[
https://issues.apache.org/jira/browse/ARROW-6005?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16900102#comment-16900102
]
Hatem Helal commented on ARROW-6005:
------------------------------------
I'm using this issue to add a unittest for this behavior since the reported bug
was fixed in ARROW-6065.
> [C++] parquet::arrow::FileReader::GetRecordBatchReader() does not behave as
> documented since ARROW-1012
> -------------------------------------------------------------------------------------------------------
>
> Key: ARROW-6005
> URL: https://issues.apache.org/jira/browse/ARROW-6005
> Project: Apache Arrow
> Issue Type: Bug
> Components: C++
> Affects Versions: 0.14.0, 0.14.1
> Reporter: Martin
> Assignee: Hatem Helal
> Priority: Major
> Labels: pull-request-available
> Time Spent: 10m
> Remaining Estimate: 0h
>
> GetRecordBatchReader() should
> "Return a RecordBatchReader of row groups selected from row_group_indices, the
> ordering in row_group_indices matters." (that is what the doxygen string
> says),
> *but:*
> Since change ARROW-1012, it ignores the {{row_group_indices}} argument.
> The {{row_group_indices_}} in the {{RowGroupRecordBatchReader}} that is
> created are never used.
> Either the documentation should be changed, or the behavior should be
> reverted. I would prefer the latter, as I do not know how to make sure to
> read a specific row groups anymore...
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
