Github user xuchuanyin commented on a diff in the pull request:
https://github.com/apache/carbondata/pull/3045#discussion_r245509909
--- Diff:
integration/spark2/src/main/scala/org/apache/spark/sql/execution/command/preaaggregate/PreAggregateTableHelper.scala
---
@@ -110,7 +110,29 @@ case class PreAggregateTableHelper(
// Datamap table name and columns are automatically added prefix with
parent table name
// in carbon. For convenient, users can type column names same as the
ones in select statement
// when config dmproperties, and here we update column names with
prefix.
- val longStringColumn =
tableProperties.get(CarbonCommonConstants.LONG_STRING_COLUMNS)
--- End diff --
emm, actually this line of code was added to solve the same problem
mentioned in the PR description. Previously we thought it's user's
responsibility to explicitly specify the long_string_columns in the preagg
datamap.
Please @kevinjmh also check this.
---
