private ParquetRecordReader<FakeParquetRecord> createParquetRecordReader(
Configuration configuration,
Path path,
long start,
long length,
List<HiveColumnHandle> columns,
boolean useParquetColumnNames) {
try {
ParquetMetadata parquetMetadata = ParquetFileReader.readFooter(configuration, path);
List<BlockMetaData> blocks = parquetMetadata.getBlocks();
FileMetaData fileMetaData = parquetMetadata.getFileMetaData();
PrestoReadSupport readSupport =
new PrestoReadSupport(useParquetColumnNames, columns, fileMetaData.getSchema());
ReadContext readContext =
readSupport.init(
configuration, fileMetaData.getKeyValueMetaData(), fileMetaData.getSchema());
List<BlockMetaData> splitGroup = new ArrayList<>();
long splitStart = start;
long splitLength = length;
for (BlockMetaData block : blocks) {
long firstDataPage = block.getColumns().get(0).getFirstDataPageOffset();
if (firstDataPage >= splitStart && firstDataPage < splitStart + splitLength) {
splitGroup.add(block);
}
}
ParquetInputSplit split;
split =
new ParquetInputSplit(
path,
splitStart,
splitLength,
null,
splitGroup,
readContext.getRequestedSchema().toString(),
fileMetaData.getSchema().toString(),
fileMetaData.getKeyValueMetaData(),
readContext.getReadSupportMetadata());
TaskAttemptContext taskContext =
ContextUtil.newTaskAttemptContext(configuration, new TaskAttemptID());
ParquetRecordReader<FakeParquetRecord> realReader =
new PrestoParquetRecordReader(readSupport);
realReader.initialize(split, taskContext);
return realReader;
} catch (IOException e) {
throw Throwables.propagate(e);
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
throw Throwables.propagate(e);
}
}
public static void setSchema(Job job, HAWQSchema schema) {
hawqSchema = schema;
HAWQWriteSupport.setSchema(ContextUtil.getConfiguration(job), hawqSchema);
}