You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@arrow.apache.org by "Javier Luraschi (JIRA)" <ji...@apache.org> on 2018/11/03 05:40:00 UTC
[jira] [Updated] (ARROW-3693) [R] Invalid buffer for empty
characters with null data
[ https://issues.apache.org/jira/browse/ARROW-3693?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Javier Luraschi updated ARROW-3693:
-----------------------------------
Summary: [R] Invalid buffer for empty characters with null data (was: [R] Invalid buffer for null characters with null data)
> [R] Invalid buffer for empty characters with null data
> ------------------------------------------------------
>
> Key: ARROW-3693
> URL: https://issues.apache.org/jira/browse/ARROW-3693
> Project: Apache Arrow
> Issue Type: Improvement
> Components: R
> Reporter: Javier Luraschi
> Priority: Major
> Attachments: Screen Shot 2018-11-02 at 10.27.11 PM.png
>
>
> I'm hitting this from data coming from Spark while retrieving this data frame:
>
>
> {code:java}
> <database> <name> <temp>
> default a TRUE
> b TRUE
> {code}
>
> Error:
> {code:java}
> error: Failed to fetch data: invalid data in buffer 2
> {code}
> The problem is that is possible to have a NULL character array with offsets set to 0, notice that in the example above Spark returns two batches, so effectively, we are trying to parse:
>
> {code:java}
> <database> <name> <temp>
> b TRUE
> {code}
>
> Where the data array is NULL and the offsets 0s for the <database> columns.
> Here is a snapshot while debugging StringArray_to_Vector:
> !Screen Shot 2018-11-02 at 10.27.11 PM.png!
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)