Peng-Lei commented on a change in pull request #32452:
URL: https://github.com/apache/spark/pull/32452#discussion_r629887096
##########
File path:
sql/core/src/main/scala/org/apache/spark/sql/execution/columnar/ColumnAccessor.scala
##########
@@ -145,6 +151,8 @@ private[sql] object ColumnAccessor {
case DoubleType => new DoubleColumnAccessor(buf)
case StringType => new StringColumnAccessor(buf)
case BinaryType => new BinaryColumnAccessor(buf)
+ case YearMonthIntervalType => new YearMonthIntervalColumnAccessor(buf)
+ case DayTimeIntervalType => new DayTimeIntervalColumnAccessor(buf)
Review comment:
Yeah，when we reuse IntColumnBuilder/LongColumnBuilder instead of
YearMonthIntervalColumnBuilder/DayTimeIntervalColumnBuilder to build the column
that contains the data of the YearMonthIntervalType/DayTimeIntervalType data
type, It seems very consistent that use IntColumnAccessor/LongColumnAccessor
instead of YearMonthIntervalColumnAccessor/DayTimeIntervalColumnAccessor to
access the column. But that means we don't need the column type of
YEAR_MONTH_INTERVAL/DAY_TIME_INTERVAL. We use INT/LONG to define the type of
column that contains the data of YearMonthIntervalType/DayTimeIntervalType,
does it cause some confusion? @yao
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
