[SUPPORT]Error after applyed HUDI-4851 for InSet

[Zhangshunyu]

Caused by: java.lang.RuntimeException: Unsupported literal type class org.apache.spark.unsafe.types.UTF8String HelloWorld
	at org.apache.spark.sql.catalyst.expressions.Literal$.apply(literals.scala:95)
	at org.apache.spark.sql.hudi.DataSkippingUtils$.$anonfun$tryComposeIndexFilterExpr$35(DataSkippingUtils.scala)

after apply HUDI-4851 for InSet, we query the table thorws this error
@alexeykudinkin could you pls have look at this issue? thank you!

[Zhangshunyu]

select count(*),count(distinct word) from test_table where word in ('HelloWorld', 'OK', ... etc. 1000 words here)

[yihua]

@Zhangshunyu to clarify, does the InSet contains a word with special characters that require UTF8?

[xushiyan]

@Zhangshunyu to clarify, does the InSet contains a word with special characters that require UTF8?

@yihua the error stack indicates it's the user-input string HelloWorld taken in as spark's UTF8String caused the problem.

[Zhangshunyu]

UTF8

1. The data type of this column is STRING in hudi table schema
2. the values in IN set are normal string values in SQL .
3. the UTF8String is generated by spark

[nsivabalan]

Fixed it via #6739