Github user ajayyadava commented on a diff in the pull request:
https://github.com/apache/orc/pull/160#discussion_r134393721
--- Diff: java/mapreduce/src/java/org/apache/orc/mapred/OrcInputFormat.java
---
@@ -58,7 +58,7 @@
*/
public static boolean[] parseInclude(TypeDescription schema,
String columnsStr) {
- if (columnsStr == null ||
+ if (StringUtils.isBlank(columnsStr) ||
--- End diff --
Hi @dongjoon-hyun,
Thank you for reviewing the patch and pointing me in the right direction. I
had earlier fixed only the exception. Based on your comment, a bit of reading
through the code and the spark test that you have linked in the JIRA my
understanding is that the desired behavior is to not select/project any columns
when the `orc.include.columns` is a blank string. I have added a test case for
what I understand is the desired behavior.
I have left the behavior for null(default) value to be as before which
seems bit inconsistent(null & blank should not be different) but I am not sure
on why it was changed and if it's ok to modify that behavior.
I am completely new to ORC and this is my first patch so my understanding
might be completely off. Please feel free to point me in the right direction.
Thanks!
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
