BELUGA BEHR created PARQUET-1514:
------------------------------------
Summary: ParquetFileWriter Records Compressed Bytes instead of
Uncompressed Bytes
Key: PARQUET-1514
URL: https://issues.apache.org/jira/browse/PARQUET-1514
Project: Parquet
Issue Type: Bug
Reporter: BELUGA BEHR
As I understand it, the class {{BlockMetaData}} has a {{getTotalByteSize}} and
a {{getCompressedSize}}. However, in the case of {{ParquetFileWriter}}, there
is one place where the compressed size is being written into {{totalByteSize}}
instead of the uncompressed size. This looks incorrect.
{code:java|title=ParquetFileWriter}
private BlockMetaData currentBlock;
...
public void appendRowGroup(SeekableInputStream from, BlockMetaData rowGroup,
boolean dropColumns) throws IOException {
...
currentBlock.setTotalByteSize(blockCompressedSize);
}
{code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)