GitHub user mgaido91 opened a pull request:
https://github.com/apache/spark/pull/22111
[SPARK-25123][SQL] Use Block to track code in SimpleExprValue
## What changes were proposed in this pull request?
`SimpleExprValue` carries some java code which is a rvalue. Now the code is
represented by a String. If the code references a variable, this means that we
are loosing track of its usage.
This is particularly important as the `value` of an `ExprCode` is
(correctly) a `ExprValue`. Thus, if we have current code referencing a variable
in it, we are loosing track of the referenced variable.
So the PR proposes to represent the code in `SimpleExprValue` using a
`Block`, in order not to loose eventual references present in the generated
code.
## How was this patch tested?
added UT/existing UTs
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/mgaido91/spark SPARK-25123
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/22111.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #22111
----
commit 7e91f93bfcd03537f02fdb9ae05b4807cb09dd87
Author: Marco Gaido <marcogaido91@...>
Date: 2018-08-10T14:50:57Z
[SPARK-25123][SQL] Use Block to track code in SimpleExprValue
----
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
