ParquetQueryWorker onChunk should include entire chunk

[bleakley]

If you call parquetQueryWorker({ rowStart, rowEnd, columns, onChunk }), the resulting chunk contains columnData which only contains the rows from rowStart to rowEnd, but ideally you would be able to get the entire chunk. For example if I request rows 0 to 50 because I want to render them, it would be nice to get the entire row group to preprocess. That way I don't have to continue to request rows 51 to 100, etc.