[PARQUET-1772] [C++] ParquetFileWriter: Data overwritten when output stream opened in append mode

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: cpp-4.0.0
• Component/s: parquet-cpp
• Labels: pull-request-available
• Assignee: Gawain BOLTON
• Reporter: Gawain BOLTON
• Created: 18/Jan/20 21:28
• Updated: 22/Jan/20 16:24
• Resolved: 22/Jan/20 16:24

Description

An arrow::io::FileOutputStream can be opened in append mode.

However, when the output stream is used by the ParquetFileWriter the data already present in the file is overwritten instead of being appended.

From what I can see, Parquet does not have currently the functionality to append data.  As such I suggest detecting when an append is attempted to give an error rather than overwrite existing data.

Issue Links

• links to: GitHub Pull Request #6228