[SPARK-51830] Exception handling for partition datatype conversion call

[Madhukar525722]

What changes were proposed in this pull request?

Exception handling, similar to other call for castPartValueToDesiredType

Why are the changes needed?

Lets say a partition is of int datatype and a partition in legacy table was created as
partition_name=partition_value

Then, when we perform following operations,

val df = spark.sql("select * from db.table");
val modifiedDf = df.withColumn("partition_name", lit(2))
modifiedDf.show(false)
modifiedDf.write.mode(SaveMode.Append).insertInto("db.table")

it throws error, even if spark.sql.sources.validatePartitionColumns=false.

java.lang.NumberFormatException: For input string: "partition_value"
  at java.base/java.lang.NumberFormatException.forInputString(NumberFormatException.java:67)
  at java.base/java.lang.Integer.parseInt(Integer.java:668)
  at java.base/java.lang.Integer.parseInt(Integer.java:786)
  at org.apache.spark.sql.execution.datasources.PartitioningUtils$.castPartValueToDesiredType(PartitioningUtils.scala:535)
  at org.apache.spark.sql.execution.datasources.PartitioningUtils$.removeLeadingZerosFromNumberTypePartition(PartitioningUtils.scala:362)
  at org.apache.spark.sql.execution.datasources.PartitioningUtils$.$anonfun$getPathFragment$1(PartitioningUtils.scala:355)

Does this PR introduce any user-facing change?

No

How was this patch tested?

Re-ran the above failing commands

Was this patch authored or co-authored using generative AI tooling?

[HyukjinKwon]

Could we have a JIRA and unittest? See also https://spark.apache.org/contributing.html

[Madhukar525722]

Hi @HyukjinKwon , addressed your suggestions. Please review