions but found %s", partitionBatch.size(), partitionSplitInfo.size()));
        }
        Map<String, Partition> partitions = partitionSplitInfo.entrySet().stream().collect(toImmutableMap(Entry::getKey, entry -> entry.getValue().getPartition()));
        Optional<Map<String, EncryptionInformation>> encryptionInformationForPartitions = encryptionInformationProvider.getReadEncryptionInformation(session, table, allRequestedColumns, partitions);
        ImmutableList.Builder<HivePartitionMetadata> results = ImmutableList.builder();
        Map<String, Set<String>> partitionsNotReadable = new HashMap<>();
        int unreadablePartitionsSkipped = 0;
        for (HivePartition hivePartition : partitionBatch) {
            Partition partition = partitions.get(hivePartition.getPartitionId());
            if (partitionSplitInfo.get(hivePartition.getPartitionId()).isPruned()) {
                continue;
            }
            if (partition == null) {
                throw new PrestoException(GENERIC_INTERNAL_ERROR, "Partition not loaded: " + hivePartition);
            }
            String partitionName = makePartName(table.getPartitionColumns(), partition.getValues());
            Optional<EncryptionInformation> encryptionInformation = encryptionInformationForPartitions.map(metadata -> metadata.get(hivePartition.getPartitionId()));
            if (!isOfflineDataDebugModeEnabled(session)) {
                // verify partition is online
                verifyOnline(tableName, Optional.of(partitionName), getProtectMode(partition), partition.getParameters());
                // verify partition is not marked as non-readable
                String reason = partition.getParameters().get(OBJECT_NOT_READABLE);
                if (!isNullOrEmpty(reason)) {
                    if (!shouldIgnoreUnreadablePartition(session) || !partition.isEligibleToIgnore()) {
                        throw new HiveNotReadableException(tableName, Optional.of(partitionName), reason);
                    }
                    unreadablePartitionsSkipped++;
                    if (partitionsNotReadable.size() <= 3) {
                        partitionsNotReadable.putIfAbsent(reason, new HashSet<>(ImmutableSet.of(partitionName)));
                        if (partitionsNotReadable.get(reason).size() <= 3) {
                            partitionsNotReadable.get(reason).add(partitionName);
                        }
                    }
                    continue;
                }
            }
            // Verify that the partition schema matches the table schema.
            // Either adding or dropping columns from the end of the table
            // without modifying existing partitions is allowed, but every
            // column that exists in both the table and partition must have
            // the same type.
            List<Column> tableColumns = table.getDataColumns();
            List<Column> partitionColumns = partition.getColumns();
            if ((tableColumns == null) || (partitionColumns == null)) {
                throw new PrestoException(HIVE_INVALID_METADATA, format("Table '%s' or partition '%s' has null columns", tableName, partitionName));
            }
            TableToPartitionMapping tableToPartitionMapping = getTableToPartitionMapping(session, resolvedHiveStorageFormat, tableName, partitionName, tableColumns, partitionColumns);
            if (hiveBucketHandle.isPresent() && !hiveBucketHandle.get().isVirtuallyBucketed()) {
                Optional<HiveBucketProperty> partitionBucketProperty = partition.getStorage().getBucketProperty();
                if (!partitionBucketProperty.isPresent()) {
                    throw new PrestoException(HIVE_PARTITION_SCHEMA_MISMATCH, format("Hive table (%s) is bucketed but partition (%s) is not bucketed", hivePartition.getTableName(), hivePartition.getPartitionId()));
                }
                int tableBucketCount = hiveBucketHandle.get().getTableBucketCount();
                int partitionBucketCount = partitionBucketProperty.get().getBucketCount();
                List<String> tableBucketColumns = hiveBucketHandle.get().getColumns().stream().map(HiveColumnHandle::getName).collect(toImmutableList());
                List<String> partitionBucketColumns = partitionBucketProperty.get().getBucketedBy();
                if (!tableBucketColumns.equals(partitionBucketColumns) || !isBucketCountCompatible(tableBucketCount, partitionBucketCount)) {
                    throw new PrestoException(HIVE_PARTITION_SCHEMA_MISMATCH, format("Hive table (%s) bucketing (columns=%s, buckets=%s) is not compatible with partition (%s) bucketing (columns=%s, buckets=%s)", hivePartition.getTableName(), tableBucketColumns, tableBucketCount, hivePartition.getPartitionId(), partitionBucketColumns, partitionBucketCount));
                }
            }
            results.add(new HivePartitionMetadata(hivePartition, Optional.of(partition), tableToPartitionMapping, encryptionInformation, partitionSplitInfo.get(hivePartition.getPartitionId()).getRedundantColumnDomains()));
        }
        if (unreadablePartitionsSkipped > 0) {
            StringBuilder warningMessage = new StringBuilder(format("Table '%s' has %s out of %s partitions unreadable: ", tableName, unreadablePartitionsSkipped, partitionBatch.size()));
            for (Entry<String, Set<String>> entry : partitionsNotReadable.entrySet()) {
                warningMessage.append(String.join(", ", entry.getValue())).append("... are due to ").append(entry.getKey()).append(". ");
            }
            warningCollector.add(new PrestoWarning(PARTITION_NOT_READABLE, warningMessage.toString()));
        }
        return results.build();
    });
    return concat(partitionBatches);
}

Example 2 with PARQUET

use of com.facebook.presto.hive.HiveStorageFormat.PARQUET in project presto by prestodb.

the class TestHiveFileFormats method testParquetPageSourceSchemaEvolution.

@Test(dataProvider = "rowCount")
public void testParquetPageSourceSchemaEvolution(int rowCount) throws Exception {
    List<TestColumn> writeColumns = getTestColumnsSupportedByParquet();
    // test index-based access
    List<TestColumn> readColumns = writeColumns.stream().map(column -> new TestColumn(column.getName() + "_new", column.getObjectInspector(), column.getWriteValue(), column.getExpectedValue(), column.isPartitionKey())).collect(toList());
    assertThatFileFormat(PARQUET).withWriteColumns(writeColumns).withReadColumns(readColumns).withSession(parquetPageSourceSession).withRowsCount(rowCount).isReadableByPageSource(new ParquetPageSourceFactory(FUNCTION_AND_TYPE_MANAGER, FUNCTION_RESOLUTION, HDFS_ENVIRONMENT, STATS, METADATA_READER));
    // test name-based access
    readColumns = Lists.reverse(writeColumns);
    assertThatFileFormat(PARQUET).withWriteColumns(writeColumns).withReadColumns(readColumns).withSession(parquetPageSourceSessionUseName).isReadableByPageSource(new ParquetPageSourceFactory(FUNCTION_AND_TYPE_MANAGER, FUNCTION_RESOLUTION, HDFS_ENVIRONMENT, STATS, METADATA_READER));
}

Example 3 with PARQUET

use of com.facebook.presto.hive.HiveStorageFormat.PARQUET in project presto by prestodb.

the class TestHiveFileFormats method testOptimizedParquetWriter.

@Test(dataProvider = "rowCount")
public void testOptimizedParquetWriter(int rowCount) throws Exception {
    TestingConnectorSession session = new TestingConnectorSession(new HiveSessionProperties(new HiveClientConfig().setOrcOptimizedWriterEnabled(true).setOrcWriterValidationPercentage(100.0), new OrcFileWriterConfig(), new ParquetFileWriterConfig().setParquetOptimizedWriterEnabled(true), new CacheConfig()).getSessionProperties());
    // A Presto page can not contain a map with null keys, so a page based writer can not write null keys
    List<TestColumn> testColumns = TEST_COLUMNS.stream().filter(testColumn -> !testColumn.getName().equals("t_map_null_key") && !testColumn.getName().equals("t_map_null_key_complex_value") && !testColumn.getName().equals("t_map_null_key_complex_key_value")).collect(toList());
    assertThatFileFormat(PARQUET).withSession(session).withColumns(testColumns).withRowsCount(rowCount).withFileWriterFactory(new ParquetFileWriterFactory(HDFS_ENVIRONMENT, FUNCTION_AND_TYPE_MANAGER, new NodeVersion("test"), HIVE_STORAGE_TIME_ZONE)).isReadableByPageSource(new ParquetPageSourceFactory(FUNCTION_AND_TYPE_MANAGER, FUNCTION_RESOLUTION, HDFS_ENVIRONMENT, STATS, METADATA_READER));
}