[
https://issues.apache.org/jira/browse/SPARK-10352?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14963194#comment-14963194
]
Harsh Rathi commented on SPARK-10352:
-------------------------------------
Why this is not a problem ?
I am writing custom explode function.
If I try to use CatalystTypeConverters for type conversions, it gives error in
StructConverter since InternalRow is not added as a case there.
If I don't use CatalystTypeConverters, it gives casting error saying
java.lang.String cannot be cast into UTF8String.
> Replace SQLTestData internal usages of String with UTF8String
> -------------------------------------------------------------
>
> Key: SPARK-10352
> URL: https://issues.apache.org/jira/browse/SPARK-10352
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Reporter: Feynman Liang
>
> Running the code:
> {code}
> val inputString = "abc"
> val row = InternalRow.apply(inputString)
> val unsafeRow =
> UnsafeProjection.create(Array[DataType](StringType)).apply(row)
> {code}
> generates the error:
> {code}
> [info] java.lang.ClassCastException: java.lang.String cannot be cast to
> org.apache.spark.unsafe.types.UTF8String
> [info] at
> org.apache.spark.sql.catalyst.expressions.BaseGenericInternalRow$class.getUTF8String(rows.scala:46)
> ***snip***
> {code}
> Although {{StringType}} should in theory only have internal type
> {{UTF8String}}, we [are inconsistent with this
> constraint|https://github.com/apache/spark/blob/master/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/ScalaReflection.scala#L131]
> and being more strict would [break existing
> code|https://github.com/apache/spark/blob/master/sql/core/src/test/scala/org/apache/spark/sql/test/SQLTestData.scala#L41]
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
