Github user dongjoon-hyun commented on the issue:
https://github.com/apache/spark/pull/22359
@wangyum @gengliangwang
Does this PR aim a correct behavior for `INSERT OVERWRITE LOCAL DIRECTORY`?
In SPARK-25313, it may make sense because the generated file are under the
table and respects the table definition. However, in this PR, this accidentally
introduces **case-sensitivity** in SQL SELECT statement.
```
The schema should be StructType(StructField(ID,LongType,true)) as we SELECT
ID FROM view1.
```
This case seems to require a new SPARK JIRA issue and more discussion on
the goal. Also, this should be allowed only when
`spark.sql.caseSensitivity=true`.
cc @gatorsmile @cloud-fan
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
