diff options
author | Herman van Hovell <hvanhovell@databricks.com> | 2016-08-02 10:09:47 -0700 |
---|---|---|
committer | Reynold Xin <rxin@databricks.com> | 2016-08-02 10:09:47 -0700 |
commit | 2330f3ecbbd89c7eaab9cc0d06726aa743b16334 (patch) | |
tree | a92bd8441b13fa090a8c4f17fb402700aaefb674 /sql/core/benchmarks | |
parent | 146001a9ffefc7aaedd3d888d68c7a9b80bca545 (diff) | |
download | spark-2330f3ecbbd89c7eaab9cc0d06726aa743b16334.tar.gz spark-2330f3ecbbd89c7eaab9cc0d06726aa743b16334.tar.bz2 spark-2330f3ecbbd89c7eaab9cc0d06726aa743b16334.zip |
[SPARK-16836][SQL] Add support for CURRENT_DATE/CURRENT_TIMESTAMP literals
## What changes were proposed in this pull request?
In Spark 1.6 (with Hive support) we could use `CURRENT_DATE` and `CURRENT_TIMESTAMP` functions as literals (without adding braces), for example:
```SQL
select /* Spark 1.6: */ current_date, /* Spark 1.6 & Spark 2.0: */ current_date()
```
This was accidentally dropped in Spark 2.0. This PR reinstates this functionality.
## How was this patch tested?
Added a case to ExpressionParserSuite.
Author: Herman van Hovell <hvanhovell@databricks.com>
Closes #14442 from hvanhovell/SPARK-16836.
Diffstat (limited to 'sql/core/benchmarks')
0 files changed, 0 insertions, 0 deletions