[BUG] [Spark 4] Exceptions from casting to DATE/TIMESTAMP do not match the Spark's exceptions, with ANSI enabled

[mythrocks]

Description
With ANSI enabled, when an invalid STRING row is cast to DATE, the exception string from Spark does not match that from Spark.

Edit: Looks like this isn't exclusive to Spark 4. Reproducible on Spark 3.5, and likely others.

Repro

sql(" SELECT '7452-68-35' AS str ").write.mode("overwrite").parquet("/tmp/myth/repro")

spark.conf.set("spark.rapids.sql.hasExtendedYearValues", false )
spark.conf.set("spark.sql.ansi.enabled", true )

spark.read.parquet("/tmp/myth/repro").selectExpr(" CAST(str AS DATE) ").show

With Apache Spark 4, the exception looks as follows:

org.apache.spark.SparkDateTimeException: [CAST_INVALID_INPUT] The value '7452-68-35' of the type "STRING" cannot be cast to "DATE" because it is malformed. Correct the value as per the syntax, or change its target type. Use `try_cast` to tolerate malformed input and return NULL instead. If necessary set "spark.sql.ansi.enabled" to "false" to bypass this error. SQLSTATE: 22018

The plugin fails thus:

java.time.DateTimeException: One or more values could not be converted to DateType
        at com.nvidia.spark.rapids.GpuCast$.$anonfun$checkResultForAnsiMode$5(GpuCast.scala:1331)
        at com.nvidia.spark.rapids.GpuCast$.$anonfun$checkResultForAnsiMode$5$adapted(GpuCast.scala:1329)
        at com.nvidia.spark.rapids.Arm$.withResource(Arm.scala:30)

Expected behavior
The exceptions for the same Spark version with/without the plugin should match.

This needs to be sorted out whenever ANSI mode behaviours are tackled.