[SPARK-38123] Unified use `DataType.catalogString` as `targetType` of `QueryExecutionErrors#castingCauseOverflowError`

• Type: Improvement
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 3.3.0
• Fix Version/s: 3.3.0
• Component/s: SQL
• Assignee: Yang Jie
• Reporter: Yang Jie
• Created: 07/Feb/22 05:39
• Updated: 10/Feb/22 09:56
• Resolved: 10/Feb/22 09:56

Description

SPARK-33541 introduces `QueryExecutionErrors#castingCauseOverflowError` and 

there are 2 ways to input parameter `targetType` in spark:

• Use `DataType.catalogString` as `targetType`, such as use in `Cast.scala` and `IntervalUtils.scala`
• Use custom literal such as `short`,`int` and `long` in `Decimal.scala` and `numberics.scala`

 

And we can use `DataType.catalogString` uniformly

 

 

Issue Links

• links to: [Github] Pull Request #35412 (LuciferYang)
• links to: [Github] Pull Request #35412 (LuciferYang)