[jira] [Updated] (ARROW-16431) [C++][Parquet] Improve error message in append_row_groups() when appending disjoint metadata

["Joris Van den Bossche (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/ARROW-16431?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Joris Van den Bossche updated ARROW-16431:
------------------------------------------
    Component/s: C++
                 Parquet

> [C++][Parquet] Improve error message in append_row_groups() when appending disjoint metadata
> --------------------------------------------------------------------------------------------
>
>                 Key: ARROW-16431
>                 URL: https://issues.apache.org/jira/browse/ARROW-16431
>             Project: Apache Arrow
>          Issue Type: Improvement
>          Components: C++, Parquet
>            Reporter: Michael Milton
>            Priority: Major
>
> Currently if you try to append together metadata from row groups with different schemas (?), you get the following error: 
> {code:java}
>   File "/home/mmilton/.conda/envs/mmilton/envs/driverpipe/lib/python3.9/site-packages/dask/dataframe/io/parquet/arrow.py", line 52, in _append_row_groups
>     metadata.append_row_groups(md)
>   File "pyarrow/_parquet.pyx", line 628, in pyarrow._parquet.FileMetaData.append_row_groups
>     self._metadata.AppendRowGroups(deref(c_metadata))
> RuntimeError: AppendRowGroups requires equal schemas.
> {code}
> What would be useful here is to actually pass the schema difference in the error object in terms of which columns disagree. This information should _also_ be in the error message.
> For example if it said:
> {code:java}
> RuntimeError: AppendRowGroups requires equal schemas. Column "foo" was previously an int32 but the latest row group is storing it as an int64
> {code}



--
This message was sent by Atlassian Jira
(v8.20.7#820007)